Copy link to clipboard
Copied
This thread is now closed. Please update to LrC 14.x or LrD 8.x. If you wish to provide feedback, please go to the new article.
The recommended order for applying edits is:
Copy link to clipboard
Copied
Hola, solo queria saber si esto consume creditos firefly porque parece que cada vez que lo uso en lightroom esta consumiendo. es correcto?
Copy link to clipboard
Copied
Merci pour la description de la procédure à suivre, le résultat est là et plutôt convaincant !
Copy link to clipboard
Copied
Great feature. I would love feathering, as well as the object rectangle selection for this tool. After a few times the Refresh option is missing.
I don't see any information in the documention/responses about what de-selecting the checkboxes "Generative AI" and "Object Aware" do. Is it correct that you must also uncheck the latter to remove that person/object/etc.? Maybe some people do not see that. Being aware of selecting everything etc., I had an issue with removing an intruding elbow from the border on a non-cropped image.
Selecting a person's glasses with "Object Aware" off, I get the person without her glasses which is the purpose. However, the person eyes looks at the camera now, from an extreme angle and the eyes are weird (too big, ugly, angrilly) every time. There should be at least some variations with the eyes looking as the original.
Copy link to clipboard
Copied
@ThierryH: "I don't see any information in the documention/responses about what de-selecting the checkboxes "Generative AI" and "Object Aware" do. Is it correct that you must also uncheck the latter to remove that person/object/etc.?"
No, the two options are independent. The Generative AI option says to use the new Adobe Firefly service to generate a replacement for the selection. The Object Aware option controls how to make the initial selection -- when unchecked, it's exactly what you've brushed, but when checked, it uses the same algorithm as Object Mask to identify an "object" that becomes the selection to be replaced.
Copy link to clipboard
Copied
[This post contains formatting and embedded images that don't appear in email. View the post in your Web browser.]
@ThierryH: "I had an issue with removing an intruding elbow from the border on a non-cropped image."
Generative Remove did a very good job of quickly removing the elbow from the JPEG you uploaded:
That indicates you were working with a raw to which Lens Corrections or Transform had been applied, or in less frequent cases, the Camera Raw engine had automatically cropped a border of 8 - 12 pixels around the edge. In all cases, Remove is (confusingly) looking at those hidden pixels. Temporarily disable Lens Corrections and Transform and then do the remove. See here for more details:
https://www.lightroomqueen.com/generative-remove-replace/
If that doesn't help, upload the original raw to Dropbox, Google Drive, or similar and post the sharing link here.
Copy link to clipboard
Copied
[This post contains formatting and embedded images that don't appear in email. View the post in your Web browser.]
@ThierryH: "Selecting a person's glasses with "Object Aware" off, I get the person without her glasses which is the purpose. However, the person eyes looks at the camera now, from an extreme angle and the eyes are weird (too big, ugly, angrilly) every time."
With your low-resolution JPEG, I got better results using Object Aware. I looked at over 50 variations, and here are some of the better ones (hard to know if they'd be acceptable on the full-resolution raw):
The Refresh button never disappeared for me, but maybe there's a difference when using raws.
Copy link to clipboard
Copied
The disappearing of the Refresh was probably due to me creating snapshots to make the point. There is definitelly a difference between Raw/CR3 and JPG in the results. Even in my JPG examples, the person looks at the camera 99% of the time. Looking at the position of the head, it is not a very natural nor elegant result. I would be happy with 50%.
Copy link to clipboard
Copied
@ThierryH, perhaps you could attach a full-resolution JPEG version of the photo here. With the low-res screenshot, Remove was generating a large fraction of replacements with the eyes looking forward.
Copy link to clipboard
Copied
Copy link to clipboard
Copied
@ThierryH, using Object Aware to select the glasses and then lightening the face, here are the best 10 out of the first 33 variations (the first one was the first-generated variation) -- to my taste, they look acceptable, especially when looking at the entire photo:
Copy link to clipboard
Copied
@ThierryH, The best one to my taste is this one -- I like how the man and woman are looking away from each other:
Copy link to clipboard
Copied
I was trying to remove my shadow of a photo I took. It replaced it with a different looking shadow nine times.
Working on the remaining shadow, it took three more runs, each time selecting the "least" of the remaining shadow to finally remove the shadow. In other words, three on the initial shadow and three more for each remaining shadow.
Copy link to clipboard
Copied
@gary_sc, please attach a full-resolution JPEG exported from the original unmodified photo. Also, are you editing a raw or a JPEG original? If a raw, which exact camera model?
Copy link to clipboard
Copied
Copy link to clipboard
Copied
Well, that's interesting, it failed. That was send via email since when I clicked on the "Reply" button it shifted to email. Now that I see I'm here, let me try again. If you want/need the raw, let me know.
Copy link to clipboard
Copied
[This post contains formatting and embedded images that don't appear in email. View the post in your Web browser.]
@gary_sc, Remove with options Generative AI and Object Aware immediately removed the shadow on the first try:
This usually indicates that there are pixels along the edge that have been cropped out. Did you apply Crop, Lens Correction, or Transform, all of which can crop out pixels? If so, temporarily disable all three tools and redo the Remove.
If that doesn't help, please upload the raw to Dropbox, Google Drive, or similar and include the sharing link here. (The forum platform doesn't accept raws!)
Copy link to clipboard
Copied
Filling empty space is something on our backlog. Thanks for the feedback.
Barry
Copy link to clipboard
Copied
Hate it! A simple spot removal that used to take a second now takes several minutes to process. Totally unacceptable.
Please put it back the way it was.
Copy link to clipboard
Copied
@silversnaps I would not be using Generative Remove for a simple spot.
None of the existing tools were removed. I recommend using the previous Heal/Clone/Remove tools for simple spots and leaving Generative AI off. They are very different tools for very different purposes.
Here is a nice guide: https://helpx.adobe.com/lightroom-cc/using/generative-remove-faq.html
Copy link to clipboard
Copied
Agree. 2 + minutes to remove a minor small glare on a window with a blue background is unaccepatable. In Photoshop it takes less than a few seconds.
Copy link to clipboard
Copied
@MackeyRJ, I don't see significant time differences between PS and Lr on Generative Remove on the same image. Can you post a video showing equivalent edits on similar images, demonstrating the difference in time?
Copy link to clipboard
Copied
Same here. Progress bar takes about as long as in photoshop, but then after reaching 100% it takes a few minutes to display the result.
Copy link to clipboard
Copied
It would be helpful to mention which Lightroom Application you use when seeing this delay. Lightroom Classic? Lightroom Desktop, Mobile, Web?
Copy link to clipboard
Copied
YesterdayI saw this behavior in Lightroom classic. Progress bar as quick as in photoshop, but it takes a few minutes to display the result.
Surprisingly today the complete processs was as fast as in Photoshop.
Copy link to clipboard
Copied
Running LR Classic (13.3.1)on a Mac Mini M2 Pro running Ventura 13.6.7
In general, remove object works very well. However, fairly often a random image is inserted in the image. The original object to be removed is correctly done and the background is created correctly but then an image is inserted in that masked space. Often it is a bird or birds. In the image I'm working on, I wanted to remove an object on the far left left of the image. That was some light equipment against the backdrop of a stage. What I should have had was pretty much a dark fill. Instead I got a half of a drinks cup inserted. Half as in vertical.