Posted by sajomes 3 days ago
"There is a crazy amount of image processing going on behind the scenes in each smartphone"
I bought the Pixel 10 because it said it had great zooming capabilities. After 5X (that means 10X), it's completely stupid. Basically it takes a very blurred photo and then tries to reconstruct it with AI. The final results are completely awful and really distant from reality.
Any advice in how to improve that?
Edited to add: I have a pixel 10 pro which has a better zoom lens, so I could be having a different experience than you...
Makes for a fine image if "looking good" is the objective (which it often is), but I wouldn't trust it to disambiguate anything uncertain.
So the ENHANCE is no longer just some lazy movie writing!
The biggest of the "sharpening with AI" models, if you let 'em go, they'll turn you into a celebrity on accident.
If you go to the Topaz AI forums on Facebook, you'll see examples of where the AI sharpening tool went off the rails and turned a blurry photo of someone (typically the user who bought the software, who is trying to clean up their own photos) into whichever celebrity that is in the AI model who resembles them.
IE, it might turn a photo of YOU into a photo of Robert Downey Jr.
Jory Miller?
Also wonder how how long until photos are no longer court-admissible evidence because they get overwhelmed with the burden of the amount (and the cost) of the "non-AI" certification?
The only solution I see is some kind of hardware-linked signature showing that the image was produced at a certain time, by a certain device. In theory, you could make this signature somewhat “provable” and anchored to a specific time by combining something like 1) the latest Bitcoin block hash, and 2) a blockchain anchor, thereby proving that the photo was taken within a certain time widow after the hash was known and before the anchor was published. But even that arrangement is highly susceptible to hardware attacks.
Because of the limitations of capturing multiple lenses at once through isVirtualDeviceConstituentPhotoDeliveryEnabled, there is a second capture mode in the app for 'Quality' which takes a simultaneous capture then takes sequential shots from each lens which do benefit from Deep Fusion, then uses those for the stitch (as long as the Deep Fusion images outperform the simultaneous capture results). That's also what I do for our Night mode.
The result when done outdoors (like, standing in a dark forest at midnight) will be a photo where you can actually see what's around you, and details show up as if it was a brightly full moon lit night.
The math is surely more complex than aligning same-lens photography even when you consider the change in angle and perspective between two shots from the same lens.
https://www.google.com/search?client=firefox-b-d&q=ARGUS+pho...
The reborn Nokia/HMD was also doing multi-sensor image stacking back in 2019 on the Nokia 9 Pureview.
Google has some great blogs about their camera technology https://research.google/blog/night-sight-seeing-in-the-dark-...
From the authors of Halide.
Edit: actually I can't seem to find the setting that aligns them. I seem to remember it being able to do that but maybe it doesn't... which would be disappointing...
I agree, and I built it!
The app launched in 2019. Ideally I would have said ML, but it felt like AI had more recognition among non-technical users.
If I launched it in 2026, it would 100% say "ML" not only because it's more technically accurate, but to make it abundantly clear it doesn't generate slop.
Since at least the iPhone 12, holding the phone still and letting scene sit for a few seconds will result in a sharper photo. I’m not sure if it’s a subtle version focus stacking or a sharpening algorithm, but it’s definitely there.
On my 16e, immediately looking at a photo just taken can result in the image noticeably sharpening after a second or 2, so this may be a post processing step as well.
Subtle movement between shots can be used to create an image sharper than any of the inputs.
I’m guessing it has to do with how intensely focused / scared my brain was with the little guy
https://apps.apple.com/us/app/twincam/id6448645623 https://play.google.com/store/apps/details?id=org.evergreenl...
Except for the "have full network access" permission according to the Google Play Store.
The fact that they have to resort to manipulation to sell their product tells me that they are not trustworthy and that their product is probably underwhelming.
Not to mention the developer account subscription is $99/year
If people are working on a full-time job and trying to create something in the free (!!) time (it still needs to be valued for the time and effort!) and they could chose to or not to charge for the app. It's more personal and a general view.
