PodSearch

ATP

575: Walk a Little Faster

 

00:00:00   So, are we all returning our vision prose as the news stories seem to indicate is the big wave?

00:00:06   This is what you decide to do to start the show. I was saving this and now you're just going to drop this bomb right up front.

00:00:14   Yeah, you already returned yours?

00:00:16   No, of course not. But I was trying very hard to make it sound believable. I'm a terrible liar, so I don't think I succeeded, but I was trying very hard.

00:00:23   [MUSIC]

00:00:25   No, I didn't return mine. I mean, truth be told, now we are getting into a topic that I don't think you intend to.

00:00:29   But it is exceedingly overpriced. It is exceedingly heavy. I don't think my particular nose construction is terribly compatible with it.

00:00:39   And I'm stubborn and obstinately refusing to use the probably more comfortable two-sided strap. I'm still on the crank strap.

00:00:48   I feel like my particular nose construction is that all of the weight really wants to rest on my nose.

00:00:54   And part of the problem is, and I've heard other people not complain as much about the nose thing, but complain about the fact that the Vision Pro really wants to sit lower on your face than I think I would choose to naturally put it.

00:01:06   And so if I put it where I think is most comfortable, and not on like, I don't know, my nose ridge or whatever, I'm sure there's an anatomical term for it.

00:01:15   But if I sit it where I want to sit it, then it's like, please move the headset down. It's too high. Or I forgot the messages.

00:01:20   But if you have a Vision Pro, you've seen it. And if you've seen any of the videos, you've probably seen it.

00:01:23   That's the Vision Pro equivalent of your hand is covering up the face ID camera on your iPad Pro?

00:01:28   Yep, yep, yep, yep. Exactly. Exactly. I could not think of a better analogy. And so I have to have it ride a little bit lower than I want.

00:01:35   And so for me, all of the weight tends to sit on my nose unless I really crank the crank strap really tight, in which case then I can transition some of that weight onto my forehead or sometimes my cheekbones.

00:01:49   And that's better. But yeah, I was in a room at the library again today. And yeah, after like an hour and a half, two hours, it was hurting my schnoz a little bit.

00:01:59   And comically, Erin hadn't had a chance to try it until this past weekend. We've just been so very busy.

00:02:05   And for her, all of the weight was on her cheekbones. And immediately I knew I should have put the dual strap or whatever it's called on there for her.

00:02:15   But, you know, we just wanted to go plow ahead with what I was trying to show her. And she it was so uncomfortable for her that she ended up like, you know, having her pointers as like load bearing fingers under the Vision Pro just to keep it from like slamming into her cheekbones.

00:02:31   And I mean, I don't think Erin has exceedingly prominent cheekbones. You know, I think, you know, they're regular-ish cheekbones.

00:02:38   Well, this is this is part of what the different like, did you see people linked around that there was a thread on Reddit where the people were compiling basically what the light seal sizes mean?

00:02:48   Oh, I did not see that.

00:02:49   So there's there's, you know, the two digits and then the end of the W at the end. They've figured out like through various sleuthing and trial and error and stuff that it basically encodes like each digit is separate.

00:02:58   It's not like 23 millimeter versus 33 millimeter. Like it's it's the two and the one or whatever. Like those are two different indicators. They can both change up or down. And one of the factors I forgot the specifics, but one of the factors is like whether it like sits high or low on your cheekbones or something like that.

00:03:14   And so chances are this isn't it. This is just, you know, for Erin to comfortably wear it. She would most likely need a different light seal, which I would totally buy that being said, the scan that she did said the same as me.

00:03:27   That doesn't mean that that's, you know, the right answer. It's just the automated scan says she and I have the same one. But it very well could be that if we had all of them arrayed out in front of us, I mean, hell, maybe even not mine would be different.

00:03:37   But I concur that that, you know, it's very likely that maybe a different light seal would work better for her. But certainly if she were to buy her own based only on the experience she had by scanning her face, she would end up with the same one I've got.

00:03:49   I think I think one of the things that's that's hurting the Vision Pro, you know, initial sales and reactions is they've made this system so complicated of how to fit it.

00:04:01   And it seems like the whatever the app is saying should fit you maybe is not always accurate or maybe they just are not good at presenting alternatives to people or whatever it is, it's a highly, you know, fit dependent device.

00:04:14   And I was thinking, like, why do we not hear about this so much with the Quest headsets? I mean, obviously part of it is I think we are applying stricter standards to Apple because everyone does and their stuff always gets more scrutiny.

00:04:28   But I think part of it might also just be like there's such a weight difference. Like Apple chose to make a very high end headset. It's a very heavy headset compared to its competitors.

00:04:37   And maybe therefore it is more sensitive to like different fit adjustments. And I was thinking, like, is it a mistake to have made the Vision Pro in such a way that the light seal or whatever, whatever the fit mechanics of it are, are not adjustable on the device?

00:04:56   Now, obviously, that would introduce more mechanical complexity, probably a little bit more weight as a result of that. But like, would it have been a better choice? I know Apple would never do this, but would it have been a better choice to have like some adjustment knobs or whatever, like on the actual light seal to have it be somewhat adjustable?

00:05:12   I don't know. But it seems like there's a lot of areas where Apple's approach to this is you find your perfect fit or more often we automatically find it for you. Here it is. Period. And if your perfect fit is a little bit different, they're just like, no, this is your perfect fit. Period.

00:05:29   And there's just no alternative. I wonder if that's something that they'll tweak over time.

00:05:34   It's a third party opportunity, though. It's kind of like how third parties sell tips for the AirPods Pro, right? Like all day you can buy foamy tips or plastic tips or tips where you take a mold of your ear.

00:05:44   Because as far as I know, there is no weird DRM or parts pairing with the light shield. It's just fabric and plastic and a couple little magnets. I would imagine that if this product ever becomes popular enough that it can sustain the third party ecosystem for light shields.

00:05:59   Like if they start selling to lots of people, there'll be third party things that you can snap in there that are very differently shaped, maybe are adjustable.

00:06:07   You know, like I think the rigidity of the main screen part is tough to change at this point in the technology curve because there's a lot of stuff going on in there that really kind of has to be carefully aligned.

00:06:18   But the replaceable light shield is actually a good design for, if not Apple, doing this for third parties to say, "Yeah, we'll sell you a thing for less than $200 that goes between you and the main unit. And maybe you can find one of these that fits you better."

00:06:33   I mean, certainly like the Quest line does have many third party, like, you know, head gaskets or whatever they're called.

00:06:39   Head gaskets, yeah. It's in your car engine and it's on your VR headset, too.

00:06:44   Much easier to replace the VR headset ones, though.

00:06:47   Very true. But I don't know. I do think that, I don't know how I got on this tangent, which I believe I did to myself, but I do really like this device.

00:06:56   I wish it was cheaper. I wish it was lighter. But it is really cool.

00:07:01   And since I do feel like it is a compulsory purchase for this, for my job, both in terms of ATP and call sheet, I feel like I kind of had to spend the money.

00:07:13   And I will whine about the money I've spent until the end of time because, hi, have we met? Hi, this is Casey.

00:07:18   And I will edit out half of it until the end of time.

00:07:20   Fair. But I do really love the device. And, you know, it is incredible for media consumption.

00:07:28   I think you and I will never see eye to eye as to whether or not it's good as a secondary monitor for your replacement monitor for your Mac.

00:07:34   And we're going to talk about that more later. But it is extremely, extremely cool.

00:07:38   And no, I don't I don't play. I mean, my time has run out at this point or I think is running out or has run out, but I don't plan to return it.

00:07:45   And I do really like it. It is not a perfect product by any stretch.

00:07:49   And I don't know if I'd go so far as to say I love it quite yet, but I do really, really like it.

00:07:54   And I'm enjoying having it at hand and using it.

00:07:58   See, I think it's cool, but I'm treating mine. I'm ending up using it more like a dev kit than like a product that's going to really have a huge place in my life so far.

00:08:09   Because, I mean, first of all, look, Apple reaped what it sowed. There are no apps. There's no games. There's very little content.

00:08:19   Apple did this 100 percent to themselves. The developers are just largely not there.

00:08:25   So there's not much for me to do with it besides watch movies.

00:08:30   And the reality is I don't watch that many movies. And when I do watch movies, I'm rarely watching them alone.

00:08:37   I'm not sure I'm going to have all use for it. That's not to say that no one else has use for this product.

00:08:41   Just me. Like, I don't have a ton of use for it. I was really hoping the Mac screen angle of it would be very useful to me, especially now.

00:08:49   I'm in this weird housing transitional period where I really want a big monitor, but I can't really fit one in this rental house.

00:08:55   There's all sorts of uses that I thought I might be able to squeeze out of it that ended up not really working that well for me.

00:09:03   So for me, it's really just a dev kit, and it's a dev kit for a branch of my app that I'm not really working on yet because I don't have time because I'm working on the iPhone version.

00:09:13   And I think this is going to be kind of a story we keep hearing, something like that, where it's going to be difficult. Again, over time, it's going to be difficult for developers to justify investing time into the Vision OS platform if there's no customer base for it.

00:09:28   And it's a chicken and egg problem. If there's not a lot of apps and content on it, it's going to be difficult for a lot of people to justify buying one, especially with the other hard-to-justify factors like the price and the single-person experience and some of the version 1 challenges and limitations.

00:09:42   Maybe I'm totally wrong, and maybe this is selling like gangbusters way above Apple's expectations, and maybe they can keep thinking that they don't need developers.

00:09:51   But the reality is what Apple has shown in their actions around app store policy over and over again is they believe that they grace us with a platform full of users, and we should kiss their feet and thank them and give them a third of our money or whatever, you know, because these are their customers they're bringing to us.

00:10:10   Effectively, therefore, we are not really bringing us any value to their platform, that they are bringing the platform full of users, and we should be very thankful for that.

00:10:16   And I think what we're seeing here is maybe this is demonstrating the value of developers to the platform.

00:10:24   If not a lot of people are finding much to do with it, that's because there's not a lot of apps and content for it, and that's because there's not a lot of third-party development for it.

00:10:36   And so maybe, I mean, I know Apple will not change their app store policies, but maybe at least this helps show them a little bit, like kind of where it hurts, because they just did this huge launch of this huge platform, and I think it might be maybe a little below expectations.

00:10:55   It's probably too early to say, but I'm kind of just getting that feeling, like I think they thought there would be a lot more apps than there are, and I think they thought there would be a lot more developer interest than there has been.

00:11:05   Maybe they thought there would be a lot more content deals from the big content publishers and stuff, and maybe that's not really happening as quickly or yet either.

00:11:12   This platform is starving for apps and content. It's starving for them. It's just not there yet, and I hope it's coming, but what if it's not?

00:11:22   Apple's going to have to just make this work largely on their own, because they've done such an effective job of alienating everyone else in the industry.

00:11:31   Other than broad changes to app store policy, like C-level changes, I'm not sure what I would have done differently, but I do think, and I can't help but wonder, if ceding dev kits to a handful of indies would have gone a long way, especially if it was in concert with "talk about it."

00:11:51   Because I think the problem is if they had ceded, let's say, underscore for the sake of discussion, they surely would have said, "Well, here's your dev kit. You may tell no one."

00:12:02   And then that's not really accomplishing much. Yes, then that gets an even more polished version of Widgetsmith in the store on day one than was already there, which is good, but I think the better approach would have been to hand underscore one and say, "Go to town, man. Talk about it. Get people excited if you can."

00:12:19   I mean, not an edict. You know what I mean? If he's going bananas, doing all of this work and having all this fun, and underscore would talk about it, underscore would be effusive about it, not because he's full of garbage, he's not at all, but because he's just a person who finds the good in things.

00:12:36   And so I think, you know, you seed an underscore, you seed a James Thompson or whatever, and you start to build a little more enthusiasm in the indie community. And I think, and again, I mean, admittedly, as I'm saying this, I'm like, "Well, of course you think that because you're in the indie community."

00:12:51   And maybe that's true, but I don't know, I feel like it would have done a lot to build enthusiasm amongst anyone other than Disney and like Unity or whoever it was that I don't even know if they got dev kits, I just have to assume.

00:13:01   But, you know, whoever it is that got the dev kits, it certainly didn't seem to be indies, from what I can tell. And I feel like that was a missed opportunity. And it could have been that they wanted to and they just couldn't produce them in time or something held it up, I don't know.

00:13:14   But it sure seems like if you were doing these labs, which admittedly were pretty controlled, by the way, I went to a lab, but I can't tell you anything else.

00:13:21   Even despite these controlled labs, I feel like if you had those labs, you must have had the quantity of devices at a stage in which they were complete enough that you could have seeded a handful of trusted indies as well as the Disneys of the world.

00:13:36   And I wish they did and they didn't. And so, like you said, they're reaping what they sowed. And there isn't a lot on there. There's really not.

00:13:44   And I think part of this is exacerbated by your average iPad app is really not great on the Vision Pro, in part because all the iPad apps are shown in light mode and all the Vision Pro apps, the native apps, are not literally in dark mode, but effectively in dark mode.

00:14:00   And part of it is because these iPad apps are designed for touch targets that are much smaller than eye targets. And I've found that with almost every iPad app, it's often prohibitively difficult to grab the right target without using a cursor of some sort.

00:14:17   And Slack is to me the epitome of this, and I think I've already brought this up several times, but like changing between different Slacks, which is something I do constantly in that app, is very, very difficult. And I don't necessarily fault Slack for this, but it's just that the targets are too small.

00:14:32   And then it becomes difficult to use Slack. And then it's like, well, I'm just going to get my work done on my computer then.

00:14:37   And so all of these iPad apps are kind of eh. And there's not many Vision Pro apps because none of us had them. And so now what? And it's exactly what you said, Marco. Like, where do you go from here, Apple? What are you going to do?

00:14:50   What we've seen is like, I think the launch of the Vision Pro kind of is that dev kit program. You know, obviously a lot of people are buying them for their own uses, their entertainment use. A lot of it's just kind of status or YouTubers playing with it. But whatever it is, a lot of people are buying them as early adopters.

00:15:08   But I also think a lot of the early purchases are companies and developers who are wanting to start experimenting with it or get on board with it or try to look at porting their apps or whatever. And this is just, I think this is their dev kit program largely.

00:15:23   And it just so happens that you'll also see them like, you know, people watching movies in first class on airplanes as well.

00:15:29   The way it is now with there being almost no software and almost no content, it's not like a failure per se. It just really hurts the argument to buy it. And it really hurts the experience of owning it when, you know, I think people are predisposed to assume any new tech product is a fad that will fail and then laugh at it.

00:15:51   That's a very common thing in media and tech commentary culture. And I think if everyone who buys the Vision Pro at first ends up not using it very much a few weeks later because they kind of ran out of stuff to do on it, that's not great for the reputation of that product and its launch.

00:16:11   None of this should come as a surprise to Apple. Like, they saw coming up to the launch, like they knew how many apps there weren't. You know, they knew which apps were being built native and which ones weren't. They knew, you know, Netflix didn't have their native app submitted to them or whatever.

00:16:26   All the things that are missing, they knew that going into it. So it isn't like this is a surprise to Apple. Again, I hope that they are stepping on the gas behind the scenes in terms of their own content efforts. They're going to have to do most of this on their own.

00:16:41   They're not going to get a lot of help from third-party developers or third-party content makers on this. They have to be making a ton of the 3D content. If you're making a ton of the environment content, any kind of experiential, virtual travel stuff, they have to be the ones to kickstart that themselves.

00:17:00   Because no one else is going to do it with these numbers and with Apple having really alienated so many people over the last decade.

