> https://www.dxomark.com/wp-content/uploads/2026/09/DuoTable_...
Oh god that plasticization of the faces is absolutely horrifying. Like a bad image gen AI from 2022. Or reptilians wearing cheap rubber masks.
Edit: what's with the downvotes?
Oh god that plasticization of the faces is absolutely horrifying. Like a bad image gen AI from 2022. Or reptilians wearing cheap rubber masks.
Edit: what's with the downvotes?
Some people really like that style, though. I think it's a matter of preference, and it seems like the people who like that style are exactly who Apple is optimizing their image algorithms for.
This particular iteration seems to increase the hallucinations a bit more than previous models, though. Unless one of the models did actually have one eye with eyelashes and one without for the iPhone shoot.
This is just because the eye without eyelashes is on the side of her face that's in shadow. That means there's less contrast between the black of the eyelash and her skin tone (which is darker in the shadow area), and so the denoising algorithm wipes out the lash. All of this image processing happens way too fast on an iPhone for anything deserving the label of 'AI' to be involved. People are just going nuts over the unsurprising artefacts of a completely standard image processing pipeline.
The only surprising part is how good the results are: there is no leftover color noise that I can see, the window doesn’t have any color fringing, the high-frequency detail of the table setting and the drapes hasn’t been destroyed, the small backlit hairs on the sides of the face of the girl in front haven’t been smeared over oil-painting style. It’s strange how one eyebrow (in the back) got completely obliterated while the other looks completely fine, and it’s even stranger how the texture of the collar (in front) is basically perfect on the left where it’s unobstructed but basically destroyed on the right where some hair escaping from the braids was in front of it, so perhaps there is some image-gen-adjacent stuff going on here.
The skin looks bad, don’t get me wrong. But the degree of badness seems fair for the situation, and as a whole the result has no business looking as good as it does to my 2016-camera-trained eye.
(Also, don’t photograph people with a wide-angle lens from up close, it’s always freaky at best—like here—and ugly at worst. Just don’t.)
As I mentioned in another comment, even this isn't strange when you consider that the areas where detail is lost are in shadow (compared to the similar areas where more detail is preserved). That means less contrast, which means more detail lost to denoising.
Presumably because others disagree. I also do not see it here. Zooming in there are clear pores and skin texture. Is it somewhat smoothed? Yeah, it has that feel. Does it look like a cheap rubber mask? No. And that phrasing is frankly very insulting to the pictured subjects.
It's smartphone cameras, squeezing blood from a stone is pretty hard.
This is a heavy walk-back from reptile wearing a rubber mask. It should not be hard to understand why that statement can be seen as insulting.
> You can clearly see that the arbitrary part of the face of the girl in front is literally blurred
Not really. It’s a bit smoothed but the smoothest parts of her face are also where the skin is stretched tightest from smiling. It is actually smoother in those areas. I don’t see a ton of difference in the foreground between the skin texture in the different photos.
Background is different but as you said, that’s blood from a stone. The data just isn’t there.
It's very hard to understand. It doesn't make sense that a criticism of the camera would be an insult to the subjects of the photo.
The criticism is that the image isn't an accurate representation of the subjects. If B is bad, and B != A, it doesn't follow that A is bad.
Your reaction is?
a. “Yeah, the slight over processing definitely turned my wife into a lizard person.”
b. “Fuck you.”