Natural view perception for phone cameras
Wide-angle photos make distant objects smaller than you saw them. Sjona brings them back to the size you saw and keeps the whole wide view. From one photo, on the phone.
Sjón (Icelandic) — sight, vision. Sjón er sögu ríkari — seeing is believing.
Phone photoSjonaThe eye sees sharply across about 40–60°, the field of a ~50 mm “normal” lens. Phones shoot wider, at 13–24 mm, to fit the whole scene, and distant objects shrink.
Barrel distortion is a separate defect, and manufacturers already correct it. This is the geometry of a wide view: a full-frame camera with a wide lens shrinks the mountain exactly like a phone.
Photographers step back and zoom in, so near and far keep their proportions. Often there is no room to step back, and no camera fixes it afterwards.
Photography forums, blogs, Reddit and a newspaper describe the same thing, long before anyone gave the problem a name.
“How can something so massive… look so tiny and unimpressive in a photo?”
“Why do photographs of beautiful scenery never do it justice?”
“What ends up on my camera or phone isn’t the spectacular sight I witnessed with my two eyes.”
“I personally do not think pics do any ranges justice.”
“Some of my photos fail to show the true depth of the features.”
“Why does the moon look smaller on photos than it does in real life?”
Sjona applies the brain’s correction to the photo: distant objects at the size we saw, the wide view kept.
Two objects of the same angular size look different when depth cues place them at different distances: the far one looks at least 17% larger. The difference also shows in the primary visual cortex (V1).
Murray, Boyaci & Kersten, 2006. Nature Neuroscience 9, 429–434Participants compared a real scene with pictures of it. Images with distant objects enlarged by a perception-based rule matched the scene better than standard perspective projection.
Nagata, Osa, Ichikawa, Kinoshita & Miike, 2008. Japanese Psychological Research 50(3), 117–127People judge each part of a picture from where they look, the way the eye moves from object to object in a real scene.
Martin, Gutierrez, Masia, DiVerdi & Hertzmann, 2025. Scientific Reports 15:15414Each pair shows the original phone photo and the same photo processed by Sjona. The mountain photos on this page were taken by Andrey, our technical co-founder, on his trips.
Phone photo
Sjona
Phone photo
SjonaEach step so far works with light, colour or pixels. No camera today, phone or professional, restores how near and far relate in the frame.
Several exposures merged for light and shadow
Depth is estimated and used to blur the background
Multi-frame noise reduction in low light
AI moves, resizes and removes objects
Depth rebuilds how near and far relate, as the eye saw it
First mass-market launches: HDR (iPhone, 2010) · Portrait mode (iPhone 7 Plus, 2016) · Night Sight (Pixel 3, 2018) · Generative Edit (Galaxy S24, 2024).
Sjona already runs as a working alpha app: one photo in, the corrected photo out, on the phone, with no cloud.
Public launch on iOS and Android in 2027. Real usage on thousands of real scenes proves demand.
A version calibrated for flagship phone cameras, ready for phone makers.
Sjona as a native Camera or Gallery feature, under a licence, a development contract or a deeper partnership.
A camera feature reaches every owner of the phone by default. We are looking for a first phone-maker partner to bring Sjona into the native camera.
Contact the foundersSjona is raising a pre-seed round. We send investors a detailed pitch with the technology, the plan and the numbers.
Request the investor pitchICF-certified coach and HR professional with 19 years in corporations and consulting firms. Runs Morerange’s operations and its HR practice, which funds Sjona’s development. Based in Barcelona.
LinkedIn20+ years in corporate leadership and entrepreneurship: strategy director at a national telecom with an €80M portfolio, product lead for IT and telecom projects. Certified Python developer; builds the company’s websites and the software around its products. Regularly works as a 3D designer and engineer on TV and film productions. Leads product, strategy and partnerships. Based in Barcelona.
LinkedInMaster’s degree in Applied Mathematics, Irkutsk State University. 20+ years in IT services delivery, operations management and digital transformation. Develops Sjona’s C++ core and leads the core algorithm and optical calibration methodology. A photographer and climber: two of the before/after scenes on this page are his mountain photos.
LinkedIn