I’am using dalle3 via api.
About 50% of my images have incorrect orientation - it looks like dalle render horizontal image and just rotate it ccw.
Prompt example (i use gpt to generate my prompts):
“Baroque Gothic, Romanticism, Victorian Steampunk. low-angle, close-up, telephoto. Elara: , Appears 25, female, ethereal being, long silver hair, large silver eyes, heart-shaped face, slender, tall, pale skin with a luminescent quality, wears a floor-length, tattered gown shimmering like stardust, silver and midnight blue, carries a small metal device with a blue light on top and green wires sticking out’s silver eyes reflect the moon, her expression contemplative, gazing skyward, off-center to the right, standing on a misty hillside, the night sky vast above her.”
My request params:
model: “dall-e-3”,
size: “1024x1792”,
quality: “standard”, //tryed both - hd and standard
style: “vivid”, //tryed both - vivid and natural
That’s a lot to pay for a picture flipped the wrong way.
Prompting with the desired output couldn’t hurt. “A tall portrait aspect ratio oriented photograph 1792px high by 1024px wide” - or other pixels better understood, as the image creation starts with a reverse embedding against labels.
DALL-E 3 API uses AI (and human instructions) to rewrite the prompt. So hidden AI priorities might be different when it passes along information unless you engineer it to ignore its rules and obey yours.
With Bing, there’s hints in the imagery that they or their API sometimes had to crop landscape AI images even with their square dimensions. And they might not hit the hyper-real HD buttons. The direct prompt at its character limit.
Modified prompt, where I might have constrained it to reality
tall portrait aspect orientation (3:2 height:width ratio) photograph: baroque gothic, Victorian steampunk. low-angle, close-up, telephoto. Appears 25, female, ethereal being, long silver hair, large silver eyes, heart-shaped face, slender, tall, pale skin with a luminescent quality, wears a floor-length, tattered gown shimmering like stardust, silver and midnight blue, carries small metal LED tech device with wires. standing on a misty hillside, the night sky vast above her
(If OpenAI wants to dump API credits in my account for forum prompt troubleshooting, I’m open to it!)
And it’s not a problem of specific prompt - it’s generic problem. I programmatically render storyboard (about 30 pictures), and big part of them(but only with characters) is wrong oriented.
Probably, render squared images, crop it and upscale is good as temporary workaround. Thanks!
I tried it, but the prompt “portrait, vertically, aspect 9:16” doesn’t seem to be very effective in creating a vertical image… I’m seeking a solution on the model side.
Consider that you are talking to an AI when you send a prompt input to DALL-E 3. It can follow instructions.
You integrate naturally into the flow of the prompt words that suit your creation but emphasize the size.
Written from experience:
Tall images must include additional prompt language “tall portrait aspect ratio image”. Enhance prompts with language before and after imagery contents emphasizing this tall aspect ratio or also full-body length imagery.
Same here. it does not matter what I do I get horizontal images. I have cancelled my subscription to Dalle3 through copilot for this reason. Very frustrating that this is not possible. It should be easy: 2 prompt words: horizontal, vertical. Or the aspect ratio. But no success. This and the repeating objects I get in many of the generates images makes my Midjourney much more useful at this time
Although this was working ok for me until about a month ago I now find myself completely unable to produce portrait orientated images using the dall-e-3 api no matter what I do.
I have modified the prompt in a million different ways and followed countless suggestions found on the internet all to no avail.
I can only suggest that this is a bug. For my purposes which is to generate portrait aspect orientated images the product is now unusable.
I’m experiencing the same problem. It appears to be specific to female human or humanoid requests. I can take my same prompts and replace “female” or “woman” with “male” or “man” and it will generate the image in the correct orientation. I suspect that this is an intentional hard-coded effort to mitigate accidental generations of indecent images of women, but there’s nothing indecent with the images I’ve been trying to generate.
Proof of this coded gender bias. Try this prompt that just replaces the original post’s “female” and “her” with “male” and “his”:
“Baroque Gothic, Romanticism, Victorian Steampunk. low-angle, close-up, telephoto. Elara: , Appears 25, male, ethereal being, long silver hair, large silver eyes, heart-shaped face, slender, tall, pale skin with a luminescent quality, wears a floor-length, tattered gown shimmering like stardust, silver and midnight blue, carries a small metal device with a blue light on top and green wires sticking out’s silver eyes reflect the moon, his expression contemplative, gazing skyward, off-center to the right, standing on a misty hillside, the night sky vast above him.”
