@@cognibuild I already use facefusion since long time. Mostly doing shit job. Even the roop_unleashed gave better results with zluda for me. Facefusion also VERY problematic (random freezes, gradio crashes, it's drop the 'server' what created by itself etc)
If anyone got stuck at the Conda part, you need to use the miniconda terminal to install conda create -n Rope python=3.10.13 , & conda activate Rope, & conda install conda-forge::cuda-runtime=11.8.0 cudnn=8.9.2.26 conda-forge::gputil=1.4.0, then cd.. back to your temp\Rope-Live\ folder to complete the rest
@@cognibuild no normal powershell doesnt work, I had to use the miniconda terminal, additionally the gui is now stuck and cant be moved or closed unless I kill the terminal. I think it has something to do with the fact IM using Nvidia GPU, and some files I have installed via the requirements.txt should not have been ie tensorrt-cu11 tensorrt_lean-cu11 tensorrt_dispatch-cu11, & torch==2.0.1+cu118 torchvision==0.15.2 torchaudio==2.0.2, how do I remove them to clean it up?
this version is laggy. i dont know why, but my vga ram quickly get exhausted when i use this version, compared to the previous one. still id like to say that your explanation is fun n easy to follow. though i found some trouble on installing cudnn cos i have no folder c:\programfiles vidia. after some google i move it to folder nvidia gpu computing. i wonder iff its the cause of the laggyness
that very well could be.. if its not getting the proper cudnn folder that would certainly be an issue. tbh, that folder should have been made already if you have your cuda installed. If you come over to Discord we can talk more and you can show me what you've been dealing with and I'll give it a look later: discord.gg/JfWJXXUu
@@cognibuild i had my cuda 11.8 installed properly, also have check and test it the way you teach me. have also reinstall n restart 3 times. but still no folder of c:\prog vidia. so google tells me to just put all the cudnn folders to c:\prog vidiagpucomputing etc.
hey man, i got everything working until the obs part. I can't get rope to send to obs or I don't know how to set up the obs part properly since you didn't cover in the video. Do i add a new video capture source to the scene? also I have 3 cams and rope live only ever detects one even if I put 3 in the settings.
@@tetragrammaton3 so what you do is you open up OBS and click on start virtual camera. Then you click on stop virtual camera. Then you close obs. Then open rope live and it should be on the top left. To send to the virtual camera itself there is a radio button on the top right that says send to virtual camera. Click that and then open up whatever app uses your virtual camera and you're gold
@@cognibuild please i need a guide so I don't waste my money:- can I buy HP Victus 15 AMD Ryzen 5 5600H 8GB RAM 512GB SSD RX 6500M Gaming Laptop Windows 11 Home 15-inch FHD 144Hz IPS 250 nits display 8GB RAM and 512GB SSD storage AMD Ryzen 5 5600H processor Discrete: AMD Radeon RX 6500M 4GB Full-size, backlit, blue keyboard for deepface live mode stream???? will it work perfectly? please help guys
@@cognibuildplease I have been battling with this for days. I do everything right as you explain here but when I want to start virtual camera, it gives me an error message: starting the output failed. Please check the log for details.
@@cognibuild thank you 🙏 I figured it out, I realized I didn’t need to keep my Obs studio open while using the third party app, I didn’t know I just had to start virtual cam and stop then close OBS, I did that and it worked fine for the third party app without errors. I’ve joined discord community regardless, thank you 🙏
I’ve tried this, but for some reason no mater if I turn on the restorer my face swap changes but it still be blurred, all the models are in the models folder, any idea what I’m doing wrong?
@@buike9306 you need to open up obs and click on start virtual camera. Then click on stop virtual camera. Then close obs and go back to rope live. Click on start rope and a little screen at the top left will pop up amongst the other videos. Click on that and it will bring the webcam up to the main window. Then click play
@@buike9306 I've been working on a software called bark and of course there's voice craft... Actual deepfake voice cloning is very difficult to get going. But both of those are available on my patreon site. The most recent thing I got working was xtts which does a pretty good job of creating voices. Maybe the easiest I've used
@@Ganeshkumar-ch1kg they're all good. My favorite is rope l live because I like the manual install rope next is awesome as well but it's twice the size because of the portable install. If you want the best webcam go with rope next just because that's what he keeps improving. You can get them all at at my patreon site. Join us a farmhand and you can download all versions
FRIEND!!! Thank you so much for such detailed instructions!
glad it helped
Working great and on the new update.👍😁
i love you omg
haha.. thnx
Hi, cool video man, which graphics card are you using?
