You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi,
first of all: Thank you C0untFloyd for your work. Great project!
I used roop unleashed for a month now and I'm very impressed by the result.
But there are a few 'issues' I can't figure out nor find the answer.
Info: I'm on Win10, unfortunatly my 750 Ti is not supported so everything runs on CPU. I use roop unleashed for images only. No video.
Questions:
How does batch process (on images) work exactly?
Here is what I do:
.load a faceset I created into the 'source box'
.drag&drop about 20 target images into the 'target box'
.hit 'START'
I expected that every target image will get processed and saved either into the output or temp folder but thats not the case. Only the first image is processed.
Is there something I overlooked?
ATM I use this 'workaround' to manage more than one image
.drag&drop the target images into the 'target box'
.click the 'Face swap frames' checkmark below the preview window (first image is processed)
.click the delete (x) beside the current image in the target list (so the next image will get processed)
.wait for the image to be processed
.click the delete (x)..........
.at the end -> search all images in the temp folder, rename and copy them to a specific folder.
While I'm writing this I realize (again) that this really can't be the way. :) I wouldn't call this batch.
Are there more options for auto-naming the final image in the settings tab? date/time is nice. I would love to see something like postprocessor, source original filename, target original filename
Are you using an older version perhaps? At least it sounds as if you're doing everything correct, except that it's called 'target files'. I tested this real quick:
As you can see I processed 3 images at once and have 3 resulting images. You don't need to do anything with the preview.
The progress can also be seen in the console output:
In my case 3/3
Yes, there are
{timestamp} -> date and time appended
{i} -> index in list (number)
{file} -> original filename without extension
{date} -> date only
{time} -> time only
I'm not interested in adding more and target filename doesn't even make sense, as you define the target filename with this template. Feel free to create a PR the main work is done in template_parser.py
Older version of python (I see you're screenshot says 3.11.4 mine is 3.10.) but roop unleashed is 4. I updated it yesterday from 3.90 (same issue with 3.90)
I tried chrome instead of my standard broser (Firefox). same results.
I'll try it on another pc (clean Win10 in a VM) when I get home.
The console show the correct number of processed images. They're just not saved in the output nor the temp folder
thanks
EDIT: Ok I made a fresh 4.0 install and it works now. I guess something was wrong with the 3.90 and the update yesterday didn't change that. Time to run a big batch now :)
Hope you don't mind another question in this thread:
For the final file format I choose jpg. Is there a way to change the compression settings (I guess roop unleashed does this via ffmpeg right?)
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hi,
first of all: Thank you C0untFloyd for your work. Great project!
I used roop unleashed for a month now and I'm very impressed by the result.
But there are a few 'issues' I can't figure out nor find the answer.
Info: I'm on Win10, unfortunatly my 750 Ti is not supported so everything runs on CPU. I use roop unleashed for images only. No video.
Questions:
Here is what I do:
.load a faceset I created into the 'source box'
.drag&drop about 20 target images into the 'target box'
.hit 'START'
I expected that every target image will get processed and saved either into the output or temp folder but thats not the case. Only the first image is processed.
Is there something I overlooked?
ATM I use this 'workaround' to manage more than one image
.drag&drop the target images into the 'target box'
.click the 'Face swap frames' checkmark below the preview window (first image is processed)
.click the delete (x) beside the current image in the target list (so the next image will get processed)
.wait for the image to be processed
.click the delete (x)..........
.at the end -> search all images in the temp folder, rename and copy them to a specific folder.
While I'm writing this I realize (again) that this really can't be the way. :) I wouldn't call this batch.
Beta Was this translation helpful? Give feedback.
All reactions