Hello. Thank you for your work, for helping to explain it clearly with your videos. The workflow you added works correctly and is very useful. The size module is especially well integrated. I am not a specialist in image generation, so please help me. Maybe one of the comment readers can do it too. As I said, the workflow works great. However, I cannot preserve the original image when expanding the image. Not in terms of size, but in essence. For example, if the original image has a certain person, then after expanding it is a different character. Maybe something needs to be switched? Thanks in advance for the answer. And have a nice day to you as the channel owner and to everyone who will try to help with the answer.
Sometimes it seems to me that many ComfyUI bloggers deliberately create the most complex and incomprehensible workflows possible :) Isn’t it possible for such an elementary thing as outpaint to create a simple and understandable workflow, and not just that?
Hi, I have a question...why put Florence if it doesn't work well? It doesn't detect well sometimes. Can't you do it manually? Thanks Apart from the fact that Florence does not detect certain aspects well, it makes a prompt for the visible image and we need the prompt for the non-visible image, that is, the one that is going to be generated again. Thanks
Hi. Thanks for the video and the workflow. I don't understand why? But it creates a totally different new image for me, she is a young white girl and she appears black with a totally different face, costume and environment. It's not outpainting (extending the original image), it's creating a totally different new image.
Hello. Thank you for your work, for helping to explain it clearly with your videos. The workflow you added works correctly and is very useful. The size module is especially well integrated. I am not a specialist in image generation, so please help me. Maybe one of the comment readers can do it too. As I said, the workflow works great. However, I cannot preserve the original image when expanding the image. Not in terms of size, but in essence. For example, if the original image has a certain person, then after expanding it is a different character. Maybe something needs to be switched? Thanks in advance for the answer. And have a nice day to you as the channel owner and to everyone who will try to help with the answer.
Sometimes it seems to me that many ComfyUI bloggers deliberately create the most complex and incomprehensible workflows possible :) Isn’t it possible for such an elementary thing as outpaint to create a simple and understandable workflow, and not just that?
will be good to add mask fix or something like that, as the original image looses its quality when outpainting
Hi, I have a question...why put Florence if it doesn't work well? It doesn't detect well sometimes. Can't you do it manually? Thanks Apart from the fact that Florence does not detect certain aspects well, it makes a prompt for the visible image and we need the prompt for the non-visible image, that is, the one that is going to be generated again. Thanks
Hi. Thanks for the video and the workflow. I don't understand why? But it creates a totally different new image for me, she is a young white girl and she appears black with a totally different face, costume and environment. It's not outpainting (extending the original image), it's creating a totally different new image.
the workflow is not the same as in the video I downloaded
fixed it