IMHO, any and all subscription fees are justified based on the thoughts of the developer and how he values his effort, time and development complexity.
Not stating anything about it even if it is not appealing or convincing. It is purely a personal choice and the dynamics of value, usage, benefits, cost and the problem it solves from the buyer's side will also kick into the thought process and decision making.
I'm just saying only one side of the story and want to record my thoughts. Peace out!
If you had customers buying last year, it’s likely you’ll have even more customers buying next year. If you stop maintaining your app, that stream of new sales almost immediately dries up.
Offline mobile apps which charge a subscription are a very silly fad to me. The few that get away with it are being steadily edged out by someone else who wants to eat their lunch.
I love the OP’s idea and absolutely see the use case. I also think it’s very promptable. My unsolicited advice is to strike while the iron’s hot at $0.99, because I’d be very skeptical if they’ll attract enough subscribers before getting starved out by copycats.
In general, when I buy something I want to pay for it once and then have it. I don't want to rent or lease it, even if it's consumable. When I eat a restaurant I I pay for the meal when it's over; if it was delicious I might come back again. Even if I like it so much that I eat there every week, that doesn't mean I want a weekly subscription.
An engineer with an app can come to work 365 days a year, fix bugs, add nice new features, unbreak the app every time iOS updates, and, though you use the app every day, they can't count on you to give them another $5 every year - or ever again - unless you enter into a subscription.
This isn't to say you have some obligation to do a subscription to any app you don't think is worth it, but for apps from non-scam developers, they're usually like $20 or less a year - about a nickel a day. A nickel is 3¢ in year-2000 dollars btw.
It's on the developer to prove to you that the app is worth paying a nickel a day to have! But ruling it out just because we wish it didn't have to be, well, that just rewards the developers who monetize exclusively with ads or sell their 'free' apps to sketchy operators.
I'd like it if Uber was free or just cost a buck a ride flat rate, but I know that anyone willing to give me a ride on terms like that is probably looking to rob me.
After a year, I don't think anyone is bothered if an app becomes unsupported for whatever reason. Usually what happens is the opposite - developer unnecessarily cripples an existing, perfectly working app, to try to milk some more revenue out of it.
This is an especially moot point now that most platform API upgrades are $1 of LLM work away.
I was just listing one more thing a developer who wasn't slacking might spend his time doing, that's all.
> Keeping the app compatible with new OS releases is necessary to keep sales going.
Relying on constant sales to new users sounds like a pyramid scheme. Software licenses don't wear out like shoes or become consumed like butter, with happy customers returning to buy more. Eventually, most of the people that need most apps have bought it and sales don't "keep going" more than a trickle, except for apps that are such mainstream hits that your new users are just "everyone who became old enough this month to get their first phone." And because the App Store has ruined organic discovery, paid apps rely almost exclusively on paying for installs either via Apple's ads or elsewhere. So, that eats a good part of each new sale's revenue right away.
Check the iOS top charts right now. For paid apps, the top-selling app is Shadowrocket, a rule-based proxy utility. How long until all developers who want a rule-based proxy utility have already paid the $3? If 100,000 developers buy that app, then that developer nets $210,000 pre-income-tax. About on par with a good corporate SWE job outside of FAANG. But that will slow down. There just aren't that many people who are looking for that. #2 is just a client of a subscription service for businesses. #3 is AnkiMobile Flashcards at $25. See, at that price I agree it can afford to never charge again. #4 is an ephemeral app for some art event in Portland so it doesn't need to be sustained. #5 is Procreate Pocket. I guess there are enough new aspiring artists appearing every day who can afford these apps (I think it's $13 for iPad).
> After a year, I don't think anyone is bothered if an app becomes unsupported for whatever reason.
Okay, so subscribe to the apps you use, and then immediately go in and turn off renewal in Settings -> (your Apple ID) -> Subscriptions. There. You now have a one-time purchase of an app which will stop working in a year - just like apps that have no recurring revenue stream do.
> This is an especially moot point now that most platform API upgrades are $1 of LLM work away.
