Please enable the feature for users to select the option to generate images using either the Image 1 model or Image 1.5 model.
#Sora - Please reinstate Image 1 into the Old Sora
teal cape
normal elk
Yes, please.
teal cape
It's not just the image quality that I'm concerned about, but mainly prompt generation in general as Image 1 apparently better understands command prompts pertaining to certain specifics. For example, back in October, when I once asked Sora to generate a picture of a drummer captured at a certain camera angle, it generated that image flawlessly which couldn't have fit my exact vision any better. Nowadays with Image 1.5, it's merely impossible to do the same thing using the exact same prompt unless you upload a reference image of your choice and remix it which I didn't have to do with my initial generation. Plus nowadays it'll make the drummer's legs angled off to the side unlike how they should be when you would imagine someone sitting behind an actual kit if that makes sense.
flat arch
Agreed. Please let us use the image-1 model on Sora as an option.
finite lynx
why would anyone wanna use image1 though
spice jetty
why would anyone wanna use image1 though
same as many still want to use dalle
spice jetty
👍
teal cape
why would anyone wanna use image1 though
Because of the visual aesthetics as well as how it responds better to certain prompts at least from my experience