00:17:08   And it's funny, too, and we should probably move on from this, but I think what's kind of unfortunate about it is even though I'm, you know, kvetching a little bit about everything, like this is an amazing device.

00:17:19   Leaving aside the physical comfort, which is a big deal, leaving aside the cost, which is a big deal, like if you can get past that or just forget it for a minute, this is a truly incredible device.

00:17:28   And, you know, the 3D stuff, like consuming a 3D movie in it is very cool, but the immersive content of which there is very little, but we're hearing more and more rumblings that there's more coming, in fact statements even that there's more coming.

00:17:40   The immersive content is, what is the Tim phrase? It's blow away. It really just knocks your socks off.

00:17:47   Was that a forestallism? I think it was a forestallism.

00:17:49   Maybe it was. I think you might be right, actually, now that you say that. I think you're right.

00:17:52   It has caught on since then. It has spread.

00:17:54   I think you might be right. But anyways, the immersive stuff is just, it's unlike anything I've ever experienced. It's tremendous.

00:18:01   And for me, and I'm not saying it's true for you, Marco, or anyone else, but for me, I really like the Mac virtual display thing and, you know, universal control.

00:18:09   It works pretty darn well. It's not perfect, but it works pretty darn well.

00:18:12   And so, and, you know, briefly using this on a train a few weeks ago was amazing.

00:18:17   And so this is a truly incredible, incredible device. And even though, you know, we've kind of accidentally enumerated some of the crappy parts of it, it is incredible.

00:18:28   And whether or not it's the future, it is a future that I am on board with and is super neat.

00:18:34   And I don't want to lose sight of that because I think we're coming across as two grumpy old men, which is accurate.

00:18:38   But there's a good side to this as well that we're not giving, you know, we're not shining enough light on.

00:18:43   Like, it is incredible. And if you are lucky enough to be able to have one, it is very, very cool.

00:18:50   And I really think that there's a lot of potential here. It's just a question of whether or not we'll realize it.

00:18:54   And I think it's going to have a slower start than anyone thought.

00:18:58   Like, you know, because we were just saying a few weeks ago, like, this is going to be, they're going to sell as many as they can make.

00:19:03   It's going to be backordered for, you know, the whole year it's going to be backordered.

00:19:07   Like, you know, it's going to be supply constrained or whatever.

00:19:09   I just looked and I can pick one of these up tomorrow or I can have it shipped to me next week.

00:19:14   That's not good for the sales figures, I think.

00:19:19   Speaking of things with a lot of potential that may or may not have been realized, two different things.

00:19:24   First of all, the 2015 movie Steve Jobs. And second of all, our new member special about that movie.

00:19:29   So we recorded this month's members only special about Steve Jobs, the 2015 movie with Michael Fassbender and Kate Winslet and a bunch of other people.

00:19:39   This is a movie about Steve Jobs, what, four years after he passed away.

00:19:44   And so we, like I said, we did a member special on it.

00:19:47   If you are not a member, Jon, what do you need to do in order to become a member?

00:19:50   You're waking me up from my slumber to pitch the membership program?

00:19:54   Yeah, man. Hey, you could have bought a Vision Pro. You could have been a part of this. You opted out of the conversation, sir.

00:19:58   Well, you just got through telling me why I shouldn't get one.

00:20:00   No, I was trying to bring it back around. That was the whole point.

00:20:03   But if you want to get one, it's really easy.

00:20:05   It sounds like there's no apps for it.

00:20:07   Unless I want to watch Major League Soccer, games that have already taken place, because Apple just announced they're providing that content. Anyway.

00:20:12   See, I put you to sleep before I talked about the good parts, apparently, because there are good parts, for sure.

00:20:16   ATB.fm/join if you'd like to become a member. Not all of our member specials are about movies, but some of them are, and this one is.

00:20:23   Great sales pitch.

00:20:25   It is accurate. No, it was a lot of fun watching this and talking about it, and I don't want to give anything away.

00:20:32   But yeah, remember that if you go to ATB.fm/join, you can join on a monthly or yearly basis.

00:20:39   You can also go to ATB.fm/gift, if I'm not mistaken, to gift yourself or someone else a membership. Hint, hint, hint.

00:20:50   But yeah, we had a lot of fun recording this one, and if you become a member, for any amount of time, you can go back in the history books and listen to any of our member specials.

00:21:00   And you can do that as long as you are a member, so you can check that out.

00:21:03   We've done one a month for almost a year now, I think, or something like that.

00:21:07   I don't have an account in front of me, but we have a fair bit of member specials in the can at this point.

00:21:12   So check it out, ATB Movie Club, Steve Jobs, and we'll put links in the show notes for the relevant information, ATB.fm/join.

00:21:20   Let's start some follow-up at 30 minutes into recording.

00:21:24   Chase writes, "Regarding the blur in the Vision Pro when you turn your head, I'm pretty sure that this is just typical sample and hold display blur.

00:21:31   This also affects televisions and where the idea of motion resolution comes from.

00:21:36   Impulse displays like CRTs and plasma have much higher motion resolution. To combat this, LCDs can use backlight strobing, and OLEDs can use black frame insertion."

00:21:48   And BlurBusters, which apparently is a website I learned today, has a really good resource where you can read about blur of all kinds.

00:21:57   "A lot of headsets, like the Quest 3, use very low persistence to get much better motion resolution.

00:22:05   The downside is in brightness. The upside of the microLED displays and pancake lenses is that it allows the displays to be very close to the eyes, and having the weight closer to the head is better for comfort.

00:22:14   The downside of pancake lenses is that they swallow much more light of the light coming off the displays than Fresnel."

00:22:20   I think it's Fresnel?

00:22:21   Fresnel, yeah. It's those lenses that are flat, but they look like they have a bunch of concentric circle ridges on them.

00:22:26   Okay, or other aspheric lenses as well. "Even if the displays are 5,000 nits, once you add color filters, polarizers, and the pancake lenses, the brightness we see can still end up very low.

00:22:36   The other consumer headset using pancake lenses and micro OLED displays is the big screen beyond 3, which I had never heard of, and it is very dim, especially if you turn down the brightness to get acceptable persistence."

00:22:46   Chase continues, "I believe Apple is pushing persistence further than they should in order to get more brightness back because they want to push HDR as a thing on the Vision Pro."

00:22:56   So on this topic, for the motion blur in motion, I wish I had known to look for that when I had my demo, because I would have.

00:23:03   I didn't notice it, but clearly Marco has, and I've heard it from other people as well.

00:23:08   And I do wonder if, I still wonder if this is what they're talking about. So the sample and hold thing, this happens on OLED TVs as well.

00:23:16   The deal with OLEDs is you light up a pixel and it stays whatever color you made it until you change it, and it changes color really, really fast.

00:23:25   And you're like, "That sounds great. This is a great display technology. What's the problem?"

00:23:28   The problem is, if you watch something like a 24 frames per second movie, it will show a frame, and the whole TV will show that frame, just the exact frame, exactly the way it is, until the next frame comes.

00:23:43   And again, you may be thinking, "That sounds like what it's supposed to do, right?"

00:23:46   Well, not really, because if you think about what a movie projector does, or what a CRT television does, is both of those things will show the frame, and then they'll be, it will basically like blast it onto the screen, like boom, here's the frame.

00:24:00   And then the frame will either fade away or quickly be replaced by black, like with plasma, if you watch it in slow motion, it blasts color at the screen, and then it just fades away.

00:24:08   Sometimes plasmas would blast some of the color, then the second part of the color, and both of those will fade until the next frame appears, and a movie projector would show one frame of film, but then there'll be nothing as the next frame slides into view, and then it will blast that frame on.

00:24:23   So what it's really showing you is bright light, bright light, bright light, bright light, and in between the bright light, there's either total blackness or a fade to black.

00:24:33   And I think what our brain does during these intervals is say, okay, well, there's like a train going across the screen, bright light, oh, there's the train, and then there's nothing, or blackness, and then a second picture of the train appears, and now it's moved a little bit to the right, and our brain goes, oh, in between when I saw that first picture of the train, and then there was blackness, and then I saw the second picture of the train, I guess it must have moved between those two parts.

00:24:55   With sample and hold on an OLED, where it just shows the train in the first position, and just holds it there for 1/24th of a second, and then immediately shows the train in the new position, what it looks like to us, and you will see this on an OLED television if you have it set up "correctly", is it's stuttery, it looks like it's moving in segments, it looks like chunk, chunk, chunk, chunk, it's like, why doesn't it look smooth, I watched the same movie in the movie theater, the train smoothly moves from left to right, but suddenly when I watch it on my OLED TV, it's stuttering or something, and it's not stuttering, it is, carefully, if you have it set up correctly,

00:25:24   it's showing 1/24th of a second, and then 1/24th, and the thing is, it never goes black between the frames, it instantly changes from frame number 1 to frame number 2, instead of showing frame number 1 for a tiny fraction of a second, and then showing blackness, and then showing the next frame.

00:25:39   So this has been a thing for headsets since they've been rolled out, and one of the innovations of Oculus, you can see John Carmack talking about this, we need displays that can blast that frame really brightly for a tiny, tiny fraction of a second, and then fade to black, and do nothing until the next frame is ready, because we want the brain to essentially fill in the blanks, because if we show the frame the whole time, until the next one is ready, even though we can do that with OLED screens, it looks jerky,

00:26:07   because your brain doesn't get a chance to fill in any of the intermediary spots, like if you think about the train moving, you see the train in position 1 and train in position 2, but if you blank out in between then, your brain will fill in train in position 1.1, 1.2, 1.3, 1.4, your brain will fill those in for you, those frames don't exist, but your brain will fill you them in.

00:26:24   But if you never do that, your brain will say, train is in position 1, still in position 1, still in position 1, oh my god it's in position 2, what happened in between, it instantly moved to position 2, and that appears jerky. This is a big thing with OLED televisions, which is why some people say I have to turn on motion smoothing, I can't have it set up "correctly" because it looks wrong to me, because when there's a slow panning shot, I see every one of the 1/24 frames that it looks jerky to me.

00:26:49   So that's what I would expect you would see in the headset if that was the problem, Marco would be saying, I turned my head and everything looks jerky, but that's not what people are saying, they're saying it looks blurry, maybe it's the same thing, maybe it's a misinterpretation, again I wish I had known to look for this, I didn't notice either jerkiness or blurriness, but my eyes aren't great and it was just a half an hour demo, so I'm interested to see how this develops, but this story about Apple pushing brightness rings true to me because things are very bright in there.

00:27:18   And obviously you can get more brightness by holding that image longer and not fading to black as long.

00:27:26   I'm curious John, because when I went from plasma to OLED, I noticed this too, I remember watching The Office, this regular TV show shot, I assume on film or whatever, but I remember panning shots, I noticed immediately the difference, like oh wait motion looks bad on OLED, everything else looked great but motion looked worse.

00:27:47   And my OLED, it's probably now seven or eight years old, so it's nowhere near cutting edge now, but I'm wondering, do modern OLEDs today, are they better with things like black frame insertion, because I believe mine was one of the first ones that supported it, but it just is not fast enough of a TV, like I tried it and it just looked terrible, I could almost see the black frames, it was not smooth enough, is it better now?

00:28:11   So, modern TVs are better at black frame insertion than your old ones, but still not good enough that I think you would ever want to use it.

00:28:18   Because the nature of OLEDs is they change really fast, and I think that makes it harder to do black frame insertion, like the way they do it now to try to make it better is they like double or triple the frame rate so that the black frames are like faster, but OLED's "problem" is they change so fast.

00:28:37   Like LCDs, there's all these tricks we have to play to make the pixels change from one color to another really fast, OLEDs you don't need to play any tricks, they change really fast.

00:28:45   So that means they're changing really fast to blackness, which means you have only a brief time that the light exists, and then immediately it's completely 100% black, and that's why you felt like you could see the blackness, because it's not like it smoothly fades like a CRT, like a dying star, that's not how it works.

00:29:01   So if we crank the frame rate up enough, can it fade to black between each frame?

00:29:06   Yeah, it's really fast.

00:29:08   No, I was saying, can you just insert a faded frame, and then another faded frame, and there'd be a heck of a frame rate.

00:29:16   Yeah, so what you're getting at there is alternate solutions, which is basically what they do, but just to finish the black frame insertion, there are ones now that are fast enough that you can't see it flickering, but they do hit the brightness, and for the most part you don't want to make your TV dimmer, especially OLEDs, right?

00:29:30   Because up until recently, OLEDs have not been, you know, people would consider them not bright enough for a very bright room. Now they are very bright, but then their competition is brighter still.

00:29:40   Anyway, what they actually do, the actual solution to the stuttering problem is, the good televisions have essentially added a form of motion smoothing that is like the most delicate form, like you don't detect it as a soap opera effect, but it smooths that out just enough for it to not look jumpy to you.

00:29:58   So it's basically like motion smoothing, but turned down to the lowest possible setting, lower than you'd ever imagine, and that basically cures the problem, and what that is doing is instead of making faded frames, it is essentially interpolating between frames, but just barely, just barely enough to make it, because it doesn't take much, 24 is close, like if the motion picture, or even 30 frames, which is probably what the office was, you don't need much more frame rate above that, so they don't need to fill in a lot for it to just smooth out.

00:30:27   I still watch mine in straight up 24 frames per second mode, I can still see it on slow panning shots in movies sometimes, but it doesn't bother me as much as it bothers some people, but yeah, that's one of the reasons you want a fancy TV, because they will have a setting that says, "Motion smoothing for people who hate motion smoothing."

00:30:42   And it's like, turn it up to one, or low, or super low, or whatever, and sometimes they also separate the different aspects of motion smoothing, so you don't have to apply both of them at the same time, or you can set them to different values, that is essentially the solution to dealing with 24 frames per second content if you are sensitive to it, or 30 frames per second in the case of the office, I imagine.

00:31:03   Speaking of display technology, and screens, and trade-offs, and the Vision Pro, as we were speaking about a minute ago, there's been this great series ever since the Vision Pro was announced, and everything, there's a blog called "KG on Tech" by somebody named Karl Guttag, and Karl goes through and is very knowledgeable about VR, and displays, and hardware, and optics, and how all this stuff works.

00:31:27   This kind of helps show the various trade-offs involved in the Vision Pro's screens and optics, and what it does, why it is limited in certain ways, what the other, you know, headsets like the Quest and other things, like what they do sometimes the same way, or certain choices they make differently, and why they choose differently, and what the trade-offs are there.

00:31:47   It's very, very interesting. So Karl had this article the other day about some of the trade-offs about the resolution inside the Vision Pro. There was a shot in the iFixit teardown video, the second iFixit teardown, where they actually show the raw image on the actual little tiny screen panel, and the image looks like it's being viewed through a fisheye lens.

00:32:10   It's like, you know, very, it's curved, it's warped, with a whole bunch of, you know, resolution spent on the middle of it, and towards the edges it warps out.

00:32:19   The screens have to kind of warp the image to make up for what the lenses in front of them are going to do as they, you know, project the image around so it looks like a giant field of view around your eye, to give you that whole immersive effect.

00:32:31   Since the image on the screens has that fisheye warping effect, whatever is in the center of the screen has way more resolution than what is in the periphery of each eye.

00:32:46   So I think this kind of helps explain, first of all, some of the, you know, optical effects that you see when you're using a Vision Pro, but also I think this might be part of my problem with the max screen mode.

00:32:58   Whatever you're looking at in the Vision Pro, when you are looking straight ahead, that, it has way more pixels, like per degree, than whatever is on the edge of the screen.

