1078
Replies
Would it be possible for Indigo to still do 10x digital since the sensor size is quite large vs the previous gen, then upscales it back to 12MP? I do not know how hard this is to achieve, but Topaz upscaling does quite amazing for things like that nature.
By @nhan_8084
This is not about sensor size, but about resolution. With 10x digital zoom, we get only 0.12MP to work with... one cannot upscale that more than 2x (i.e., to 0.5MP) without hallucinating. Topaz upscaling uses hallucination for high upscale ratios, and that is OK when you are editing because you can always undo. But for a camera we do not (at least not yet) hallucinate pixels we are not confident should be there.
I am a new user to Indigo, having just upgraded from an old iPhone XR to the iPhone 17 Pro. I have started testing using the iPhone on a tripod, comparing the native phone app to Indigo. On a 14mm photo, I noticed a subtle white band in the sky. The shape of a rainbow, but not a rainbow, just white. Is there a way of sharing the file with Adobe that is not public?
By @Sweet Light
Thank you for reporting this - we are aware of some problems on 17-series devices with lens vignetting calibration and are working with Apple to resolve that. It is possible that the problem is fixed in iOS 26.1, which we are trying to verify. For sharing files privately I will look into that, as we currently do not have a mechanism for it. As you can imagine, we cannot publicly share teammembers' emails.
I am a new user to Indigo, having just upgraded from an old iPhone XR to the iPhone 17 Pro. I have started testing using the iPhone on a tripod, comparing the native phone app to Indigo. Super Resolution tests: at 200 mm Indigo gives better detail than the Apple app. However, at 48 mm, Indigo smears fine detail. The Apple app has much better detail at 48 mm (12 mp file). I will do more tests to see if these results are typical for the 200 mm and 48 mm settings.
By @Sweet Light
Thank you for doing such rigorous testing. Please note that for Super-Resolution to work, there is an expectation that the device is *not* perfectly stable (i.e., that it is not on a steady tripod). When there is a small amount of hand-shake, which is low frequency enough to not introduce motion blur in each individual frame, we "extract" the extra information that is essentially hidden in each raw image to do the 2x upscale without hallucination. If you put the camera on a tripod the ability of the algorithm to extract such information may get reduced.
Just bought a 16 Pro and have been testing the Indigo camera app. Here are my findings:
Pros:
Fantastic low-light performance; great landscape photos
Excellent dynamic range and detail
No other app currently delivers such impressive 2× and 10× (SR) shots
Clean design
Quick jpg development option in the gallery when shooting DNG only
Cons:
Touch-to-focus is broken; the camera always chooses the focus point on its own
No AE/AF lock
No magnified focus assist in manual mode
No portrait mode
The interface needs work and optimization:
No way to hide the histogram
Too many taps to access exposure compensation
No customizable/favorite buttons (the histogram area could be repurposed for this)
The “Camera” button in the top-right corner is pointless (repeats Photo/Night switch?)
Space left &right to the Photo/Night switches could be used for custom buttons
The app could really use a proper icon (Pi?)
Overall, it’s a great app to experiment with and has huge potential. But in its current state, there’s no way it can replace the stock camera app as the primary shooter.
By @multipraktik
Thank you for sharing your thoughtful pros and cons for Project Indigo. The team is very happy when we get detailed feedback on what our customers think of the app (even if we may disagree at times :)). To go point by point on the cons side:
- Tap-to-focus does have some issues, partially related to limitations of iOS 18.x. In iOS 26 Apple introduced some improvements to the API which we are adding to Indigo and will see how much they improve the experience.
- AE/AF lock is on our roadmap.
- If you go to Pro controls -> Focus, on the left side of the slider there is a magnification icon. Tap on that and you will get a inset in the top-right of the viewfinder showing you a zoomed-in area of the image for precise focus control. If you tap on the viewfinder, the inset will show the area of the image where you tapped.
- Portrait mode is a difficult feature to do well. We are working on it, and once we have something we are happy with (i.e., which does not just match what others have but offers additional benefits in some scenarios) we will release it.