@@ancapistaomarxista4071 I'm running an Nvidia 4090. But this will work with a lot of other graphics cards
@@cognibuild Yeah bro... EXCEPT the AMD and the Intel cards.
@@Hazardteam face fusion should work for you and it's awesome as well. I have an installation tutorial for that as well
@@cognibuild I already use facefusion since long time. Mostly doing shit job. Even the roop_unleashed gave better results with zluda for me. Facefusion also VERY problematic (random freezes, gradio crashes, it's drop the 'server' what created by itself etc)
@@Hazardteam I definitely had problems with pumping the quality too high and having it crash my computer. Try the original rope by Hilobar
Can this work on intel graphics
If anyone got stuck at the Conda part, you need to use the miniconda terminal to install conda create -n Rope python=3.10.13 , & conda activate Rope, & conda install conda-forge::cuda-runtime=11.8.0 cudnn=8.9.2.26 conda-forge::gputil=1.4.0, then cd.. back to your temp\Rope-Live\ folder to complete the rest
miniconda terminal? You should be able to do everything in the command line.
@@cognibuild no normal powershell doesnt work, I had to use the miniconda terminal, additionally the gui is now stuck and cant be moved or closed unless I kill the terminal. I think it has something to do with the fact IM using Nvidia GPU, and some files I have installed via the requirements.txt should not have been ie tensorrt-cu11
tensorrt_lean-cu11
tensorrt_dispatch-cu11, & torch==2.0.1+cu118
torchvision==0.15.2
torchaudio==2.0.2, how do I remove them to clean it up?
I see you are using nvidia GPU, so now I guess Ive hit a wall - bugger
heres my errors
(base) C:\temp\Rope-Live>rope.bat
(base) C:\temp\Rope-Live>call conda activate Rope && python Rope.py
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\user\miniconda3\envs\Rope\lib\tkinter\__init__.py", line 1921, in __call__
return self.func(*args)
File "C:\temp\Rope-Live
ope\GUIElements.py", line 401, in
self.button.configure(command=lambda: self.function(self.args))
File "C:\temp\Rope-Live
ope\GUI.py", line 2675, in parameter_io
self.widget[key].set(value, request_frame=False)
KeyError: 'source videos'
Unrecognized file type: C:/Users/user/Videos\desktop.ini
Unrecognized file type: C:/Users/user/Videos\Captures\desktop.ini
Unrecognized file type: C:/Users/user/Videos\desktop.ini
Unrecognized file type: C:/Users/user/Videos\Captures\desktop.ini
Unrecognized file type: C:/Users/user/Videos\desktop.ini
Unrecognized file type: C:/Users/user/Videos\Captures\desktop.ini
invalid command name "2188678259008update"
while executing
"2188678259008update"
("after" script)
invalid command name "2188678255872check_dpi_scaling"
while executing
"2188678255872check_dpi_scaling"
("after" script)
invalid command name "2188678447744coordinator"
while executing
"2188678447744coordinator"
("after" script)
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\user\miniconda3\envs\Rope\lib\tkinter\__init__.py", line 1921, in __call__
return self.func(*args)
File "C:\temp\Rope-Live
ope\GUIElements.py", line 331, in window_resize
self.frame_length = self.parent.winfo_width()
File "C:\Users\user\miniconda3\envs\Rope\lib\tkinter\__init__.py", line 1327, in winfo_width
self.tk.call('winfo', 'width', self._w))
_tkinter.TclError: can't invoke "winfo" command: application has been destroyed
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\user\miniconda3\envs\Rope\lib\tkinter\__init__.py", line 1921, in __call__
return self.func(*args)
File "C:\temp\Rope-Live
ope\GUIElements.py", line 557, in
self.textselect_label[mode].bind("", lambda event, mode=mode: self.select_ui_text_selection(mode))
File "C:\temp\Rope-Live
ope\GUIElements.py", line 575, in select_ui_text_selection
self.function(self.argument, self.name)
File "C:\temp\Rope-Live
ope\GUI.py", line 2400, in set_view
self.populate_target_videos()
File "C:\temp\Rope-Live
ope\GUI.py", line 1993, in populate_target_videos
self.static_widget['input_videos_scrollbar'].resize_scrollbar(None)
File "C:\temp\Rope-Live
ope\GUIElements.py", line 105, in resize_scrollbar
self.child.configure(scrollregion=self.child.bbox("all"))
File "C:\Users\user\miniconda3\envs\Rope\lib\tkinter\__init__.py", line 2760, in bbox
self.tk.call((self._w, 'bbox') + args)) or None
_tkinter.TclError: invalid command name ".!frame2.!frame.!canvas"
@@SupermanRLSF ok .. you shouldnt be using powershell. Just use a normal elevated command prompt
this version is laggy. i dont know why, but my vga ram quickly get exhausted when i use this version, compared to the previous one. still id like to say that your explanation is fun n easy to follow. though i found some trouble on installing cudnn cos i have no folder c:\programfiles
vidia. after some google i move it to folder nvidia gpu computing. i wonder iff its the cause of the laggyness
that very well could be.. if its not getting the proper cudnn folder that would certainly be an issue. tbh, that folder should have been made already if you have your cuda installed. If you come over to Discord we can talk more and you can show me what you've been dealing with and I'll give it a look later: discord.gg/JfWJXXUu
@@cognibuild i had my cuda 11.8 installed properly, also have check and test it the way you teach me. have also reinstall n restart 3 times. but still no folder of c:\prog
vidia. so google tells me to just put all the cudnn folders to c:\prog
vidiagpucomputing etc.