00:33:08   Now if you turn your head, obviously, you're going to turn the high resolution part of the screen towards what you're looking at.

00:33:13   But if you are looking towards the corners of the screen, only by moving your eyes, and not by turning your head, then what you are looking at has way less resolution than what's in the middle of the screen.

00:33:26   And also other, you know, there's other kind of optical trade-offs that you get when you're looking near the edges.

00:33:31   So for instance, when you're looking towards the middle, both eyes can see what you're looking at in their respective screens.

00:33:38   Their screens don't overlap perfectly. The left eye screen can see a little bit further on the left, and the right eye screen can see a little bit further on the right than their opposite screens can.

00:33:50   If you're looking past, like, the overlap area, where, like, if you're looking, you know, far to the left, only your left eye might be seeing that in its screen.

00:33:59   So that's even less information it's getting. And it's in the lower res, you know, warped edge of the screen optic trade-off, you know, pipeline.

00:34:08   So I think maybe my problem with the Mac screen mode is that you don't have to think about that kind of stuff when you're using a physical monitor.

00:34:17   A physical monitor has the same resolution across the whole thing. And you can just move your eyes and not move your whole head, and you will see pretty much the full resolution.

00:34:27   I mean, yeah, your eyes aren't super perfect in all ways either, but they're, you know, I think they have fewer trade-offs than the Vision Pro screens do.

00:34:33   That's what we said last week. I was saying about, like, when you move your eyes, your field of view moves with your eyes.

00:34:38   So even though your eyes only see things that are in focus that are directly in the center, you can move that center.

00:34:43   But when you're in the headset and you move that center by moving your eyes, the screens don't care.

00:34:47   They don't, it doesn't, imagine if that high res center of the screen followed your eyes as you move them around.

00:34:55   Like maybe the screens are motorized, they're saying in front of your pupils or something. That's what happens in reality.

00:34:59   In reality, you move your eyes and, you know, your field of view is just as janky, even probably jankier than this screen, which is why foveated rendering works as well as it does.

00:35:08   But you get to move it wherever you want. And so you can take the dead center, highest resolution part of your eyes and point it at the Apple menu.

00:35:15   But if you keep your head dead straight and you point your eyes at the Apple menu inside the Vision Pro,

00:35:21   yeah, that Apple menu is probably going to look pretty janky because it's on the corner of the screens and the screens didn't move when your eyes did.

00:35:27   I've been trying to figure out, you know, ever since I got the Vision Pro, why is it that everyone else says the max screen is sharp and it's not that sharp for me?

00:35:34   And I think this might be one of the reasons. I think the people for whom it's working well for, maybe they are just moving their head more.

00:35:43   Because they're multi-monitor people like Casey. It's all coming together.

00:35:46   Or maybe they are making the virtual window smaller in the Vision Pro field of view.

00:35:52   Oh, that's not what I'm doing.

00:35:54   But maybe are you turning your head because you're used to having your three 5K monitors, you're constantly turning your head in real life so you're used to it?

00:36:01   I guess. I mean, I haven't really thought about it that much. But yeah, I mean, it stands to reason that's true.

00:36:05   I mean, again, I was using this earlier today and I thought to myself, like, what is Marco talking about?

00:36:12   Supposedly improve this in the 1.1 beta, like the max screen sharing is a little bit sharper for people. So when that update comes.

00:36:20   And just to be clear, I'm not trying to imply that Marco you're full of it or lying or anything.

00:36:24   It's just, it's so funny to me that this device is so personal and based in a way that I don't think that any of us have ever really dealt with.

00:36:33   It's based on your own like body's abilities and physiology. Is that the word I'm looking for? You know what I mean?

00:36:38   Yeah. And habits and like habits like turning your head or not.

00:36:41   Right. And so I think I made a big stink about this last episode. I'll just briefly say again, like I'm not trying to imply Marco that you're wrong or you're lying or anything like that.

00:36:49   It's just so weird to me that your experience does not match mine on what is effectively identical hardware.

00:36:55   And it's just a funny quirk of this brave new world we're entering into.

00:36:59   It turns out our eyes and brains are not identical hardware. Maybe that's the problem.

00:37:04   This can all be solved by screens that are four times as big with eight times the resolution.

00:37:08   No problem. I'm sure we'll be right there.

00:37:10   Hey, you know, we got retina Mac monitors eventually. It's just got to be patient.

00:37:13   I think there's always going to be certain trade offs. You know, there are, for example, like certain principles of optics that are going to make some of this stuff more difficult.

00:37:20   But certainly the higher resolution they can get those screens, you can paper over a lot of those problems or you can you can dramatically minimize them.

00:37:28   Kind of like on the iPhone screens where, you know, I think most modern iPhone screens essentially use the pen tile sub pixel pattern where you don't even get an R, a G and a B sub pixel for every quote unquote pixel on the screen.

00:37:38   Right. Like if you zoom in on the screens that they use this pattern where, like, I forget which one is, but one of the sub pixels is shared with neighboring pixels.

00:37:44   You're like, how can that look good? What a garbage screen. They couldn't even give every pixel an RGB. That must look terrible.

00:37:50   And the answer is no, you can't tell they're so small. And like, nope, nobody notices. Nobody cares. Right. Because they're so small.

00:37:56   Yeah. I mean, even like, I mean, please, designers out there, cover your ears for just this moment. When I like draw icons and think about like font weights and icon stroke widths for my app, you used to have to think about like, are you going to make everything like exactly, you know, like 1.5 points or three points or whatever it was.

00:38:17   So it would perfectly line up on pixel boundaries, integer multiples. Yeah. Like, and so it would look good on retina screens and non retina screens and it would perfectly align with everything. And these days, I don't think about that anymore.

00:38:27   Like once we went to 3x density iOS screens, which I believe happened at the same time. I believe that was all with the iPhone 10 and forward. I have yet to find any stroke width that I choose to use that that looks blurry or bad compared to other ones.

00:38:42   So now I just I just do this mentally. I'm like, all right, I want this this this icon to be semi bold. I want this one to be medium. You know, I just do that and kind of let the system do what it wants with thicknesses. And they are not always perfect integer multiples.

00:38:54   And it turns out it's totally fine because we have such incredibly high density on those screens now. So I think with Vision Pro, I think you're right. Once down way down the road, I don't think this is coming soon, but way down the road, maybe 10 years from now, when they can, you know, double the resolution of the screens or more.

00:39:11   I think a lot of these problems will get a lot less noticeable. But until then, it's just going to be trade offs that we live with. And that's just the reality of the technology we have so far.

00:39:20   Tony de Taranto writes, adding to the categories of ATP listeners with jobs in every profession, I am a professional choral conductor and long time listener to the show. This is my favorite corner. It really is. I had a funny thought while listening to John describe window management VisionOS in episode 574, specifically about how our eyes normally function as input devices, not output devices.

00:39:42   In the course of my work and in my professional training at music school, I've learned an important thing that most people who aren't conductors don't think about when giving cues, conveying information.

00:39:51   It's important what you do with your arms and hands, but it's arguably more important what you do with your eyes. Eye contact and directing your gaze is an even stronger form of communication and control than arm and hand gestures.

00:40:02   From when I first studied conducting technique in college, I had to train my gaze as much if not more than my gesture to achieve the desired result from the group. I just wanted to provide this as an example of how, although I'm no Superman, I feel like I use my eyes as an output device for my work.

00:40:20   Those conductors, they just see all the people in the chorus or the orchestra as tiny instruments to be controlled by their eyes.

00:40:32   Well, there was a line for the movie, wasn't it? They play the music, I play the orchestra or something like that.

00:40:35   No, no spoilers for Steve Jobs, come on.

00:40:37   atp.fm/join. David Schaub writes, "What frustrated me is that dropping files on the Mac OS dock breaks Fitz's law. If you drag a file to the dock, you can drag it past the icons, which, according to David Schaub, is ridiculous."

00:41:02   It is ridiculous. It's a long-standing thing. It's kind of like when you put a folder alias on the dock, you can't drag things into it. It's one of those things that could be fixed in numerous ways. For example, you could just disallow dragging folder aliases into the dock if you're not going to support it.

00:41:16   But it hasn't been because apparently nobody cares. But yeah, dragging things to the dock, we get away with it because in the grand scheme of things, the dock is pretty big on most people's monitor. I bet most people don't have their dock smaller than the menu bar.

00:41:27   But yeah, try it. Take something, if you have a folder in the dock, take a file and try to drag it into the folder and go past the folder. Go all the way to the screen edge and you'll notice the folder is like, "Nope, nothing's on top of me anymore."

00:41:38   It's crappy. I might have even filed a feedback on it years ago. Maybe I'll file one again. I'm sure other people have, but I don't think Apple cares. But it's sad, they should.

00:41:48   Apple, this is wildly unrelated. Apple extends their modem licensing deal with Qualcomm through March of 2027. So to recap, Apple bought Intel's modem business a few years back.

00:42:00   Intel's cell modem business, yeah.

00:42:02   Right. And they said, without saying, "Oh, we're going to make our own modem so we don't have to continue to pay Qualcomm."

00:42:09   And then they realized, "Oh, this is harder than we thought." And they continued to pay Qualcomm. And now they've apparently realized, "No, it's still harder than we thought."

00:42:15   And so they are extending their deal even further. Reading from MacRumors, "Apple has extended its modem chip licensing agreement with Qualcomm through March of 2027.

00:42:23   Qualcomm said today, during its first earnings call of 2024, Apple's existing agreement has now been extended for two years, so we can expect to see Qualcomm modems the next several iPhone generations."

00:42:32   Yeah, not great for Apple. I still think they really should make that modem and maybe think about integrating it into the SoC or near the SoC, but first they got to make one that works. Good luck.

00:42:43   I keep in mind also, just because they have a license to deal with Qualcomm through 2027 doesn't necessarily mean that Apple is not going to ship its own modems before then.

00:42:53   It probably means that, but this could be like Apple is going to need Qualcomm's chips through March of 2027 for some parts of their lineup.

00:43:04   That's true.

00:43:05   So it's possible that next year one of the iPhones gets it, or some variant of it, whether it's the iPhone SE.

00:43:11   Yeah, but they're going to want to start this slow, I think. It's not going to debut in the Pro phone, probably. I think they're probably going to be wary and put it out in the SE first or something.

00:43:20   I don't know what they're going to do, but given how hard it's been, I don't think this is a bet your flagship product on. I don't think they can make enough of them. I guess the TSMC would do it for them or whatever.

00:43:31   But yeah, I think given how this has gone, a gradual rollout of Apple's modems is a good idea. Remember, last time they used Intel modems in their phones, and what was it? You could either get the Intel modem or the Qualcomm one, and nobody wanted the Intel one because it wasn't great.

00:43:47   And that was kind of on their flagship phone.

00:43:49   Yep, I had the Intel one. That was the AT&T iPhone 7.

00:43:53   Not a good one. It's tough, so I hope Apple does take this slow. Keep extending that deal, lock in whatever deal you have to get with Qualcomm. I know these two companies hate each other, but you kind of need cell modems for cell phones.

00:44:06   Indeed. Going back many episodes now, probably five or ten episodes, we were talking a lot about patents, and an anonymous person wrote in to say, "Apple employees are heavily incentivized to file patents. It's one of the few ways to make additional income at Apple outside of your normal job responsibilities.

00:44:21   Apple has an entire department, an online portal, to streamline the process. I can't find the exact numbers right now, but I believe employees receive at least $1,000 if a patent is accepted and approved.

00:44:30   The caveat, of course, is that Apple owns all of the intellectual rights to the invention." I don't mean to sneeze at $1,000. $1,000 is a lot of money, but like…

00:44:39   That seems chintzy, right?

00:44:40   Yeah!

00:44:41   Yeah, for Apple, for instance, that does seem lower than I would have guessed.

00:44:46   That's Apple's M.O. Their whole rep is that they don't pay as much as their competitors because everyone should just be happy to be working there because of Apple, but yeah, $1,000 is like, "Why bother? Give me one share of stock."

00:44:56   We are sponsored this episode by Celtrios. This is a shmup, a shoot-em-up game. They actually sponsored us a while ago, and they are back with even more updates.

00:45:06   So here's the idea. It's a shoot-em-up. So tiny ship, big weapons, waves of foes and power-ups and obstacles, and all of this is brought to you with no ads, no in-app purchases, no subscription, and nothing is locked.

00:45:20   You can play Celtrios over and over again as long as you want with lots of customization options to keep it fresh, and Celtrios is available exclusively for Apple platforms.

00:45:30   Mac, iPhone, iPad, and Apple TV. So Celtrios has 13 different stages. You can start from whichever one you want. You can also, of course, resume play later if you quit the app.

00:45:40   You have a million different possibilities when you're configuring your ship, dozens of abilities, and then, of course, randomizing options, full-screen options, huge amount of ways to play both mechanically and then also just how your ship is configured. It's great.

00:45:54   Celtrios also has a huge high-quality soundtrack with over 45 minutes of music, and Celtrios keeps expanding. There have been over 75 free updates to it so far, and the entire thing was made by an independent developer.

00:46:10   So, of course, this is right in my heart, you know? So Celtrios supports one or two players at a time. On the Mac, you can do various input methods. With iOS, the second player needs a game controller, but hey, it works.

00:46:19   So it's just a great shoot 'em up or shmup kind of game. I love this genre of games. It's so fun. It's like the old arcade games taken to the extreme with all the modern capabilities. It's wonderful.

00:46:29   So get Celtrios if you love traditional shmups or you just want a quality game that's fun to play again and again with none of those usual annoyances of other modern games.

00:46:39   Mac-only Celtrios is available on Steam with a free demo, or head over to the Apple App Store to get Celtrios for iOS and tvOS. Thank you so much to Celtrios for sponsoring our show once again.

00:46:50   Harvey Simon writes, in regards to the magnetically attached Apple Watch bands that are rumors, "Magnets screw with compasses. Apple Watches have compasses. Ergo, Apple is unlikely to add magnetic attachments for Apple Watch bands."

00:47:09   So, I mean, I'm not an expert in magnets, but that makes sense to me. If you're waiting for extremely powerful magnets to be connecting your Apple Watch band, maybe that won't work well with the compass feature of the watch. Time will tell.

00:47:25   I think working around magnetic strength limitations is not that hard, because, again, I'm not a scientist, but I'm pretty sure magnetic strength falls off dramatically with distance. And if you have fixed magnets inside the Apple Watch that are always in the same position, I would expect it would not be super hard to just calibrate that out from the sensor.

00:47:47   Hmm, third-party watch bands might throw a monkey wrench into that. Well, it depends, like, where are the magnets? Are the magnets in the body, or are they in the band? I would imagine they have to be in both places, kind of like the MagSafe cases, you know, like the ones that work well. The cases also have magnets in them, and third-party ones tend to be stronger than Apples in my experience.

00:48:03   Hmm, maybe. But, like, I think if they're in the watch body, which, I mean, actually, does the charging disk on the bottom, are there any magnets on the watch side of that? There probably are, I would imagine. There's electricity, so there's also magnetism. It's complicated.

00:48:17   Yeah. Anyway, so, I think they could design around this. I don't think it would be that big of a deal.

00:48:37   Fair enough. All right, and then, finally, for follow-up this week, more on the line or letter of credit from Brian Coffey. "As someone who also works in commercial banking, specifically problem commercial loans, I need to clarify the comments on the requirements to get a letter of credit."

00:48:51   "Drawing a letter of credit. Yes, Apple could draw on the letter of credit, but in the agreement, they must have a good and valid reason. The typical use for a letter of credit is for international shipping, where you ship the goods before getting paid.