I don't think you are a developer of serious iOS apps. Just the visual changes, that users expect apps to go along with, require major redesigns. People don't buy apps that look like they were designed in 2019.
It's not the amount of the payment. It's the subscription model I can't stand. You're correct that I might use an app I like every day. But the marginal cost to the developer of me using it every day is $0. Rather than belabor the restaurant example , consider a book or a music CD. Writing books or recording music is generally a slow and difficult process, it takes months to produce one that might be read in a weekend. But if you buy a book, you can read or refer to it every day without paying again, or you cna listen to your favorite songs as often as you like by playing the CD. Nowadays many people prefer to stream music, but I think when people buy a book (even an e-book), they consider that it's their copy to read as often as they want...
...or not. If I buy a book and I don't like it, or later I find another book that I turn out to prefer to the first one, I can just not read it again. I don't have to cancel a subscription for the book.
It aligns your incentives with the developer very well. They really want you to not cancel your subscription after only paying the first year's payment (when you may not have even paid back the customer acquisition cost!), and you want the app to be high quality and well-maintained. If the developer doesn't meet your bar for that, you cancel the subscription with 4 taps and they lose future income. That incentive to please customers after the sale just doesn't exist in a pay-once app world (except in tiny niches of old-school developers, like the Panics of the world, whom I do greatly respect).
Besides the incentives, it also gives you a chance to, after that one-year $7 spend, re-evaluate "do I even use this anymore?" If you paid for a guitar tuning app and you stopped playing guitar, you can just cancel the subscription, even before you've paid the amount that they'd have priced it for one-time-purchase. There is upside for the consumer in the apps that either let you down, or that you simply realized you no longer need, and it lowers the risk bar to try an app that you aren't sure about.
I get that if you do end up using such an app for the next 15 years, yes you'll probably pay somewhat more than any price you'd have considered paying one-off, but it also probably means that app (an app you use often for decades) couldn't have been sustained for all that time with only new customer installs paying the bills.
> if you buy a book, you can read or refer to it every day without paying again, or you cna listen to your favorite songs as often as you like by playing the CD.
We agree that this is all very great. And since the music or book you buy carries zero support burden of any kind from the artist it's a pretty great model - after their album launches the artist can devote 100% of their bandwidth to producing new music, or they can tour to get recurring income out of the same music, or a mix of the two.
If you paid for it once and only got to keep the version of it as of when you paid for it, in a couple years the platform would remove your ability to use it.
Don't want new features, and don't need an upgrade then stick with it. If you need the new version, then pay a partial amount of the app for the upgrade?
So, in the old days, you paid often $50-500 for a piece of software and the developer would usually have zero burden to update the software for you. By the time the OS completely broke it, it would have been 5+ years later. All that time, the developer was freed up to build and sell whole other applications, or, to innovate new and compelling features to get you to pay the upgrade price to get a new version.
Today that developer would have to do 5 major updates with breaking API changes, and probably at least 1 full visual refresh just to keep up with Apple's changing design whims, just for that one app to not be bricked. They have less bandwidth then to build more apps, since each app carries its own annual burden of churn.
Giving this advice to developers today is frankly absurd career advice.
1. Few apps can achieve the scale needed for a one-time fee to be sustaining for the app - most apps appeal to a finite set of people, and eventually most who are in your target market (and who are willing to spend actual upfront money), have bought the app.
2. An entire generation has grown up that has never paid more than maybe 99 cents for software, and most refuse to even pay that - they're used to adware. You can't price an "app" at $75 and expect it to be bought by essentially anyone. While adware can actually be a very financially viable route, many developers hate ads and want nothing to do with this model.
3. Unlike the multiple years of stability of old desktop operating systems, platform owners today think nothing of breaking even a plurality of apps with any major OS version - and for some insane reason they also think a new "major version" ought to come out annually.
If you want apps that cease getting updates after a few months and age very poorly, or just to have no choices out there besides adware, keep boycotting every developer who uses a subscription, regardless of how cheap it is.
