Hacker News Reader

Comment by sharkjacobs | Hacker News Reader

Was it ever explained or understood why ChatGPT Images always has (had?) that yellow cast?

My pet theory is that OpenAI screwed up the image normalization calculation and was stuck with the mistake since that's something that can't be worked around.

At the least, it's not present in these new images.

There's still something off in the grading, and I suspect they worked around it

(although I get what you mean, not easily since you already trained)

I'm guessing when they get a clean slate we'll have Image 2 instead of 1.5. In LMArena it was immediately apparent it was an OpenAI model based on visuals.

wdym it cant be worked around when there exist literal yellow tint corrector models/tools haha

Yeah, though I can imagine a conversation like this:

SWE: "Seriously? import PIL \ read file \ == (c + 10%, m = m, y = y, k = k) \ save file done!"

Exec: "Yeah, and first blogger get's a hold of image #1 they generate, starts saying 'Hey! This thing's been color corrected w/o AI! lol lame'"

Or not, no idea. i've not understood the choice either, besides very intelligent AI-driven auto-touch up for lighting/color correction has been a thing for a while. It's just, for those I end up finding an answer for, maybe 25% of head scratcher decisions do end of having a reasonable, if non intuitive answer for. Here? haven't been able to figure one yet though, or find a reason/mention by someone who appears to have an inside line on it.

There's a possibility that any automatic correction could have false positives (since the yellow tint doesn't happen 100% of the time) which creates different problems where a image could have an even weirder hue.

Meta's codec avatars all have a green cast because they spent millions on the rig to capture whole bodies and even more on rolling it out to get loads of real data.

They forgot to calibrate the cameras, so everything had a green tint.

Meanwhile all the other teams had a billion macbeth charts lying around just in case.

Also, you'd be shocked at how few developers know anything at all about sRGB (or any other gamut/encoding), other than perhaps the name. Even people working in graphics, writing 3D game engines, working on colorist or graphics artist tools and libraries.

Not really, but there's a number of theories. The simplest one is that they "style tuned" the AI on human preference data, and this introduced a subtle bias for yellow.

And I say "subtle" - but because that model would always "regenerate" an image when editing, it would introduce more and more of this yellow tint with each tweak or edit. Which has a way of making a "subtle" bias anything but.

There was also the theory that is was because they scanned a bunch of actual real books and book paper has a slight yellow hue.

That seems unlikely, as we didn't see anything like that with Dall-E, unless the auto regressive nature of gpt-image somehow was more influenced by it.

My pet theory is that this is the "Mexico filter" from movies leaking through the training data.

I never heard anything concrete offered. At least it's relatively easy to work around with a tone mapping / LUTs.

I'm guessing that it was intentional all along, as no other models exhibit this behavior. It was so it could be instantly recognized as ChatGPT

There's definitely an analysis on the net somewhere, can't remember the details though.

maybe their version of synth-id? it at least helps me spot gpt images vs gemini's

Colloquially called the urine filter

lets not mince words, its called the "piss filter"

Not always, it started at a very specific point. Studio Ghibli craze + reinforcement learning on the likes.

The Studio Ghibli craze started with the initial release of images in ChatGPT, and the yellow filter has always existed even at that time. They did not make changes to the model as a result of RL (until pontentially today, with a new model)

That's not how it works the model doesn't just update in real time to likes and besides it was already yellow upon release

Reply on news.ycombinator.com