00:49:01   The letter of credit ensures the overseas recipient won't stiff you. The shipper who got stiffed must go to their bank to work with the ship ease bank to claim that the shipper never got paid.

00:49:10   And with regard to getting a letter of credit, a letter of credit is a credit product of the bank. Once a letter of credit is drawn upon, it instantly becomes a loan.

00:49:18   A company can get an unsecured letter of credit if they're producing enough cash flow to show their 'good for the money' if the letter of credit ever gets drawn upon.

00:49:26   If a potential borrower doesn't have strong enough cash flows to support this, there are alternatives. Smallish businesses could put up their homes or retirement accounts as collateral.

00:49:33   You could also be 50% secured or really any percent secured, depending on the strength of your cash flows.

00:49:38   All this is to say, you don't absolutely have to put a million in cash in the bank, but you do have to show a bank that you could reasonably come up with a million if you had to and prove it.

00:49:48   So it's not exactly handing the bank a million bucks or a million euros, but it's not that far away either.

00:49:54   Yeah, I mean, like any bank thing, if you can convince some bank to do something for you, then fine. But yeah, I feel like the bank basically wants to know if you have something that we can get a million dollars from, even if we need to repossess your home or whatever.

00:50:07   So it's not as bad as you must have a million dollars in cash, but I feel like you still kind of have to have some way to get a million dollars for a bank to agree to this, because that's kind of the whole deal.

00:50:17   Unless you have a really friendly bank and it's like, "We like your face. We think it's fine."

00:50:22   Yeah, we'll front you a million bucks because we like your face. That sounds reasonable.

00:50:25   Yeah.

00:50:26   Alright, let's move on to some topics, and of course we have a little bit more Vision Pro to talk about. And Jon, you seem to be very enthusiastic to talk about personas, baby. So what you got?

00:50:37   This is kind of like when we talked about eyesight last week. The reason this feature exists is obvious and I think is not really going to go away.

00:50:45   So personas are the little fake computer people that you use to represent yourself when you're on a FaceTime call, and the reason for them is obvious.

00:50:53   Being on a video call in a Zoom meeting for your work in a FaceTime or whatever, it's so common today, it's a very common part of using computers, that if you've got weird ski goggles strapped to your face, how do you get yourself into a video call?

00:51:07   Do you let your Mac's webcam show your weird ski goggly face? I think people would find that off-putting, even with the creepy eyesight things on it.

00:51:16   So Apple's solution is, "Hey, we'll make a little computer version of you, and when you're in your FaceTime call or your Zoom meeting, the little computer puppet of you will talk."

00:51:24   And that problem is not going to go away until we're actually wearing glasses that just look like regular glasses, which is many, many, many years in the future, if it ever comes in any form.

00:51:36   So you're going to need some way to show your face in these meetings. I guess the other solution is, "Oh, we're having a Zoom meeting. Why isn't your video turned on? Oh, I'm wearing a headset, so you can't see me. I'm not sure that's necessarily going to fly."

00:51:53   And you wouldn't want it to, because there's a lot of bandwidth, a lot of communication that happens from your facial expressions.

00:51:59   So Apple's solution to this is these weird, creepy computer models, and as weird as they are, I think Apple's going to keep plugging away at this.

00:52:12   They've taken a lot of flack for this during the rollout, because they do look kind of creepy and scary.

00:52:18   I don't think Apple is going to be scared away, nor should they, because I think video conferencing is not going away, and it's going to be a long time before these things on our face don't look like sea goggles and hide most things about our face.

00:52:35   So I think they just need to keep plugging away at it, as embarrassing as it is now.

00:52:39   And as for how embarrassing it is, like last episode, when we tooted about it, we showed Casey's persona and the graphic for that episode.

00:52:48   They look silly, but I think you have more appreciation for them if you have any experience with this type of thing before.

00:52:58   And most of my experience with this type of thing comes from video games and similar tech, where they would put your face in the game, make a skin for your player character or whatever.

00:53:08   And the things that have been built into games have been so much worse than what Apple did.

00:53:13   They really are one of the best instances I've ever seen of this particular technology.

00:53:18   And we'll put a link in the show notes to someone demonstrating this.

00:53:21   The thing I think that's most impressive about Apple stuff, well, two things.

00:53:24   One, that they're able to make something that looks as good as it does.

00:53:28   I know you're hearing that and you're like, what do you mean?

00:53:30   They look terrible, look like a death mask.

00:53:32   It could be so much worse.

00:53:34   Yes, they all blur the edges to hide their sins or whatever, but they do an amazing job with a face scan that you can do, not a professional scan.

00:53:41   Go look at things where they have a famous Hollywood actor in a video game and they have that person go into a full motion capture studio and have them stand there with balls on their shoulders and lasers shooting them for a million angers.

00:53:53   It's an incredibly controlled environment where they spend the entire day getting their face scanned and they put them in the game and they look awful.

00:53:59   And this is, oh, just hold the ski goggles in front of your face for two seconds, turn to the side, and Apple's doing a better job.

00:54:04   So that's impressive.

00:54:06   But the second thing is, it's demoed in this video, how well it tracks the expressions you're making with your face.

00:54:13   So you're wearing ski goggles and you're raising your eyebrows and you're blinking and you're twisting your mouth and you're sticking out your tongue and you're smiling and you're frowning and you're furrowing your brow.

00:54:24   And somehow Apple's able to detect all those things. Some of those things are happening inside the headset, some of those things are happening outside the headset.

00:54:32   It is phenomenal what they do.

00:54:34   Now, does the little puppet look like you when it's doing that?

00:54:37   You know, it's got fake, everyone's got the same perfect fake teeth and everyone's got the same weird artificial tongue.

00:54:43   And it's not, you know, if you can make the W shape with your tongue, the avatar is not going to, right?

00:54:47   But like, it does a really good, I think it does a good enough job, kind of like a really well articulated puppet would do, of letting the people who are on the Zoom call with you or whatever, know what expression you're making.

00:54:59   Are you, you know, are you happy about that? Are you skeptical? Are you angry? Are you not paying attention?

00:55:06   Like, I feel like they do an amazing job of matching the movements of the various parts of your face and reflecting those.

00:55:14   No, it's not perfect. It's not capturing all the subtleties of your acting performance. It might not even really look like you, but kind of like I was, you know, talking about the cartoon eyes, sometimes just sort of a, not a cartoon eyes, but like a decomposed, less granular version of you captures a lot of it.

00:55:33   It's the reason animation can look so good. Animation doesn't look photorealistic, but the right lines in the right places can be very expressive in animation.

00:55:41   And I feel like that's what Apple's going for. This isn't exactly you, but if they catch enough, if they put the right lines in the right places, they can convey most of the information your face is expressing.

00:55:52   And again, I'm really impressed that some of that expression is underneath these, the goggles, and some of it is outside and they put it back together into a cohesive whole.

00:56:01   So I don't relish seeing one of these personas talking to me, but I think this is just one of those hard problems that Apple and anybody who wants to do what Apple is doing has to be resigned to tackling over the next several decades, because it's not going to go away again until we just get plain old glasses.

00:56:18   Until then, people are going to want to be in meetings and people are going to want to see their faces and they're going to want to be able to use their faces, again, talk about our coral conductor, as an output device, because that's part of the way we communicate with other people.

00:56:32   You want to be able to scowl at someone meaningfully and have that have an effect on them in the meeting.

00:56:37   Yeah, no, I want to build on what you were saying earlier, like as a technical achievement, it is stunning how good these are at expressing what your face is expressing.

00:56:50   Because remember, you've got cameras on the inside that are figuring out when you're blinking because the persona reflects that, that figure out when you're raising your eyebrows because the persona reflects that.

00:56:59   I don't recall if it reflects where your eyes are looking, but I think it does.

00:57:04   Yeah, it does, pretty sure.

00:57:05   And it's certainly, you can turn your face left and right and up and down and then your smile, like that's happening outside the device.

00:57:12   And granted, there are cameras pretty much everywhere on this thing, but still, it's outside the device that you're smiling or sticking your tongue out, like you had said.

00:57:20   And whether or not, leaving aside the creepiness factor or the uncanny valley, which again, just like I was saying in the beginning of the show, that's a big thing to just push under the carpet.

00:57:30   But leaving that aside, the technical achievement is really just phenomenal and stunning how good it is.

00:57:38   And I do a monthly FaceTime call with James Thompson.

00:57:41   We had one this morning and because we're both idiots, we jumped on the call, unbeknownst to either of us, but I think we both kind of assumed it.

00:57:48   We jumped on the call, started off the calls in our vision pros with our personas and so on and so forth and had a good laugh about it and chuckled about how ridiculous we both look.

00:57:59   And then we hung up and got on the computer like we usually do.

00:58:02   But I think if we had stuck with the personas, it would have been awkward for a few minutes, and this is a very common refrain from people who've done it.

00:58:09   It would have been awkward for a few minutes and then it would have felt pretty normal all in all.

00:58:13   And I think that's how it typically ends up.

00:58:17   I'm not saying it's not weird, I'm not even necessarily saying it's not creepy, but I don't know, you settle into it and again, as a technical exercise, it is beyond compare.

00:58:28   Did you look at the Charlie Chapman video, by the way? The one I was referring to? Did you have a chance to look at it?

00:58:34   I think I skimmed through it super quickly if memory serves.

00:58:37   Some of the things he does, for example, are like puffing air underneath his, like, into his, behind his lips and into his cheeks, and also talking out of like the side of his mouth.

00:58:48   Like, that's what I'm talking about, where it's not just like a puppet where it's like, "I can tell when you're opening your mouth and you said a T sound, so I'll make it look like you're making a T sound with your tongue."

00:58:56   Like, you can do weird stuff with your face that is not normal, like, you know, puffing your lips up and, you know, talking weird, like, and it's, I mean, is it tracking it exactly?

00:59:06   No, but they're accounting for the fact that you might do that, and they're mapping it to whatever their little puppet model is of your face, going far beyond just simply making it so that your mouth moves when you talk.

00:59:16   Please, everyone should definitely look at this video. And again, ignore the fact that his whole head is all fuzzed out and he looks like a weird death mask of himself.

00:59:25   I mean, but I think you are correct at how incredibly impressive this accomplishment is, but it's still not good enough, and I think this is kind of, this is largely the story of Vision Pro in general right now.

00:59:38   Most products, like most new groundbreaking products, you tend to have mostly, like, routine, you know, stuff that has been done before, plus, like, one or two big new challenges. That's not the Vision Pro. The Vision Pro is a very small number of things that have been done before, like, here's an M2-based, you know, iPad-based OS in a computing environment with Windows and stuff.

01:00:02   And then they tackled ten different massive challenges at least, and they have achieved remarkable things. They're way ahead of the industry in so many ways with the Vision Pro.

01:00:16   But the problem is, they're not selling the Vision Pro and pitching the Vision Pro to people in the VR industry. They're not selling and pitching the use of personas to only gamers.

01:00:29   They've tackled these massive problems, and they've done very respectable jobs in their solutions to them, but the problem is, they're still not where most people want them to be.

01:00:43   And they probably have some kind of, like, resentment whenever people criticize it on some level. Like, I bet there's people on Apple who are like, "How can you criticize the personas, because look at how amazing they are compared to the state of the art."

01:00:55   And that's true, but when you're presenting them as an alternative to people, then people are going to hold them to a much higher standard.

01:01:03   Maybe that's why they made the studio display camera so cruddy.

01:01:07   Like, the Vision Pro, in many ways, it's pitching itself as an alternative to various aspects of reality that people are very good at noticing the differences between the real stuff and the fake stuff, and they have achieved remarkable stuff.

01:01:26   And yet, it is still not good enough for what most people expect as, like, the basics. "Oh, you're going to show a virtual version of me? Okay, it needs to look like me." Like, it needs to be a stand-in replacement, and it's sort of, you know, it's in the ballpark for a lot of people, but it's not, you know, a stand-in replacement.

01:01:42   In many cases, it looks creepy and weird, and, you know, still, even with 1.1. And so, I think this is going to be, you know, just part of the Vision Pro's uphill battle over the coming years, is like, even though they have achieved remarkable things, they still need to push it even further than what they've already done to match what most people's expectations are, who are not VR industry pros.

01:02:06   Yeah, I think they picked mostly, like, the right challenges, because there are so many different challenges they could have chosen, and you might look at this, I think a lot of people look at the outside, like, "Why did Apple even try to do this?"

01:02:14   And the answer is, because this is one of the problems they eventually need to solve, and if you don't start working at it now, don't expect you're just going to snap your fingers sometime 10 years from now, it'll be perfect.

01:02:23   You've got to make the janky version first, right? So they do it, like, but they can't ignore this problem. Like, they can't ignore it unless they see a time horizon of just wearing, like, clear glasses instead of these goggle things.

01:02:34   Otherwise, this is just going to be out there as an issue, and so they better start plugging away at it, and I bet this was, like, a big time sink, and a big, you know, like, a lot of technology went into this.

01:02:44   And, to your point, Margo, a lot of technology and time went into it, into a feature that we know regular people are going to look at and go, "Ugh."

01:02:50   I look at it and go, "Ugh."

01:02:53   But it's like, you just got to, like, this is a problem that has to be solved. If you want to make this product, there are a small number of problems that you basically just have to solve, and you're not going to be able to do a great job on them,

01:03:03   even if you do better than anyone else has ever done before, but you better start cracking on it, because next year you make a better version, next year, like, they can't ignore this one.

01:03:10   They can't, like, even more so than eyesight, or eyesight where you can say, "Okay, we'll make the cheaper one without eyesight," or, "We'll use cartoon eyes," or whatever.

01:03:16   Like, you could maybe sweep that one under the covers, but people are on video calls all the time, and I don't think it's acceptable to them to say, either, "You just can't show your face," or, "Your face is going to have these giant ski goggles on it."

01:03:29   And all of those are less acceptable than even this janky thing, so they just have to be like, "Okay, we're going to put ourselves out there, and we're going to say we gave it our best shot, because we recognize this is a problem that we need to solve, and next year we hope we'll do better."

01:03:42   And, you know, it's going to be a while, but like I said, it is impressive from a technical perspective what they've achieved, and I think they struck a reasonable balance in the beginning.

01:03:50   People were like, "Why don't you just use Memojis?" I think Memojis would be worse. Like, I know they have Memojis, I know there's this whole scene kit versus reality kit, political, internal, API thing that they have going on there, but setting that aside, Memojis are not as expressive as personas are.

01:04:04   Again, watch the video, we'll link in the show notes, of Charlie Chapman showing his different facial expressions. Memojis are cartoons that are not what I was using before, the example of animation where you draw just the right lines to be expressive.

01:04:14   Memojis are not that. Memojis are bad, rigid-headed, Chuck E. Cheese, Pizza Time Band, whatever things.

01:04:21   And you know, Slack and the Memoji team, what they did is amazing too. Memojis had to walk so personas could walk a little faster.

01:04:28   But personas are better able to communicate you across a computer while you're wearing that goggles than Memojis would be.

01:04:40   And what people want is personas, but not bad. You could either just never ship, or you could ship what you have, which is personas, which are better than Memojis, but still not good enough.

01:04:55   So I give Apple an E for effort here, and if they have to choose where to add the resources to pursue better things, I would say put more resources into personas than eyesight.

01:05:08   For going forward.

01:05:10   Yeah, I don't know. I stand by eyesight as well, and we don't need to belabor this, but I get why people are turned off by both personas and eyesight.

