Computational Photography: Project Indigo Launch
Adobe has launched Project Indigo (Computational Photography), a free experimental camera app for iPhone, created by former Google Pixel camera engineers Marc Levoy and Florian Kainz. Indigo aims to deliver a natural, SLR-like look, full manual controls, and advanced computational photography features, capturing up to 32 frames per shot for superior image quality.
Computational Photography: Former Pixel Engineers’ Vision
Marc Levoy and Florian Kainz, renowned for their work on Google Pixel’s camera, left Google in 2020 to join Adobe and form the “Nextcam” team. Their expertise in computational photography, especially in night mode innovation, is central to Project Indigo. Indigo sets itself apart from AI-heavy competitors by using multi-frame super-resolution to capture real detail, not hallucinated content. Notably, the app features a “Remove Reflections” tool to eliminate glass and window reflections. Currently, Indigo is available for iPhone 14 and newer, with an Android version promised.
Computational Photography: SLR-Like Image Quality
Project Indigo prioritizes a natural, SLR-like aesthetic over the heavily processed look of most mobile camera apps. The app applies only mild tone mapping, color boosting, and sharpening, following feedback from photographers seeking a more authentic style. Indigo outputs both SDR and HDR images using a hybrid JPEG format, as described by Adobe. Users can toggle between SDR and HDR in the app, and shared images adapt to the recipient’s display. RAW shooters benefit from DNG files with computational enhancements as rendering suggestions, preserving editing flexibility.
Computational Photography: 32-Frame Image Stacking
A highlight of Indigo is its 32-frame computational photography stack. The app captures and merges up to 32 underexposed frames per shot, delivering images with enhanced dynamic range, reduced noise, and preserved natural textures. This process takes slightly longer but results in superior quality. Indigo applies this stack to both JPEG and RAW/DNG files. Users can choose between Photo mode for instant shots and Night mode for low-light scenes, with manual control over the number of frames per burst for optimal noise reduction and detail.
Computational Photography: Future Developments
Project Indigo is currently exclusive to iPhone, but the team has confirmed an Android version is in development. With its innovative approach to computational photography, Indigo is set to reshape mobile photography for enthusiasts and professionals alike.
References:
