Rechazo imágenes

[Recorridos virtuales 360]

Buenas tardes, tengo muchas fotos de AI que realmente no sé porque Adobe me las echa para atrás. Os adjunto una prueba en este mensaje. ¿Una causa podría ser que no las he catalogado como ilustración? ¿Hay que  catalogar todas las imágenes generadas por AI como ilustración? ¿Si las mejoro, puedo volver a subirlas como ilustración? Gracias 

[Abambo]

No, these should be categorized as photos. What is the reason of the refusal?

[Abambo]

It's too small. It's a PNG instead of a JPEG. And I strongly suspect that there are technical issues.

[Recorridos virtuales 360]

Te paso el fichero pequeño para no pasar el grande, pero son jpg con las medidas que pide Adobe, más de 4000 px

[Abambo]

Yes, but when you change the pictures and submit here, we can't judge what is wrong as we see something different from what the moderators saw. So it's pretty useless to us.

Quality issues is not a refusal because you posted in the wrong category.

[Recorridos virtuales 360]

[Jill_C]

The Gen AI tool is attempting to simulate a shallow depth of field here, but didn't quite get it right. For instance, the woman on the left is quite blurry, but the wall just behind her head is sharply focused. 

[Recorridos virtuales 360]

Problemas de calidad

[Jill_C]

You are allowed to resubmit images after re-editing them. I think this image has a white balance issue - too yellow.

[daniellei4510]

As well as the other issues mentioned, the woman in the middle has some severe quality issues in her eyes that would need fixing. The screen shot I've attached should answer your main question.

 

[Recorridos virtuales 360]

Ahora lo veo y ya estoy arreglando algunas fotos rechazadas. Gracias os pasaré algunos ejemplos arreglados como este. Esta bien la imagen?

[Abambo]

You really need to critically check the details:

Your picture may or may not pass with the defects shown, but still, I would address those issues.

[daniellei4510]

You probably need some additional touchup on the middle's woman's face (like those hairs on either side of her forehead), but if you were also to throw the wall behind the woman in front out of focus (as @Jill_C noted), that will also help. You might also consider adding some catchlights in her eyes to match those of the middle woman.

Just how far you actually want to take it is up to you, but I would also use generative fill to fix the tangled hair near the shoulder of the woman in the foreground and remove what could either be hair or a necklace appearing across her neck.

Also, the middle woman again...there is no strap holding up  the left side of her dress. More gen fill might fix that. Myself, I'd probably spend upwards to 2 hours editing this image. That's more time than I would like to spend, but you have to decide for yourself if it's worth taking the time to fix it or to chuck it. I always approach AI as follows "That's a really great result...were it not for all the defects."

[Abambo]

Yes, there is some major cutout issue with the front woman.

@daniellei4510,

Concerning the strap, who's left are you talking about:

As for addressing the issues and the effort to put into the asset: I can easily lose 2 hours on a real portrait… 😉

[daniellei4510 · Accepted solution]

Yes. You pointed to the correct left. 🙂

As for two hours on a real portrait, I hear you. After being without a darkroom for years, I finally took it upon myself to scan and edit my entire body of usable portraits and figureative work. It took me well over a year and a half. I actually became a fixture at my local bar, where I'd spend hours doing the menial work on my laptop or iPad just dust spotting, then putting the finishing touches on my REAL computer at home. Editing straight out of the camera is intesive enough, but correcting less than perfect scans probably helped prepare me for editing AI.