01:05:21   But I genuinely think this would be a far worse product without both of those components.

01:05:28   We are brought to you this episode by Squarespace, the all-in-one website platform for entrepreneurs to stand out and succeed online. Whether you're just starting out or managing a growing brand, Squarespace makes it easy to create a beautiful website, engage with your audience, and sell anything from your products to your content to your time, all in one place and all on your terms.

01:05:48   Let me tell you how great it is building a website with Squarespace. So I've told you over the years how whenever I or someone in my life needed a new website for something, especially businesses, I would always point people to Squarespace first, and every time they would try it and they would realize, "Oh, I'm done. This is all I need."

01:06:04   This just happened again. A friend of mine runs a small business, and she had some very, very high quote for some custom work, and I said, "Hey, why don't you try it on Squarespace first?"

01:06:14   And she didn't know about this. And the next time I saw her, she was like, "Oh, my God. You just saved me thousands of dollars."

01:06:20   And she's not technical, but she was able to go to Squarespace and build her own site without me telling her how to do it as the nerd.

01:06:28   You literally just point people to Squarespace, and it's so easy. Even non-tech people have no problem figuring it out and then doing it themselves.

01:06:36   So you're actually not only saving yourself as the nerd, you're saving yourself work and time that you might think they would need your help,

01:06:44   but then you're empowering them to build their own site. And every feature, she's asking, "Oh, can you change this?"

01:06:51   "Yes." "Can you change the template, put in your own stuff?" "Yes." "Does it support buying things, scheduling things?" "Yes."

01:06:56   "Does it support image galleries?" "Yes, of course." So much on Squarespace is built right in, and it's so easy anybody can do it.

01:07:03   I strongly suggest you check out Squarespace, and if people in your life need websites, point them to Squarespace too.

01:07:09   Go to squarespace.com to start a free trial. When you're ready to launch, go to squarespace.com/ATP, and you will save 10% off your first purchase of a website or domain.

01:07:19   So once again, squarespace.com, free trial. When you're ready to launch, use squarespace.com/ATP for 10% off your first purchase.

01:07:26   Thank you so much to Squarespace for sponsoring our show.

01:07:29   There was news today that Apple is already defending iMessage against tomorrow's quantum computing attacks. I'm sorry, what?

01:07:43   So Apple and security professionals have probably known about this for a long time, but I don't often have to think about this.

01:07:50   But one of the things that security professionals are thinking about is, "Hey, there will one day be a quantum computer."

01:07:56   And what happens if you somehow, probably, nefariously capture a bunch of encrypted traffic today, but what if you save it off for years and years and years?

01:08:08   And eventually, you finally get your hands or build that quantum computer, and you can go back to all of that years and years and years of data that you captured with our comparatively weak encryption that we use here in 2024.

01:08:25   Well, suddenly your quantum computer can just decrypt all of that, right? That's how it's going to work.

01:08:31   And nobody really knows if that's true or not. It certainly stands to reason that it's true.

01:08:35   And Apple is already starting the process of defending iMessage against tomorrow's quantum computing attacks.

01:08:41   So they had a blog post about this, which I'll link in the show notes, and The Verge covered it, and I will read from The Verge a little bit.

01:08:47   Apple's security team claims to have achieved a breakthrough that "advances the state of the art of end-to-end messaging."

01:08:55   With the upcoming release of iOS 17.4 and Mac OS 14.4 and the equivalent iPad Watch OS, the company is bringing a new cryptographic protocol called PQ3 to iMessage that purports to offer even more robust encryption in defenses against sophisticated quantum computing attacks.

01:09:11   So now from Apple's blog post, today we are announcing the most significant cryptographic security upgrade in iMessage history with the induction of PQ3, a groundbreaking post-quantum cryptographic protocol that advances the state of the art of end-to-end secure messaging.

01:09:23   PQ3 is the first messaging protocol to reach what we like to call Level 3 security, which I'll explain in a second, providing protocol protections that surpass those in all the other widely deployed messaging apps.

01:09:34   To our knowledge, PQ3 has the strongest security properties of any at-scale messaging protocol in the world.

01:09:38   PQ3 employs a hybrid design that combines elliptic curve cryptography with post-quantum encryption, both during the initial key establishment and during re-keying.

01:09:48   Thus, the new cryptography is purely additive, and defeating PQ3 security requires defeating both the existing classical ECC cryptography and the new post-quantum primitives.

01:09:58   I thought that bit was interesting because basically they're covering their butts to say, look, what if we screwed this up? What if we came up with this new, you know, quantum, you know, post-quantum encryption, but we are, we have a bug in the implementation.

01:10:12   We're just rolling this out and oops, we got something wrong and there's some kind of bug or buffer overflow or whatever.

01:10:17   Making this additive to the existing encryption hopefully makes it so that if they really screwed this up and, oh, it's trivially easy to crack this post-quantum encryption because of a bug in Apple software.

01:10:29   Well, once you crack it, what you're left with is the stuff that was encrypted the way it's currently encrypted now.

01:10:34   Like that it's layered on top of it is my understanding, which I think is a really smart thing to do to sort of, you know, we all kind of wish we could do this.

01:10:42   Like you kind of want to have the old way there as a fallback, even if you totally screw up the new way.

01:10:47   I'm not entirely sure if that's true, but my reading this paragraph makes me think that it might be, and it's a clever idea, which is like belt and suspenders.

01:10:55   We're not getting rid of the old encryption. We're just encrypting it one more time, even better.

01:11:00   Yep. Yep. And so Apple, as mentioned, has come up with its own four plus stage, I guess five plus stage level system.

01:11:10   They've defined classic classical cryptography, which is not quantum secure. There's level zero, which is no end to end encryption by default.

01:11:17   And by the way, you can kind of tell it's kind of weird that like Apple's marketing department really doesn't ever really like to name competitors at all ever.

01:11:25   Like, even though we know what they're talking about when they allude to something, but whoever's doing the security blog has no problem naming names.

01:11:31   So underneath the level zero thing, level zero, no end to end encryption, they just name names, Skype, Telegram, WeChat, whatever that QQ is. Yep.

01:11:40   Level one, end to end encryption by default, which includes LineViber, WhatsApp, Signal previously, and iMessage previously.

01:11:46   Level two. Now that now we're in the post quantum cryptography or PQC. This is and also has an encryption by default.

01:11:53   So level two is PQC key establishment only, which is signal with PQXDH, whatever the heck that means.

01:12:00   And then level three, which is PQC key establishment and ongoing PQC rekeying, which is the new iMessage with PQ3.

01:12:08   And then in the future, not given a level, but I guess it would be level four, PQC establishment plus ongoing PQC rekeying plus PQC authentication.

01:12:16   And then there's potentially even more after that. So this is really cool.

01:12:20   I'm happy that Apple was working on this, even though I don't think anyone gives a crap about what I say in iMessage.

01:12:26   But I do think that that's really cool that this is something that they're actively working on, hopefully and presumably so long before it's ever going to be necessary.

01:12:34   It's interesting in the fact that they have like the level four written there.

01:12:37   If you read the big document, the blog post, they explain like that they essentially chose not to do the thing that makes this level four,

01:12:47   because it's just like level three, except for they add the PQC authentication and they explain why they chose not to do that.

01:12:53   They said like, look, we don't want to do that right now.

01:12:58   And we don't think it's necessary right now because it happens. It happens in the moment.

01:13:02   So it's not one of those things. The authentication is like, you know, establishing authenticity of who you're communicating with before you start communicating.

01:13:09   And that has to happen each time. And so what they say is like, this is not like you could save this for later and then decrypt it later because you're not there's no useful information exchange yet.

01:13:18   So saving that, saving that, you know, that exchange is useless to you because that exchange has already happened.

01:13:24   So even if you crack the encryption out with your quantum computer 10 years from now, it's pointless.

01:13:27   There's no data there. And that conversation happened long ago and those keys are all useless. Right.

01:13:32   And so they said, well, until someone can get a quantum computer that can intercept your traffic and then in real time, crack it.

01:13:38   We don't have to worry about that. So that's why they push it off to the future.

01:13:41   And the other thing they talk about in implementation tradeoffs is the post quantum cryptography stuff. The data is bigger, like the keys are bigger and whatever, whatever other info they have to exchange is significantly bigger than their old encryption scheme.

01:13:54   And their old encryption screen scheme sends new keys like every single message.

01:13:58   But to like basically tamp down on the bandwidth use of, you know, if you're going to tack something onto every message that sent that adds up real fast because a lot of my messages are sent per hour, per day, per minute, per second.

01:14:09   Right. So they say with the post quantum stuff, they only are going to do like the rekeying periodically.

01:14:17   And they do it based on an algorithm of like if you're on a crappy connection, we won't try to shove these new keys down as fast as we normally do.

01:14:24   If you're on a faster connection, we'll send them more frequently. And what they basically said is we're trying to narrow the window that an attacker could do something.

01:14:32   So if an attacker someday cracks this, the only thing that they'll be able to see is the brief period before we rekeyed.

01:14:40   They cracked just that little segment. So I don't know how big that is. They didn't actually say whether it's like two lines of text or five minutes or whatever.

01:14:46   But that's another tradeoff they made. And that's another fun thing of like seeing security people write something instead of marketing people, because they'll tell you here were the engineering tradeoffs and here's why we made them.

01:14:58   Instead of just saying this is the best and no one else has anything like this and it's super secure, let engineers write it and they'll tell you about the tradeoff.

01:15:05   So I thought this was super interesting. I highly recommend everybody read the blog post we link because it is, it might seem like it's kind of got a little bit of technical jargon, but they do a really good job of explaining it well enough for you to follow what they're saying.

01:15:17   And it's pretty cool stuff.

01:15:19   All right. It's been a little while, so let's do some Ask ATP. And let's start tonight with Ian Nalkuszewski, who writes, "What is your advice on how to best communicate to non-tech people the value and benefits of native Mac-ass Mac apps?

01:15:33   I work at a small under 10 people company where everyone works on a Mac, but some are new to the platform and most of their software experience is using Electron apps and other apps that are at best so-so citizens of the platform.

01:15:44   We're contemplating hiring a developer to make an app to help us with an internal project and want to be able to make the case to hire someone who knows how to make good Mac software.

01:15:51   If I put an app like Fantastical next to Outlook, most of my teammates just see two calendar apps with cosmetic differences and shrug off the idea that there's anything notably different between them.

01:16:00   I'd love any advice on how to make the case to non-technical people that Mac-ass Mac apps have a real user-facing benefits beyond just feeling better."

01:16:07   Hot take, I don't know that you really want a Mac-ass Mac app in this context. If you're just writing stuff for your own team of 10 people, I wouldn't spend the time personally.

01:16:17   And I know that's probably going to make everyone shudder and hate me, but there are bigger problems and more important problems to solve than making a Mac-ass Mac app.

01:16:24   But that's my opinion, wrong as it may be. Marco, correct me.

01:16:28   I don't think I'm going to disagree with you on this. It depends so much on the nature of the app that you're going to build. Ian said this is an internal app.

01:16:39   I think it will be challenging for the higher-ups to justify what it would take to make a really good "Mac-ass Mac app."

01:16:49   And what that means is basically native Mac code, native Mac controls, the standard Mac UI design paradigm, things like that, as opposed to things like Electron.

01:17:03   And I think for most software, let alone most internal-use software, it's very difficult to justify that kind of investment on the Mac.

01:17:13   Because, first of all, Ian mentioned wanting to hire someone who could do that. That's difficult.

01:17:19   There's not a lot of programmers out there who are qualified to make this style of high-quality, native, traditional-style Mac app. It's a very small talent pool.

01:17:31   Including inside Apple.

01:17:33   Yeah, Apple can't even make them anymore.

01:17:35   Sorry, but it's true.

01:17:36   It is. So problem number one is, could you even find someone to do this? Problem number two is, would you be able to pay them what they are probably worth?

01:17:44   And then problem number three is, can you convince the higher-ups in your company that that's worth doing?

01:17:49   And I think the only way that is really easily done is if the higher-ups in your company are Mac nerds and also not good business people.

01:18:01   Because what you're ultimately looking at, and by the way, I'm both of those things, so if it was my internal app, I would absolutely do this.

01:18:10   But the problem is, when you're talking about having that style of app with the realities of today and the markets and the tech needs around it today,

01:18:21   it's more of an indulgence than something that you can make a good business case for.

01:18:27   So if you're able to convince them, "Hey, indulge me in having this thing that we're going to build be very nice in these ways that you don't care about but I do,"

01:18:38   then good for you, that's great. I think you're in for an uphill battle.

01:18:42   And then also, even if you can get someone to build it and you can get the higher-ups to agree to indulge you in this,

01:18:50   what happens down the road when you have to change it or update it? How hard is it going to be to get someone in to do that down the road?

01:19:00   Because it's already hard enough now. So I think it's going to be a tough sell.

01:19:05   These two just don't have enough experience dealing with giant corporate bureaucracies. Let me tell you how to do this.

01:19:10   So the first thing you need to do is establish things that are not in this question that aren't in the answer to.

01:19:20   For example, this seems to imply that you're going to hire a developer to make an app for an internal project and that you only need a Mac app.

01:19:27   If that's really true, confirm that and say, "Just so we're clear, we're not planning on making a Windows version of this app later.

01:19:34   It's an internal app. Are we ever going to need a Windows version? Do we need a Linux version?" And if you can get past that hurdle of clarifying the requirements and they say,

01:19:42   "No, no, no. We're never going to make a Windows version. We're never going to need a web version. This is going to be a Mac app.

01:19:47   It's an internal thing. We're just only ever going to run it on Macs. There's never going to be any other version."

01:19:52   Then you're set. Because then what you can do is not say what Marco just said because that will discourage them.

01:19:58   What you have to say then is, "Okay, if it's going to be a Mac app, the reason we want to hire an experienced Mac developer is because

01:20:07   the straightest path to make a Mac app is to use Apple's frameworks in a straightforward way.

01:20:14   No weird custom stuff. No, I've come up with a framework of my own for doing GUIs.

01:20:19   Just use..." And you'll have to hash this out. Do they want you to use SwiftUI? Do they want you to use AppKit? Whatever it is they pick.

01:20:25   Whatever path you go down, find someone who will do that in the most straightforward way possible.

01:20:33   And the pitch is, when this person disappears, I want anybody to be able to look at this and say,

01:20:39   "Oh, this is a straightforward SwiftUI app, straightforward AppKit app that doesn't do any weird custom stuff,

01:20:45   that has no custom controls, that Apple's documentation explains how to do it, that it's really easy to find,

01:20:52   example code documentation, that it is straightforward and they're going to be done faster."

01:20:58   This was the old pitch with the Next stuff, when it was Next and Objective-C and everything.

01:21:03   "You can make a highly functional app with fewer lines of code and in less time because the frameworks do so much for you."

01:21:11   That's the pitch you make. Your goal is, "I want to get an experienced Mac developer in."

01:21:17   And the pitch is, we need someone who's like, "I don't even know what APIs Macs offer.

01:21:24   I'm just going to use the Quartz 2D drawing API and draw my own GUI because I don't know what this whole AppKit thing is.

01:21:29   It's confusing to me, but I'm a good programmer, so I'm going to make my own UI framework like Lauren Brickter out of OpenGL."

01:21:35   That is not what you want. And so that's how I would pitch it, I would say.

01:21:39   And you're never going to get to know the culture and the idioms, but at the very least you should get someone in there.

01:21:44   For example, if you decide it's going to be AppKit, get someone who knows AppKit.