@@IkeikeIkeike77 ive dealt with that before as well but dont know what causes it.. yes, just create the folders
@@cognibuild yup. seems fine for me. thx bro for all your sharp clear yet still fun tutor
The link isn’t working to your two step installer
hey man, i got everything working until the obs part. I can't get rope to send to obs or I don't know how to set up the obs part properly since you didn't cover in the video. Do i add a new video capture source to the scene? also I have 3 cams and rope live only ever detects one even if I put 3 in the settings.
@@tetragrammaton3 so what you do is you open up OBS and click on start virtual camera. Then you click on stop virtual camera. Then you close obs. Then open rope live and it should be on the top left. To send to the virtual camera itself there is a radio button on the top right that says send to virtual camera. Click that and then open up whatever app uses your virtual camera and you're gold
@@cognibuild please i need a guide so I don't waste my money:- can I buy HP Victus 15 AMD Ryzen 5 5600H 8GB RAM 512GB SSD RX 6500M Gaming Laptop Windows 11 Home
15-inch FHD 144Hz IPS 250 nits display
8GB RAM and 512GB SSD storage
AMD Ryzen 5 5600H processor
Discrete: AMD Radeon RX 6500M 4GB
Full-size, backlit, blue keyboard for deepface live mode stream???? will it work perfectly? please help guys
@@cognibuildplease I have been battling with this for days. I do everything right as you explain here but when I want to start virtual camera, it gives me an error message: starting the output failed. Please check the log for details.
@@Ekemini-xl1ow come over to the discord channel and share a screenshot:
discord.com/invite/6xetaBBx
@@cognibuild thank you 🙏 I figured it out, I realized I didn’t need to keep my Obs studio open while using the third party app, I didn’t know I just had to start virtual cam and stop then close OBS, I did that and it worked fine for the third party app without errors. I’ve joined discord community regardless, thank you 🙏
I’ve tried this, but for some reason no mater if I turn on the restorer my face swap changes but it still be blurred, all the models are in the models folder, any idea what I’m doing wrong?
@@A_Little_O_Everything it has to do with the quality of your input photo. Try a higher quality one
@@cognibuild I’ve tried a high quality photo for my input, I don’t know it just com out looking like a fuzzy mask
@@A_Little_O_Everything at the top there's a setting for 1:26 to 252 and 512. Try 512. Also slide ffmpeg all the way to the left
I'm still confused about how to set up the webcam.
@@buike9306 you need to open up obs and click on start virtual camera. Then click on stop virtual camera. Then close obs and go back to rope live. Click on start rope and a little screen at the top left will pop up amongst the other videos. Click on that and it will bring the webcam up to the main window. Then click play
@@cognibuild thank you
Do you know of any good voice cloning software that I can do a good job
@@buike9306 I've been working on a software called bark and of course there's voice craft... Actual deepfake voice cloning is very difficult to get going. But both of those are available on my patreon site. The most recent thing I got working was xtts which does a pretty good job of creating voices. Maybe the easiest I've used
Apologies for asking again, but have you thought about including a real-time voice cloning feature
Jim? Is that you?
Johhny Boy! It is
welcome to my world lol
What your GPU
4090 24gb
Deeplive cam of this one which one is good?
@@Ganeshkumar-ch1kg they're all good. My favorite is rope l live because I like the manual install rope next is awesome as well but it's twice the size because of the portable install. If you want the best webcam go with rope next just because that's what he keeps improving. You can get them all at at my patreon site. Join us a farmhand and you can download all versions
Donal Trump!