If you want apps that are diligently maintained and which keep working forever, reward developers who produce apps with reasonable subscriptions. I have a password manager that's $19 a year and covers my whole family, and a podcast player that's $15. That password manager comes out to a nickel a day.
I pay exactly zero a month in subscriptions.
I wouldn't have enough time in the day (or month) to use 30 apps, or even 10, once you subtract all the first-party ones and ones that are just interfaces for things that are monetized separately (grocery store app, restaurant app, Uber, etc).
I have 3 paid subscription apps.[1] Total cost $43 a year. My only point was that if there's an app that is useful why shouldn't you pay a small amount per year that sustains it, guaranteeing that it gets good attention from a developer who is accountable not to their advertisers, but their users.
[1] Note: That doesn't count streaming services, but my family paid for cable 30 years ago and that was fine. It's entertainment.
Subscriptions for software that don’t offer an external service put all the incentives backwards strictly to the benefit of the developer and expense of the user.
There’s no question why it’s a great deal for the developer. But nothing you’ve said explains why any user should buy into this at all.
Your post prompted me to donate to KeepassXC, my subscriptionsless password manager, because I would be sad if that went away, and I would hate if the software landscape became even more a subscription hell than it already is.
Pedants corner:
>most apps appeal to a finite set of people
The set of people is finite.
If the Apple store supports this, free trial + one time payment sounds like a good way. Don't optimize for freeloaders, as someone else said.
If the app relies only on iPhone hardware and/or is not a live service, an account/subscription should not be required. Even non-devs might justifiably ask "why do I need a subscription for a camera app".
https://www.lux.camera/halide-mark-iii/
As for the subscription, there’s always been a pay-once option for major releases. You just miss out on a free trial, the beta program, and anything we might release with ongoing costs.
When I emailed in, they blamed it on how the App Store mechanism changed.
Halide somehow is the only app that has this issue.
The issue is that, indeed, it actually validates that the App Store account currently logged in purchased the app. I believe this is intended behavior, as it closes a piracy loophole where you could log into the Apple ID of a friend, download all their paid apps, and then log in to a different account while retaining access.
If other apps don't run into this problem, it's possible they haven't moved on to StoreKit 2. Or the app has their own login independent of app subscription; setting aside the time it would take me to build that, I don't think most people are willing to create an account to try a camera app.
I use Analogue for day to day capture, and NoFusion and Moment apps for things I will want to edit later. All with one-time purchase. I don't mind paying for quality apps, but I do have subscription fatigue now.
Might check it out. Thanks for letting me know. But really been enjoying Analogue lately.
Unfortunately it looks like lifetime users don't get access to the beta program for whatever reason. I guess they don't want feedback from those users.
I tried Analogue too, but I stopped using third-party camera apps altogether because they all seem to suffer from an issue where if you lock the phone too quickly after taking a photo, the photo is lost forever. I lost so many photos to this that I knew I took.
Halide shouldn't be afflicted by the lost photo issue you're talking about. Normally, if you close a lock-screen extension before a photo is written, it won't be in your library. This can be a problem when the user quickly smashes the capture button and your photo queue gets backed up.
Luckily, the unlocked app gets a chance later to access the content of lock-screen capture sessions. So our strategy in Halide is to write the intermediate files to disk as soon as you initiate a capture; this should take a split second. When we have uninterruptible time, later, we make sure the photo finished writing to your library, and if not, we try again.
However, Apple's documentation isn't clear if the OS will "wake-up" your main app to process unfinished photos, or if it relies on the user launching the app. I'll just say that I dog-food Halide every day by capturing photos of my toddler, and I can't remember the last time I lost a shot.
These days, I use the regular Camera app because I need my damn photos to save. I loved Halide, but that bug made it too unreliable. I get very distressed by data loss and couldn't continue to risk it while trying to develop a way to work around it.