01:21:48   I know it sounds dumb, but I've seen corporate hiring and it's like, "Ah, you can figure it out. It'll be fine."

01:21:53   And someone comes in and says, "Oh, I've made a Mac app," and they show you an Electron app.

01:21:57   And it's like, "Mm, did you make a Mac app, though? Are you really a web developer?

01:22:01   Are you using React Native and Electron and all these other things?"

01:22:04   That's not what we're looking for.

01:22:06   And that, I feel like, will be harder to wrangle because if you're making some kind of internal tool app

01:22:12   and someone uses Electron and it takes 500 MB of RAM when you just launch the thing, that's another case against it.

01:22:17   Just find a Mac developer who can make a straightforward, simple thing that will be done quickly,

01:22:23   have lots of functionality, and be easy for any future Mac developer to understand.

01:22:28   I'm not saying it's a slam-dunk case, but that's your best shot.

01:22:31   All right, sorted.

01:22:33   Julian Gamble writes, "If you could ask Apple for one new API to help your apps this year, what would it be?"

01:22:38   Julian's guesses are Marco for Overcast, "WatchOS, an API to make syncing files, like podcast files,

01:22:44   work on demand and reliably on schedule and, in general, much easier."

01:22:47   We can stop there. I don't even write a watch app in a minute. I'm ready.

01:22:51   I don't think an API can change physics in the size of the watch's battery.

01:22:56   And I know Marco says they should loosen up a little bit, but in the end, an API that did that would burn your battery pretty badly.

01:23:02   Okay, first of all, this would not be my pick for this question.

01:23:06   But just for the sake of argument, what I would want in this area would be,

01:23:10   if a user has initiated a download while the app is in the foreground,

01:23:16   let me start a backroundable download that begins immediately and uses Wi-Fi if it has to.

01:23:24   Because right now, it will, you know, wait for a while and maybe do it later when I saw the charger,

01:23:29   or it'll use the Bluetooth connection to the phone because it's lower power.

01:23:33   And if a user, while using the app in the foreground, initiates a download, that signifies pretty clear user intent.

01:23:41   I want this to happen, like kind of now.

01:23:44   I'd make a pretty strong argument for that, but honestly, that is not my biggest problem on watchOS.

01:23:49   My biggest problem on watchOS is every few days when I get an email from a customer saying,

01:23:54   "Why don't I support the double tap gesture on the Series 9 and Ultra 2 yet?"

01:23:59   And the answer is, there is no API to do that. Apple released the double tap feature in the fall,

01:24:05   and they said, "Hey, third party apps, you can just let this do the default response on notifications."

01:24:12   And that's it.

01:24:14   So literally, there is no API to response double tap, and it's such a glaring omission

01:24:20   that my customers assume that I'm the one being negligent, not Apple.

01:24:24   So that would be my number one request on watchOS.

01:24:28   But that isn't even my number one request overall.

01:24:31   My number one request at this moment is for SwiftUI's list to have feature parity with UITableView.

01:24:43   Now, that is not a small thing.

01:24:46   No, it is not.

01:24:47   However, that would be a huge improvement to the most coding that I'm doing this year.

01:24:54   Julian said this year, that's what would help me is make SwiftUI lists have more capabilities.

01:25:00   For instance, one that I ran into most recently is the kind of drag to reorder mechanic in it,

01:25:07   does not support multiple items.

01:25:09   In UITableView, you can pick up multiple items as you drag your finger around with the second finger,

01:25:14   and then drop them all in one spot.

01:25:16   The equivalent in SwiftUI does not support that.

01:25:21   The API appears to be written to support it, because when it tells you drop here,

01:25:27   it passes you a set of indexes to drop there.

01:25:32   So the API seems to be built to support multiple things being dropped,

01:25:38   but there is no physical implementation of that.

01:25:41   It's just one case of many where I keep running into areas where SwiftUI list is oddly limited

01:25:48   in ways that UITableView is not, and UITableView is a super important API for iOS.

01:25:55   Almost every app uses TableViews in some way, and UITableView has been added to over time like crazy,

01:26:02   because it's such a huge part of interaction on iOS.

01:26:06   There's tons of features that UITableView supports.

01:26:09   SwiftUI list kind of started from scratch and kind of did the basics,

01:26:12   and has been very slowly and carefully adding little bits and pieces.

01:26:16   Oh, you want to customize this inset?

01:26:18   Okay, here's one way to do that.

01:26:20   You want to customize whether this border shows up over here?

01:26:23   Okay, fine, we'll give you another small way to do that.

01:26:25   But there's some big areas, like the multi-select drag and drop,

01:26:29   that it just has not enough feature parity, and that is making my life difficult

01:26:34   as I'm trying to work on the SwiftUI rewrite for the biggest part of my user base,

01:26:40   like the iOS app and the TableViews within it.

01:26:43   Those are major areas, and I just keep hitting walls that just aren't there yet in SwiftUI.

01:26:49   It rocked my world when, I want to say it was like four or five years ago,

01:26:54   it was maybe more than that, but somebody described all of professional iOS development

01:27:00   as turning JSON into TableViews, and I was like, "Mother of God."

01:27:05   That is exactly right, so if that doesn't mean anything to you.

01:27:08   So JSON is a plain text way of transmitting data.

01:27:11   Typically when you get something from a web server, not as a person but as an app,

01:27:17   you're going to get data back in JSON format, J-S-O-N, and like Marco said,

01:27:22   TableViews run the majority of all iOS apps.

01:27:26   So you could summarize iOS development as turning JSON into TableViews for money.

01:27:30   Web development too these days. Not TableViews, I guess, but same thing.

01:27:33   Yeah, exactly.

01:27:34   You get a JSON API, you get the result, you lay it out on an HTML page.

01:27:37   Exactly, so anyway, that rocked my world even though it was five years ago.

01:27:41   I still think it's hilarious and accurate.

01:27:43   John, continuing with Julian.

01:27:45   John, for front and center for macOS,

01:27:48   an API to enable preserving icon arrangement in folders as per classic macOS,

01:27:52   you know, pre-Mac OS X. That's Julian's guess.

01:27:55   That's misunderstanding what would be required to get that.

01:27:58   It's not an API that's missing.

01:28:00   The Mac Finder application would have to behave differently.

01:28:03   There's no API they could roll out that would make that work, unfortunately.

01:28:06   Back to what Marco was saying about lists.

01:28:08   I was telling him before, when we were talking about this in Slack,

01:28:11   they should count as blessings because a list was so under-featured

01:28:16   and buggy for me in SwiftUI and macOS that I can't even use lists

01:28:20   in my thing that has a list of things that are re-orderable,

01:28:23   and I had to basically roll my own list.

01:28:25   It's like having to roll your own UI TableView because TableView is too janky for you.

01:28:28   So it could be worse, but yeah, it could definitely be better.

01:28:31   What I actually want for an API to help make my app better this year,

01:28:35   I actually filed feedbacks on this in the fall of last year,

01:28:39   a whole bunch of them.

01:28:41   I'll just read off the titles of them because I tried to separate it into,

01:28:44   I don't know why I bother,

01:28:46   but I tried to separate it into things that I hope would be,

01:28:49   like if I put them all in one feedback,

01:28:51   they'd say, "Oh, we're not doing all this crap."

01:28:53   I tried to break it down, like maybe they'll pick one of them.

01:28:55   So the first is add a modern window list API,

01:28:58   an API in macOS that lets me list all the windows on the system.

01:29:01   There are existing APIs that do that,

01:29:04   but they're super old, most of them are deprecated,

01:29:06   and they're just terrible.

01:29:07   Well, on your computer they would just crash.

01:29:09   And when I say list the windows,

01:29:12   I don't mean get the windows contents like screen capture kit.

01:29:15   I don't mean control them like accessibility APIs,

01:29:17   which are also kind of old and cruddy.

01:29:18   I mean just literally list them.

01:29:20   List them, which apps own them,

01:29:22   what are their sizes, what are their positions.

01:29:24   It sounds so simple.

01:29:25   You can do it with the 16 APIs,

01:29:27   but Apple really doesn't want you to,

01:29:29   and there are a bunch of caveats and a bunch of this stuff is deprecated.

01:29:32   The second one was add a system-wide window layering and visibility API.

01:29:36   So you can list all those windows and you know where they are

01:29:40   and how big they are and which app owns them.

01:29:42   Wouldn't it be great if you could tell them to come forward,

01:29:44   go to the back, go behind some other window?

01:29:46   Again, still I have no idea.

01:29:47   These APIs have no idea what's in the windows.

01:29:49   They can't see the window contents at all.

01:29:51   They don't want the window contents.

01:29:52   They're just a bunch of anonymous rectangles

01:29:54   that are owned by processes, right?

01:29:56   An API where you could tell them what to do.

01:29:58   And the Mac window layering APIs are extremely limited.

01:30:02   People are shocked to hear that you can't do something as simple

01:30:04   as take one window and change its layering

01:30:06   because those windows belong to other applications,

01:30:08   and my application has a very limited ability to screw with windows

01:30:11   in other applications.

01:30:13   Again, modularly accessibility APIs,

01:30:15   which are a whole other can of worms that I file other bugs with.

01:30:17   And let's see, the ones I get even more ambitious,

01:30:22   add support for window manipulation extensions.

01:30:24   So the idea, this is more of kind of more ambitious,

01:30:28   but like the idea that when you're doing operations

01:30:31   with the Windows server, like moving windows around

01:30:33   or clicking on them to bring them to the front or whatever,

01:30:36   that there would be a plug-in system

01:30:38   where you could affect that interaction.

01:30:40   Once again, still having no idea what's in any of these windows.

01:30:43   All you know is, hey, a window is being moved.

01:30:45   It's owned by this application.

01:30:46   It's in this size and it's in this position.

01:30:48   I'm about to move it to here.

01:30:50   Is there anything you'd like to do to modify that operation,

01:30:52   like say snapping to a grid

01:30:54   or giving you a chance to draw a bunch of guides

01:30:56   like you're in a graphics application?

01:30:58   You know, like you could do so many cool things with this.

01:31:00   Apple is currently not doing them.

01:31:02   Third-party applications try to do them,

01:31:04   but it's so hard because macOS fights you at every step.

01:31:07   So anyway, my answer to this is basically like a modern,

01:31:10   you know, Swift savvy,

01:31:13   like not something that's in core foundation,

01:31:17   not something that's an ancient carbon API,

01:31:19   but a modern API that lets you participate

01:31:23   in the window management system in macOS,

01:31:26   changing window layers, knowing where they all are,

01:31:29   and being able to do things as they're moved around.

01:31:32   That would be a dream, not just for me and for my apps,

01:31:34   but I think if Apple provided those APIs,

01:31:36   all those apps that are out there now

01:31:38   that try to do this with like the accessibility APIs

01:31:41   and the janky ones that we have now

01:31:43   would become so much better, so much more full-featured.

01:31:46   I've said this in past programs, Stage Manager,

01:31:49   if Apple did my wish list of APIs here,

01:31:53   Stage Manager should have been something

01:31:55   that a third-party could have implemented,

01:31:57   because if you have all the APIs

01:31:58   and a third-party had the idea of like,

01:32:00   "I think it would be cool if Windows worked like this,"

01:32:02   a third-party should have been able to implement Stage Manager.

01:32:05   As we know, a third-party

01:32:06   could absolutely not implement Stage Manager.

01:32:08   Only Apple could do it,

01:32:09   and now only Apple can ever make it better,

01:32:11   and only Apple -- we have to wait for the next idea

01:32:13   that Apple has in five years.

01:32:14   Third-parties are out there with lots of cool ideas

01:32:16   about window management on the Mac.

01:32:18   They can't implement them,

01:32:19   because Apple does not provide robust enough APIs,

01:32:22   and like I said, you can do pretty much all of these

01:32:25   with zero access to the contents of any of these windows,

01:32:28   so it is privacy-preserving,

01:32:30   but it is, you know, annoying for Apple to implement,

01:32:32   because they may think this is not important,

01:32:34   but I'm like, "Apple,

01:32:35   your heart doesn't seem into this window management thing.

01:32:37   Every once in a while, some team manages to squeak something out,

01:32:39   and it gets added to the giant pile of Mac window management stuff.

01:32:42   Let third-parties do it.

01:32:43   We'll figure out what works.

01:32:44   Just copy whatever the most popular app is in five years."

01:32:48   That's what I want.

01:32:49   -Fair enough. Julian continues.

01:32:51   "Casey, this is a tough one.

01:32:52   It has to be Apple and not another company.

01:32:54   For VisionOS, an API and Apple TV for Call Sheet

01:32:58   to pull out the current movie or show name

01:33:00   and be able to prompt with the movie show info if required."

01:33:03   Yeah, kind of, but really,

01:33:05   the one thing I would kill for right now is,

01:33:08   I want to have a way to ask

01:33:10   any of the Apple TVs on the same network,

01:33:12   "What are you playing right now?"

01:33:14   And I get why that isn't a thing,

01:33:16   because somebody like Facebook would use it

01:33:18   for nefarious purposes,

01:33:19   but maybe you could have --

01:33:21   I mean, there's so many other freaking user prompts,

01:33:24   like, you know, Windows Vista-style.

01:33:26   Why not prompt some security prompt that says,

01:33:28   "Hey, is it cool if Call Sheet looks at what you're watching?"

01:33:32   And I would love that, and it'll never happen,

01:33:35   but that would be what I would want.

01:33:36   Please and thank you.

01:33:38   -Why don't you just do what everybody else does,

01:33:40   which is just have microphones or visual systems...

01:33:43   -Yeah, right, exactly.

01:33:44   -...detect what's playing

01:33:45   and look it up Shazam-style to figure it out.

01:33:48   -Yeah, that's what I should do.

01:33:49   Winnie Lewis writes,

01:33:50   "Could Marco please best first favorite Phish concerts?

01:33:54   I wasn't aware Phish had a lore,

01:33:56   and I'm curious where to start."

01:33:57   -Do you need me to explain best first favorite, Marco?

01:33:59   We should explain it to a listener anyway.

01:34:00   -Yeah, please. -Yeah, so best first favorite

01:34:02   is a thing from one of my other podcasts,

01:34:04   "Reconsolable Differences."

01:34:05   My co-host, Merlin, came up with it.

01:34:07   It is the idea that when you're trying to discuss a thing

01:34:13   like the band Phish or a television show

01:34:16   or, you know, a set of movies or whatever,

01:34:18   you are challenged to come up with

01:34:20   which one of these things you think is the best --

01:34:22   the best Beatles album, for example --

01:34:24   which one you think someone who is new to the Beatles

01:34:26   should listen to first --

01:34:28   so what Beatles album should I start with --

01:34:30   and what is your favorite Beatles album.

01:34:32   And they may all be the same thing,

01:34:33   or they may all be different.

01:34:34   So best first favorite Phish concert

01:34:36   is what's the best Phish concert,

01:34:37   what's your favorite Phish concert,

01:34:38   and if someone is new to Phish and you had to tell them

01:34:40   this is the concert you should start with, which one is it?

01:34:43   -And I don't really have a good answer.

01:34:45   So the reason Winnie Lewis asked this question is

01:34:48   it was shortly after New Year's this year

01:34:52   'cause Phish did a really, you know, fan service,

01:34:55   amazing thing for their New Year's Eve concert

01:34:58   that if you've been a Phish fan for a very long time,

01:35:02   this was an especially, you know, big one

01:35:04   that played upon a whole bunch of, you know,

01:35:06   stuff in Phish lore from forever ago

01:35:08   that fans really enjoyed.

