[halotrope]

I recently photographed a Tiger (The big cat, in the Zoo...) with the iPhone 13 Pro Camera. It looked like a comic drawing with exaggerated stripes and weird artifacts in the face. The whole pipeline from the photons hitting the sensor of the photographer to being emitted from your phones screen again is just layers and layers of optimization, compression and enhancement. Completely compromised. I think we are getting very close to a point where photography cannot be assumed to depict the real world by default.

It is unfortunate as the reality that the brain constructs seems to be based on this fake reality of everyone being unnaturally perfect and attractive. This seems to work even against better knowledge. No wonder we have a teenage mental health crisis. I would not be surprised if social media will get the same have the same curve of public opinion as smoking once had.

[formerly_proven]

You're looking at a 1/3.4" OF image sensor with 1µm pitch, and an f/2.8 lens - not only is the sensor tiny, but the pixels are extremely tiny*, and it's fairly diffraction limited as well. If you had a raw image from this sensor, you'd notice how it's very noisy even at base gain and brightly lit conditions. Base sensitivity images from sensors with much larger pixels tend to still be quite visibly noisy!

No surprise that the image you get in the end is mostly ML fabrication cued by some noisy inputs.

* For "pixel-scale reference", on a standard full-frame camera, 1µm pitch would result in some ~900 MP of resolution. The pixels are that tiny here.

[ducaale]

This reminds me of Huawei's moon mode https://in.mashable.com/tech/3114/are-the-moon-shots-from-th...

Smartphone Camera vs Reality https://youtu.be/MZ8giCWDcyE

[Jugurtha]

There'll be a point where you take a picture of the tiger and your phone will say "I see you're trying to take a picture of a tiger. You suck at photography. Here is a better picture from National Geographic instead".