- On the UI, which is experimental (just like the app), we hear you that there are many ways to improve. Specifically for your notes:
- You can hide the histogram. Tap on it and select an option without it.
- Exposure compensation is currently hidden in Pro Controls. It is on our roadmap to expose it in the viewfinder for Auto mode.
- UI customization is obviously a powerful tool, but it also heavily complicates QA and testing process. We are a small team, focusing primarily on innovation in the computational photography space, these kinds of items are for now lower priority. But, we do have our eye on it and will start working on it once we have completed some other features that are in development.
- Secondary mode switcher in top right is going away. That was an experiment that did not turn out so well. 😉
- We will be adding more modes around Photo and Night, so that space needs to be somehow reserved. We will play with this more.
- 'Pi' icon is because of the name (Project Indigo) - this is an experimental name for an experimental app. But I will pass your feedback along to the appropriate team.
@BorisTheBlade how do you think apple generates the 48mp proraw for the main camera? Because i think the reason they dont allow 48mp single frame dngs is because these quad bayer sensors cannot output 48mp of both color and luma resolution at the same time its either going to be heavily interpolated and is going to look very bad in terms of artifacts so they need stacking to get all the color data from other frames but what if there is a moving subject or many moving subjects? Do you think that super res applied on a full picture not a crop on a 12mp bayer sensor like 13 pro can match apple pro raw 48mp in terms of quality and sharpness? What is your team take on this especially Marc Levoy's?
By @powerful_Elixir5E29
Apple captures multiple frames using the full resolution quadbayer sensor and then they use their flavor of computational photography to try and extract a 48MP image out of it. Because they are using a quad bayer sensor, each frame does not actually contain 48MP-equivalent amount of detail, which is why you can sometimes see some artifacts in the ProRaw result. We are experimenting with using our Super-Resolution on a burst of 12MP bayer raw images, and once we are happy with how it operates (quality and speed) we will likely release it so users can try.
It is frustrating trying to do a systematic evaluation of Indigo because it does not record proper metadata for the Focal Length 35mm. I try to be systematic in evaluation photographic equipment. In the case of Indigo, this means organizing files by the Focal Length 35mm equivalent. For some reason Indigo incorrectly records this import metadata. The result is a frustrating amount of time required to cross reference files imported to a desktop computer, viewed in Lighroom, against the Indigo app's information for each file indicating what the digital zoom was. For example, I want to organize all of the images shot at 48 mm SR into a single folder. This way I can compare them to 48 mm images taken with the Apple camera app or a different camera app. Indigo incorrectly shows 24mm as the 35mm equivalent focal length for both 24 mm and 48 mm shots. This is true of 100 mm and 200 mm shots If you want a systematic and proper evaluation done of the Indigo app it would be very helpful that the metadata is recorded accurately. In analysing the iPhone 17 pro camera apps, I treat the iPhone as three different cameras. In addition I want to further differentiate between 24 mm vs 48 mm and 100 mm vs 200 mm photos. This is the basis for evaluating a camera app ability to record images using 14 mm, 24 mm, 48 mm, 100 mm and 200 mm settings.
By @Sweet Light
Thank you for sharing your thoughts. It is definitely clear how this may create some difficulties in a workflow that you describe... we'll look into the EXIF/DNG tags to see if there is perhaps a better way to note the camera characteristics in the image. This is however a bit of a philosophical question: how do you record focal length for a photo which is a combination of some optical properties of the lens AND digital zoom obtained by cropping the sensor? Typically focal lenght entails not just the FOV of the lens, but also such characteristics as minimum focus, depth-of-field, etc. When digitally cropping, only the FOV is changed while everything else remains the same. So we opted to record optical and digital factors separately. I read many articles that came out when iPhone 17 was announced (example LINK) criticizing this introduction of confusion between claiming a focal length and claiming "optical quality", which kind of goes at the heart of what we are talking about. Happy to hear any suggestions you may have on ways to handle this.
AdChoices