01:35:10   And I don't really know how to tell you

01:35:12   to get started in Phish

01:35:13   in a way that you would appreciate that

01:35:15   'cause that's kind of like saying, like,

01:35:17   "Oh, here we have a podcast with over 500 episodes.

01:35:20   How do I go about getting all of the old references?"

01:35:23   You could go listen to all 500 episodes of our show,

01:35:27   and by that point, you will understand all the references.

01:35:30   That's a bit of a commitment,

01:35:32   and I wouldn't necessarily recommend that most people do it

01:35:35   just because that's quite a lot to tackle.

01:35:37   So that New Year's one could be your favorite, though, still,

01:35:40   or it could be the best.

01:35:41   Phish has 40 years they've been playing together,

01:35:45   so no one is going--

01:35:47   Like, you're not gonna go into a band with a 40-year history

01:35:51   and get all the context to understand all the different lore.

01:35:55   I don't even know a lot of it,

01:35:57   and I've been a very diehard fan of the band

01:36:00   since, like, 2007-ish, 2008-ish.

01:36:04   Much of the old Phish lore goes over my head, even,

01:36:07   because I wasn't there in the '90s

01:36:09   when all the fandom was really building up.

01:36:12   So I'm actually not even qualified to answer this question,

01:36:17   and I have purchased the live download

01:36:20   of every show they've done since 2009.

01:36:22   You can answer "favorite" for sure.

01:36:24   Answer "favorite."

01:36:25   I can't do that either.

01:36:27   I don't frequently listen to old shows.

01:36:30   I'll pull one up occasionally.

01:36:32   And, in fact, one of the reasons

01:36:34   why I want to make a jam band listening app,

01:36:37   as I discussed in previous episodes at some point,

01:36:39   you know, the app that will have an audience of four people,

01:36:42   and so I should never make this app,

01:36:44   but one of the features that I want out of this app

01:36:47   is kind of a way to, like,

01:36:49   deep mine my collection of old Phish concerts

01:36:53   for, like, gems of shows,

01:36:55   and I have a couple of ideas on how that could be made more interesting

01:36:57   and things like that,

01:36:58   but the reality is what I usually am listening to

01:37:00   is the last few months' worth of live shows.

01:37:04   So, of course, that's a rolling window.

01:37:06   They're literally doing a show right now.

01:37:08   I cannot wait until tomorrow morning.

01:37:10   I can download it and listen to it.

01:37:11   Usually that's what I'm doing,

01:37:12   is I'm listening to whatever the last few months of shows are

01:37:16   I'll kind of go through,

01:37:17   and as I'm going through a show,

01:37:19   I will give star ratings in the Mac music app,

01:37:23   formerly called iTunes,

01:37:24   anything that I, like, rate a certain level,

01:37:26   I'll kind of go back and revisit more often,

01:37:28   and then I have a playlist in iTunes/music

01:37:31   called Best of Phish.

01:37:33   Any real standout songs, not shows,

01:37:36   I will add to that list.

01:37:38   It's not a huge list,

01:37:39   but then whenever I'm somewhere,

01:37:41   if I'm like, "Oh, shuffle my Best of Phish playlist,"

01:37:43   like, then I know I'm going to get some real,

01:37:45   you know, rockin', you know, standouts,

01:37:47   but I don't, even I,

01:37:49   am not qualified to tell you which entire performance

01:37:52   is best or my favorite,

01:37:55   and I would say for first,

01:37:57   your best off for first Phish concert,

01:38:01   just finding any of them.

01:38:03   Like, if you go, so the service that they release them through

01:38:07   is called LivePhish.

01:38:08   You can buy a whole show worth of MP3s there

01:38:11   for, I think it's 10 bucks for a whole night.

01:38:14   You can also try, they have their own streaming service

01:38:18   that spotlights certain shows.

01:38:19   You can stream the whole catalog from it.

01:38:21   They have a free trial on that streaming service.

01:38:23   So you can even just try, like,

01:38:24   sign up for the streaming service for a month

01:38:25   and just play some shows.

01:38:27   It's, I think it's, you know, 12 bucks or whatever.

01:38:29   Like, it's not, we're not talking a ton of money here.

01:38:31   That's what I would suggest.

01:38:32   Like, get into it that way,

01:38:33   'cause that's, the way most fans get into Phish

01:38:37   is maybe you'll hear one of their studio albums or two.

01:38:40   Like, you know, you can go to whatever music streaming service

01:38:42   you already have and listen to some of their studio albums,

01:38:45   but what the band really is about is live shows.

01:38:48   So getting any live show exposure

01:38:51   will give you an idea of what this band actually is.

01:38:54   Some of them are available on YouTube for free.

01:38:56   Some of them you can get

01:38:57   on certain streaming services for free.

01:38:59   Most of them you're gonna have to go to live Phish to get,

01:39:01   'cause most of them are not released

01:39:02   through the official, like, streaming service channels.

01:39:04   And then the second thing is, if you're into this,

01:39:07   go to a concert.

01:39:09   Many people, if you don't get into Phish

01:39:11   through listening to the albums first,

01:39:13   you usually get into Phish

01:39:14   because someone brought you to a concert

01:39:17   and you enjoyed it.

01:39:18   So that's what I would suggest, like, for first.

01:39:22   Go to the streaming services and listen to whatever you can.

01:39:24   Live generally is preferred to studio albums.

01:39:28   And go to a show if that's, you know,

01:39:30   if that's your jam, so to speak.

01:39:32   And that's it.

01:39:33   Unfortunately, I don't have a good answer to, like,

01:39:34   what are my things,

01:39:35   'cause my things are always shifting around.

01:39:37   I don't have one show I go back to all the time.

01:39:40   I'm constantly just listening to whatever is recent,

01:39:42   whatever I can get.

01:39:43   - This is a very on-brand for Phish,

01:39:45   and for the band, and for you in particular,

01:39:47   to essentially not be able to name best first or favorite,

01:39:51   because it's all just music, man.

01:39:53   - That's not what I said.

01:39:54   But that is the result.

01:39:56   Thank you to our sponsors this week,

01:39:58   Celtrios and Squarespace.

01:40:00   And thanks to our members who support us directly.

01:40:02   You can join us at atp.fm/join.

01:40:05   And we will talk to you next week.

01:40:07   (upbeat music)

01:40:10   ♪ Now the show is over ♪

01:40:12   ♪ They didn't even mean to begin ♪

01:40:14   ♪ 'Cause it was accidental ♪

01:40:16   ♪ Accidental ♪

01:40:17   ♪ Oh, it was accidental ♪

01:40:19   ♪ Accidental ♪

01:40:20   ♪ John didn't do any research ♪

01:40:22   ♪ Marco and Casey wouldn't let him ♪

01:40:25   ♪ 'Cause it was accidental ♪

01:40:27   ♪ Accidental ♪

01:40:28   ♪ Oh, it was accidental ♪

01:40:29   ♪ Accidental ♪

01:40:31   ♪ And you can find the show notes at atp.fm ♪

01:40:36   ♪ And if you're into Twitter ♪

01:40:39   ♪ You can follow them at C-A-S-E-Y-L-I-S-S ♪

01:40:44   ♪ So that's Casey List ♪

01:40:46   ♪ M-A-R-C-O-A-R-M ♪

01:40:49   ♪ Auntie Marco Arment ♪

01:40:52   ♪ S-I-R-A-C ♪

01:40:54   ♪ USA, Syracuse ♪

01:40:57   ♪ It's accidental ♪

01:40:58   ♪ It's accidental ♪

01:41:00   ♪ They didn't mean to accidental ♪

01:41:03   ♪ Accidental ♪

01:41:05   ♪ Tech podcast so long ♪

01:41:10   - So I went on a small adventure.

01:41:13   I mentioned a while ago,

01:41:16   and a couple listeners have called me out on it recently,

01:41:18   I mentioned that I may or may not have recently

01:41:21   bought a higher resolution camera

01:41:24   in preparation for Vision Pro content.

01:41:28   I was tempted for a while

01:41:29   to go with a higher resolution camera system,

01:41:31   and I did.

01:41:33   And what pushed me over the edge

01:41:36   happened around the time that Casey visited me

01:41:40   in New York in the fall.

01:41:42   There was a reason Casey came to New York in the fall,

01:41:44   we got pizza together and did a couple other things.

01:41:47   - That was the only reason, was the pizza.

01:41:49   - For some reason during that trip,

01:41:51   I was inspired that maybe

01:41:53   I should start capturing higher resolution photos,

01:41:57   especially interested in panoramas.

01:41:59   - Was it because of that really janky jpeg

01:42:01   that your phone took of you on the beach

01:42:02   at night trying to get your car unstuck?

01:42:05   - No.

01:42:06   (laughing)

01:42:08   - We talked about it on the show,

01:42:09   just how terrible that thing was.

01:42:10   It was like, man, you know what I need?

01:42:12   I need a higher resolution camera.

01:42:14   (laughing)

01:42:17   - No, so anyway,

01:42:18   so there was an event that Casey and I attended in November,

01:42:22   and anyway, so I decided I needed to capture

01:42:25   more high resolution content.

01:42:27   Coincidentally, not at all related to that,

01:42:30   I got a Vision Pro recently,

01:42:32   and I was able to view my panoramas

01:42:36   and my other photo and video content in the Vision Pro

01:42:39   and was disappointed in the resolution of them

01:42:42   when viewed at that scale,

01:42:43   because of course, you're looking at,

01:42:45   I'm looking at phone captured panoramas from old iPhones

01:42:49   from like five, six, seven years ago

01:42:51   being displayed in this virtual 100 foot tall view

01:42:55   I'm seeing inside the Vision Pro,

01:42:57   so I decided let me see what I can do

01:43:01   with higher resolution stuff,

01:43:03   and I wanted for a while to get into the Fuji GFX line.

01:43:08   Last year I mentioned how I had fallen in love

01:43:14   with cameras again because I had discovered Fuji cameras,

01:43:19   and I just love the way Fuji's render color,

01:43:23   especially just the out of camera JPEGs

01:43:26   that I was able finally to take pictures that I loved

01:43:30   without having to mess with them in Lightroom

01:43:33   and everything first.

01:43:34   I was just super thrilled with just the out,

01:43:37   the straight out of the camera performance

01:43:39   of Fujifilm's cameras,

01:43:42   and this kind of was inspired

01:43:44   because Tiff wanted the X100V for her birthday last year.

01:43:47   Yes, I know they just released the sequel to that

01:43:50   like yesterday, I'm very much aware of that,

01:43:52   thank you very much.

01:43:53   So I got Tiff an X100V for her birthday last year.

01:43:55   I got to try it a few times, fell in love with it,

01:43:58   got myself an X-T5.

01:44:00   The X-T5 is an amazing camera in so many ways.

01:44:04   It is by far my favorite handling and controls

01:44:08   I've ever had on a camera.

01:44:10   It has a whole bunch of knobs,

01:44:12   but for all the main stuff to adjust,

01:44:14   you don't have to go into menus,

01:44:15   you don't have to hold buttons and turn a wheel

01:44:17   and hope something changes.

01:44:18   No, it's just knobs, it's wonderful,

01:44:20   and I love the X-T5.

01:44:21   The only thing with the X-T5 that's a little bit of a downer

01:44:24   is that, well there's two things.

01:44:26   Number one, Fuji's autofocus system is not as good as Sony's.

01:44:29   We talked about this, I don't want to go too far into this.

01:44:31   The autofocus is not as good as Sony's,

01:44:33   and also because they are APS-C sized crop sensors,

01:44:36   they don't have the very high resolution

01:44:39   or the very low light abilities of full frame sensors.

01:44:43   Well it turns out, Fuji makes a larger sensor camera system.

01:44:48   They jumped right over full frame,

01:44:50   have never made one as far as I can tell.

01:44:52   Instead they make medium format digital cameras.

01:44:55   And yes, there's some asterisks on what that means,

01:44:57   but generally this is what is accepted as digital

01:44:59   for what medium format means.

01:45:01   So this is basically the next step up

01:45:04   beyond full frame in sensor size.

01:45:07   So you have the little crop sensors

01:45:09   that many of the small mirrorless cameras use,

01:45:11   then you have full frame,

01:45:13   which is what the really nice cameras use,

01:45:15   and then above that you have medium format.

01:45:17   The sensor is just giant.

01:45:19   And I've been eyeing this for a while,

01:45:21   if I could get amazing resolution

01:45:23   and low light performance with Fuji's colors,

01:45:27   that would be the best of everything.

01:45:29   Problem is they're super expensive.

01:45:31   Most of the GFX cameras are in the $5,000 and up range.

01:45:38   Oh my word!

01:45:39   And the lenses are also very expensive.

01:45:42   And don't forget, the cameras are usually pretty big.

01:45:44   Yes.

01:45:45   Well, it turned out in the holiday season this past winter,

01:45:51   there were two factors that led to a substantial discount

01:45:55   on one of the cameras.

01:45:57   One was it was the holiday season

01:45:59   and everybody was doing sales and BS like that.

01:46:02   Another was that Fuji was about to release a new model,

01:46:05   the GFX 100 II, I believe,

01:46:08   and that pushed down in the lineup and in price the GFX 100 S.

01:46:13   That's the one I got because it was on super sale.

01:46:17   And I knew it was about to be replaced,

01:46:19   but the GFX 100 II, the new one that came out,

01:46:22   was adding things that I really don't need or care about

01:46:27   and was at a price point I would never have gone for.

01:46:30   But what it did was push the GFX 100 S

01:46:33   down to surprisingly affordable, relatively speaking, price points.

01:46:38   And because I'm not a professional photographer,

01:46:42   I don't need a bunch of giant expensive lenses.

01:46:45   I got the smallest lens in the system,

01:46:47   which is its version of a 40-ish millimeter pancake.

01:46:52   It's quite large, but it is the GFX version of that.

01:46:57   It's the 50 millimeter f/3.5.

01:46:59   So I've been playing with this, I've been shooting with this

01:47:02   for the last two months or whatever it's been since I've gotten it.

01:47:05   It is amazing.

01:47:07   And it is, kind of like many things in technology,

01:47:12   it is a massive set of trade-offs.

01:47:15   The camera, as John said a minute ago, is very large.

01:47:18   Now, it is very large in terms of, like,

01:47:21   if you look at today's mirrorless cameras,

01:47:24   they are very compact compared to what, like, good SLRs used to be.

01:47:30   If you actually compare the GFX 100 S with this lens

01:47:34   to the previous era's DSLRs

01:47:38   that were of similar professional use cases,

01:47:41   like, for instance, the Canon 5D line,

01:47:43   this camera is almost exactly the same size class as that.

01:47:47   It's very, very similarly sized and weighted to a Canon 5D

01:47:51   with, like, a medium aperture lens on it.

01:47:54   So it's big by today's camera standards,

01:47:58   but if you go back even just ten years ago,

01:48:01   it was considered -- it's a normal-sized camera

01:48:04   for that time range for professionals to handle.

01:48:07   And I am incredibly happy with it in most ways.

01:48:10   It is an incredibly slow camera.

01:48:13   Like, just -- so it -- I should say,

01:48:17   it takes 100 megapixel images.

01:48:19   Oh, my grief.

01:48:21   So do Android phones, Marco.

01:48:23   They're probably the same thing, right?

01:48:24   Oh, totally, yeah.

01:48:25   Android phone takes 100 megapixel images?

01:48:27   Why did you bother getting this thing?

01:48:29   And the size of the sensor -- like, when I had to,

01:48:32   you know, open the camera up and mount the lens,

01:48:35   and I got to see that sensor, it is ridiculous how big it is.