As I said in more detail below, on a free tier I expect to be able to at least try it out in a normal way. a monthly limit of exporting 3 pictures (eg to look at on my computer or TV) seems like something a normal person would hit in the first 30 minutes of just exploring what it does. Now, I noticed you run it this way for 14 days, but I gotta say that on first view this was massively outweighed by the 3/month limit.
subscription-wise, I just don't like them. If you are doing a ton of server-side processing, I'd rather pay on an as-needed basis. If not, I'd rather just Pay $19.99 (or whatever price you determine to be competitive in the market for photo apps). I see why subscription income would be nice to have for you, but for me it's a financial leaky faucet. In that case, I'd prefer to just go without because the app is a fun thing to have rather than a professional need.
It also bothered me that it was described as a 'reasonably priced' subscription. If I don't see a $ sign and a number, I reflexively classify it as a scam even though rationally I know you're just testing the waters. I close or dismiss 10 or 20 things a day after a cursory glance into whether they're worth my time or money: fear of being ripped off (FOBRO) outweighs the fear of missing out (FOMO).
I said: note that you are not obligated to maximize profit and you can give away more features for free and that it generally makes the world a kinder place to do so.
The point being that they have a choice. Contrary to what people are saying they are supposed to do, charge as much as possible, they can still choose to do something else. This is of course the normal mindset everywhere else, but HN seems to forget about it.
Just differentiate signal from noise. Listen to your target audience not random internet comments
I think some dollars is better than 0 dollars.
Don’t overestimate the value of your products, it’s not an essential app. I’ll wait for someone else to vibecode it and release it for free.
But if there's a free tier I want to try out normal use patterns over the course of a week or a month, and I prefer one-time payments or token purchases to subscriptions. I do the latter only for my IDEs, and still don't like it that much.
My objection was to the free tier only allowing one to export 3 images a month. That seems bizarre to me, since in course of trying something for the first time you might make 5, 10, or 20 images in an hour. If I'm running into a monthly limit in the first few minutes of using a tool, my immediate reaction is that it's something I can live without.
I am also considering a lifetime unlock option.
After I buy it I am zero burden to them beyond occasional updates, which I have very modest expectations of. They are far more likely to get my money if it’s one time. I can’t imagine paying a subscription for a physical camera, let alone a camera app on hardware I already paid a lot for
Plenty of apps do this - looking at the first browser tab I have open I see gmail which does exactly this.
For a pure client app like an iphone app, you don't have any real "per user" costs, so there are no concerns about going under from the costs of attempting this. It would be very different if you had to pay for server capacity.
“Easy-to-understand framing The brackets over the viewfinder indicate the area covered by the more detailed lens. Frame your subject inside them to capture the best detail.“
Reminds me of the old 3d cameras that took two photos at once through 2 lenses (on film)
https://www.ebay.com/itm/206409326865?chn=ps&_trkparms=ispr%...
I didn't know about those Nishika stereo cameras, that looks incredible!
When you hit the little 0.5x, 1x, 4x buttons in the camera app you are switching between the physical cameras. Though Apple goes out of their way to obscure what is actually happening.
The obvious issue with the product here is that the telephoto lens will not cover the whole image that the 1x captured so only some portion of the center of the frame will be higher resolution.
One of my favorite games ever, and I'd gladly pay for an ad-free version, but they don't have it. I assume there's much more money to be made with ads, over subscriptions.
Gameboy emulators are now available again in the iOS app store or even as a a web app. Or just play on real hardware.
I think a few bucks for ad-free brings in more money for most developers, but that still pales in comparison to a recurring subscription model (or models where people keep spending in-app on consumables).
Once governments force Apple to allow sideloading we should get some kind of f-droid for ios which only lists open source user respecting apps.
https://developer.apple.com/documentation/avfoundation/avcap...
https://www.theguardian.com/technology/2014/jan/06/apple-buy...
Recently, Samsung introduced the ISOCELL platform which allows to select the area which will be recorded at higher resolution compared to other regions.
There can be all sorts of different applications build around such as extracting higher resolution of text regions, etc [2].
Of course, the lens needs to support this resolution but I think that will be default in the mid-term future.