01:48:39   It's so massive.

01:48:41   But the result is you get not only incredible resolution,

01:48:46   but just as going from a crop sensor to full frame

01:48:51   comes with a substantial increase in light sensitivity,

01:48:55   and so you get, like, much lower noise

01:48:57   and higher resolution and better color even in low light,

01:49:00   you have that same jump again going above full frame

01:49:04   into this medium format sensor.

01:49:06   So full frame is already great compared to the small sensors,

01:49:09   and this is even that additional step above that.

01:49:13   And so it is amazing.

01:49:15   I can shoot ridiculous things handheld.

01:49:18   I can crank the ISO up to, like, you know, ISO 24,000,

01:49:21   and it still looks amazing.

01:49:23   Like, it's just -- I'm very happy with this.

01:49:25   But I thought, like, hey, what if I look at some of these pictures

01:49:28   in the Vision Pro?

01:49:29   And that took me a while to finally hook up

01:49:31   and, you know, figure out, oh, panoramas.

01:49:34   There is no, like, metadata that says this photo is a panorama.

01:49:38   But if you crop any photo down to be very wide,

01:49:43   if you give it a very wide aspect ratio,

01:49:46   I think a little bit wider than 16 by 9 at the minimum,

01:49:49   it will display it in the panorama section,

01:49:51   and it will enable a panorama-like view mode in the Vision Pro.

01:49:55   Not exactly the same, though.

01:49:58   It basically gives you -- it doesn't give you, like, the full 180 view.

01:50:02   It kind of just gives you, like, a larger in-front-of-view view.

01:50:06   But it was enough for me to see the effect.

01:50:09   There were two key takeaways for me with this camera

01:50:12   and with trying to use a professional camera to shoot Vision Pro content.

01:50:16   Number one, the resolution does matter a lot.

01:50:21   The content of those pictures looked way better

01:50:26   than any panorama my phones have ever shot.

01:50:29   It wasn't even close.

01:50:31   Now, that is not surprising.

01:50:33   It's not a fair comparison.

01:50:34   These are massive sensors with amazing optics in front of them.

01:50:37   So, of course, you would expect that.

01:50:40   So, I'm not saying that the iPhone camera is crap.

01:50:42   Like, it's just a totally different beast, what I'm talking about here.

01:50:45   So, that's unexpected, and that's fine.

01:50:48   However, what I also learned is that the lens I have for the camera

01:50:54   being a roughly 40-ish millimeter equivalent focal length

01:50:58   is totally wrong for this kind of use, because it's just not nearly wide enough.

01:51:03   The panoramic display in the Vision Pro assumes that you have a very wide field of view.

01:51:10   Again, it wasn't even showing it in the full width.

01:51:12   It was kind of cropping it in and giving me just a bigger, regular window.

01:51:16   I'm not sure that might require some kind of phone-only metadata

01:51:19   to give it the full 180 view.

01:51:22   I think if you just made it really, really narrow, it might do it.

01:51:25   Oh, maybe.

01:51:26   The key thing to know about the iPhone panoramas is it's not just one capture.

01:51:29   You have to rotate your phone or whatever, and so it's many, many captures,

01:51:33   which makes it a way wider field of view in the horizontal axis

01:51:37   or whichever direction you're moving.

01:51:39   I mean, you follow the little arrow or whatever.

01:51:40   So it's kind of a shame that Fuji doesn't have a panorama mode

01:51:44   where essentially you take your giant medium-format camera

01:51:47   and you slide it around the horizon just like you would do with your phone,

01:51:50   and it does multiple 100-megapixel captures and stitches them all together just like a phone,

01:51:54   because that's what you want for--

01:51:56   I mean, I know you're saying, "Oh, if I just got a 5-millimeter lens

01:52:00   or some really wide-angle thing that it might kind of do the equivalent,"

01:52:04   but I think you'd get the best results if you literally did a panorama

01:52:08   by stitching together multiple exposures from your big medium format

01:52:11   into an actual panorama, and that would have even more megapixels in it.

01:52:15   Yeah, I think that's the way to go, and I think ultimately

01:52:19   this is probably the kind of thing that would be done in software.

01:52:22   I would assume Lightroom or something probably has a feature like this

01:52:26   where you can stitch together multiple exposures into one big panorama.

01:52:31   Sometimes they'll do--some cameras will do it in the camera.

01:52:33   Some cameras--I think Sony would force you to use some janky third-party Sony app to do it.

01:52:37   And some of them will let you export the exposures and do it in your own app,

01:52:41   but it's per camera brand.

01:52:43   It's another area where if Apple did it, it would be much--you know,

01:52:46   Apple did do it, and it's way simpler.

01:52:47   What do you do? You just move the phone, and it just does it.

01:52:49   But lots of camera companies do have some way to do this.

01:52:53   I'm just not familiar with Fuji's solution.

01:52:55   I would assume that it would definitely not involve the camera,

01:52:58   because the camera--

01:52:59   It has nothing in it, yeah.

01:53:00   Well, I mean, the camera has a lot in it,

01:53:02   but it's already very sluggish to capture and process these giant images

01:53:06   off this giant sensor.

01:53:07   Like, there is no way it has the power.

01:53:09   Compared to the iPhone in terms of how much computing is in there

01:53:12   versus how much computing is in iPhone.

01:53:14   Oh, yeah. But even--I mean, jeez, just imagine how much memory it must take

01:53:18   to stitch together photos from that sensor.

01:53:21   I mean, because the raw--I don't have the number offhand,

01:53:24   but I think the raws are, like, 200 megs each.

01:53:27   There's a lot of data being used by--you know, being generated by the sensor.

01:53:33   It's ridiculous how many pixels you have to process.

01:53:38   So I would expect this to be no small feat for software and hardware

01:53:42   to stitch those together.

01:53:43   But I think it's interesting.

01:53:45   But ultimately, even with the amazing hardware of that camera,

01:53:51   it is not suitable to do that job

01:53:53   of stitching together for panoramas for the Vision Pro.

01:53:56   What you ultimately want for the Vision Pro panoramas

01:53:58   is just better iPhone cameras and better iPhone processing.

01:54:01   That's it.

01:54:02   Because the iPhone is doing the capture in ways that no camera will ever do

01:54:07   in terms of, like, you know, being able to stitch things together so perfectly,

01:54:10   it's so easily, you know, maybe with imperfect input,

01:54:13   maybe in varying light levels across the frame.

01:54:16   Like, if you happen to sweep across the sun,

01:54:18   like, how do you deal with that?

01:54:19   With the camera, you can fix the parameters,

01:54:22   but then you've got to kind of expose for the sun

01:54:24   and make everything else darker.

01:54:25   Like, it's just this whole thing.

01:54:26   Like, the iPhone takes care of so much of that for us

01:54:30   and makes it so easy.

01:54:32   And then once you're in the Vision Pro looking at it,

01:54:34   you don't have to worry about did I properly match the, you know,

01:54:39   the perspective with how I process this photo

01:54:42   so it will display correctly in the Vision Pro.

01:54:44   It doesn't matter.

01:54:45   When you do it with an iPhone, it's always properly matched.

01:54:48   It always handles that for you.

01:54:49   So ultimately, I think we're still going to just be limited by,

01:54:53   in practice, by what the iPhone can do for the sharpness

01:54:58   and resolution and possibly depth

01:55:00   of what we're looking at in the Vision Pro.

01:55:02   -Surprised you didn't get one of those white monoliths

01:55:05   in Alicia Keys' studio.

01:55:06   That might have been cheaper than that thing.

01:55:07   Whatever, like, dual-camera system they use to capture that,

01:55:10   obviously it's not still images, but, like, that's --

01:55:12   When I think about "Contra for the Vision Pro,"

01:55:14   I remember what I said in my demo,

01:55:15   I was much less impressed by the panoramas

01:55:17   because they weren't 3D.

01:55:19   Like, the right and the left eye

01:55:21   didn't see different perspectives on the same objects.

01:55:23   They were a flat 2D billboard wrapped around me,

01:55:26   which is cool and would be better if the resolution --

01:55:29   if there was more pixels so it didn't look kind of, you know,

01:55:31   kind of fuzzy.

01:55:33   But what was so much more compelling to me were the,

01:55:36   you know, the video that was captured

01:55:38   with two different perspectives from two different cameras.

01:55:41   And I don't know how expensive the things

01:55:43   in Alicia Keys' studio were, but they seemed smallish.

01:55:46   And I know when they're doing the Major League Soccer,

01:55:49   apparently Apple recorded a bunch of those matches

01:55:52   with some fancy cameras.

01:55:54   Those are probably bigger and scarier

01:55:56   than the discrete little white pillars

01:55:59   with two little black dots on them.

01:56:01   And maybe they're more like the cameras we were talking about

01:56:05   when we talked about 3D movies on the last episode,

01:56:08   where it's two gigantic, expensive film cameras

01:56:11   arranged in a really weird way through a prism

01:56:14   or a beam splitter so that they can both get

01:56:17   the perspective they need, you know,

01:56:19   because the cameras are so big,

01:56:20   they can't get that close to each other.

01:56:22   But either way, I find that content much more compelling

01:56:25   than I would even an infinite-resolution panorama

01:56:28   because the panorama still just looks to me

01:56:30   like a giant painting that I'm looking at,

01:56:32   as opposed to actually being in the mountains, you know?

01:56:34   Oh, yeah, totally.

01:56:35   Like, the panorama is in the Vision Pro to me.

01:56:38   Yeah, it looks like I'm in a planetarium.

01:56:40   Like, okay, there's a big, static image.

01:56:42   Okay.

01:56:43   Planetarium's usually better than that,

01:56:45   but, like, you know what I mean.

01:56:46   It looks like you're in a dome and it's painted on.

01:56:48   Like, that's great for certain things,

01:56:51   but it's nothing like a 3D environment.

01:56:54   So Canon just released,

01:56:56   they kind of released this, like, double-eyed lens

01:56:59   intended for capturing VR180 format,

01:57:02   which is what the, like, the high rope thing and,

01:57:05   was it the shark that's VR180?

01:57:07   The rhinoceros.

01:57:08   Yes, anyway, so Canon just released a lens for that,

01:57:10   and they have a special mode with one of their highest-end,

01:57:13   new mirrorless cameras.

01:57:15   I believe it's the R5C that can do it with the special lens

01:57:19   and the special, like, subscription software

01:57:21   that you need from Canon to stitch the images together.

01:57:24   Wait, what is it doing for you?

01:57:25   Panoramas, you mean?

01:57:26   No, I mean 3D video.

01:57:28   Sorry, I switched gears.

01:57:29   From a single camera with a single sensor

01:57:31   or from two cameras?

01:57:32   Yes, single camera, single sensor.

01:57:34   They just released a lens that has two eyes on it,

01:57:37   and I guess it--

01:57:38   And they split up the sensor?

01:57:39   Like, one gets the right half, one gets the left half?

01:57:41   Yeah, but somehow they're doing 8K video per eye

01:57:44   at 60 frames a second.

01:57:45   But I haven't had a chance to look too much into that yet.

01:57:47   But anyway, so if I was gonna actually start creating,

01:57:50   like, 3D content, I would probably look at that.

01:57:54   But what I learned from this is, like,

01:57:57   I love this camera for lots of other reasons,

01:57:59   but it is not, it's not what I need

01:58:02   if I wanna create 3D content.

01:58:04   That's kind of a separate beast.

01:58:06   Yeah, I do wonder if you could get away

01:58:08   with just way worse cameras but two of them.

01:58:11   Like, the 3D effect will hide a lot of sins

01:58:13   because you're just so wowed that it looks 3D

01:58:16   as long as you don't move your head too much, right?

01:58:19   But, yeah, 'cause, like, I keep thinking

01:58:21   of those camera stands in the studio.

01:58:24   They didn't look that big.

01:58:25   There's no way they looked close enough

01:58:27   to even house a single one of the 8K cameras

01:58:30   that Apple's recording the Major League Soccer games with.

01:58:33   So they just, not that they look like webcams,

01:58:35   but they looked way smaller than you would think.

01:58:38   Like, there are two lenses and presumably two cameras

01:58:41   inside these little pillars that just kind of look like posts.

01:58:44   And I thought those looked pretty good,

01:58:46   but maybe I was just fooled by the fact that they were 3D

01:58:49   and I didn't notice how pixelated everything was.

01:58:51   - No, the 3D stuff, I think, was pretty good.

01:58:54   Both, you know, the immersive and just regular 3D.

01:58:56   I thought all of it was very, very well done.

01:58:59   And I want more of it.

01:59:00   Somebody, I was just listening to somebody talk about this.

01:59:02   I can't remember who it was, but--

01:59:03   - Me?

01:59:04   - No, no, I mean, I thought it was somebody else

01:59:07   that was just saying, or maybe it was dithering.

01:59:09   I don't remember, but, you know, the 3D,

01:59:11   the immersive stuff, I shouldn't say 3D,

01:59:12   the immersive stuff was so, so good,

01:59:16   or I say was, as though it's past tense.

01:59:18   I mean, it's still there.

01:59:19   It's just, I've experienced just one--

01:59:20   - It's just the same videos you've seen a million times.

01:59:22   - Exactly.

01:59:23   But the immersive stuff is so incredibly,

01:59:26   incredibly, incredibly cool.

01:59:28   And I just want more of it.

01:59:30   I want all of my stuff to be immersive.

01:59:33   And I know that's never gonna happen, but I want it.

01:59:35   And I hope Apple really does just hammer on the gas

01:59:40   in order to, you know, get more of this.

01:59:43   And we heard about the MLS stuff,

01:59:46   and we are hearing rumors about the slam dunk contest,

01:59:50   but I want all the immersive video.

01:59:53   I want all of it.

01:59:54   Please and thank you.

01:59:55   - Yeah, they need to commission, like,

01:59:56   someone to do, like, nature documentaries.

01:59:57   You know, kind of like they did when 4K first came out,

02:00:00   where you saw a lot of planet Earth type things,

02:00:02   like, oh, do you have a new 4K TV?

02:00:04   Watch this 4K account.

02:00:05   But the problem is they sold way more 4K televisions

02:00:07   much quicker than Apple is going to sell these headsets.

02:00:11   So it might be a lot.

02:00:13   And Apple has the ability to bootstrap this

02:00:15   because they do have the rights to some sports franchises,

02:00:19   so they're doing the Major League Soccer thing,

02:00:21   and they do have a studio

02:00:22   that makes television shows and movies.

02:00:24   And it's not going to be economical,

02:00:26   and they're gonna lose money on it.

02:00:27   But if you want to solve the chicken-egg problem

02:00:29   and you can't convince anyone else to make content

02:00:32   for your, you know, 200,000 of your closest friends

02:00:35   with their headsets, Apple, you just pay to make it yourself.

02:00:38   And the good thing is if you make evergreen content,

02:00:40   like a planet Earth-style nature thing,

02:00:42   that's not gonna age that badly.

02:00:43   In five years, when more people have these headsets,

02:00:45   they'll still want to watch that planet Earth thing.

02:00:47   Hopefully, planet Earth won't have changed that much by then.

02:00:50   -Hopefully. -It's evergreen.

02:00:51   And sports aren't evergreen, unfortunately.

02:00:53   But, hey, if you catch a particular dramatic game

02:00:55   or something, that might make some evergreen content.

02:00:57   And if not, you know, there's always more sports.

02:00:59   So Apple needs to practice recording it in a format

02:01:02   that looks good in their headsets

02:01:04   and just keep doing that going forward.

02:01:06   (beeping)