Hey how are you! If you use the mdx-net inst HQ 3 model that should work. Then select instrumental or vocals only and you’re good to go. If you get stuck, send it to me at mgmt@braddollar.com and I’ll do it for you.
@@BradDollar Hi. Hope you're doing well. I'm very satisfied with the result after the first two tests, with the vocals and other with instrumental. However I'm wondering if there is an other option where I could separate vocals and several instruments (keyboard, guitar, drums, etc.) in a one shot or is just possible in several proccesses. Thank you in advance.
Hi there! Thank you so much for watching! You may be able to do it with MDX-net instrumental model (see my video here), but the surefire way is to go under the "choose mdx-net model" drop down and then select "download more models." Under mdx-net, look for these: MDX-Net Model: kuielab a vocals MDX-Net Model: kuielab a other MDX-Net Model: kuielab a bass MDX-Net Model: kuielab a drums MDX-Net Model: kuielab_ b vocals MDX-Net Model: kuielab b other MDX-Net Model: kuielab b bass MDX-Net Model: kuielab b drums You can run those individual models on a track to get the stems you're looking for. If you get stuck, send me the song to mgmt@braddollar.com and I'll help you out!
Hey thank you for watching! You might be able to do that in one shot, just by making an instrumental the way I show you here. If that doesn't work, try breaking it up. Remove the vocal first, and then try the MDX model called MDX-NET Crowd HQ 1 (you'll have to download it from their built in download center, real easy). That should help manage some of the audience sound. If that doesn't work, you could potentially try a gate or automating up and down. I've mixed a lot of Grateful Dead live recordings and one of the big things to manage is audience noise. We had a lot of luck automating any sources of audience noise up and down throughout. It's tedious but made a big difference. Hope this works!
I did this but the parts where the vocals are removed end up quieter than the rest of the song :( so the volume appears to fluctuate throughout. is that just my song choice
Hey what’s up! Yea there’s bound to be a few artifacts here and there. Especially if your song is super compressed. You can also look for the mdx-net model called Inst HQ 5, which is a newer version than what was available when I made this video. You can also try increasing the segment size to something higher like 2048 or above. It will make it run a lot slower, but you will get better results. I ran into the same issue you’re having with a country song this week that was HYPER compressed and I did these workarounds to get to it. It’ll never be perfect, but it will be damn close. Thank you for watching!
Hi! As long as you’re downloading directly from ultimatevocalremover.com there shouldn’t be any issues. It’s worth noting I’m on Mac, but there’s nothing suspicious about the UVR files or downloads. I bet there is a tool you can use to scan the file though. I hope you try it! This thing is super helpful.
Great question! You sure can. There's an MDX model in their download center called KuieLab. That you can run on your track to get vocals, drums, bass and guitars/keys. Totally free as well! Thank you so much for watching and being here! Hope this helps.
Were you able to do it with one click? My first tool was StemPlayer but since the crack does not work, I have tried to use one of the songs I have splitted with StemPlayer on Spleeter, AudioLab Spleeter and I have found some faults on them. The seperation quality is not like Stem Player. UVR is a great program but still can not find to seperate them with one click.
Is there a way to separate multiple vocal tracks? Like if two people are singing together, harmonizing or speaking, that would make it so you'd have the individual tracks for their specific vocals?
Hey there! There is actually! It's not super perfect, but it can definitely separate out lead vocals from backing vocals. There's a "Vocal Splitter" option in the main menu you can play with and it works with the UVR-MDX-NET model. Hope this helps! Let me know how it goes. Thank you so much for watching!
Dang that is a bummer! Really good to know though, thanks. I bet it will catch up soon. It seems like their team updates it frequently. Also, just in case, make sure you turn off "GPU Conversion." That also doesn't work with my Mac (M1). I hope this helps! Thank you for watching.
It might be able to! I’m constantly surprised by it. It doesn’t seem to care if it’s stereo or mono. Let me know if that works! Thank you a ton for watching.
Brad: do you have tutorial that just says the basics of configuring it for notebookml vocal conversation (mono track) .mp3 format file and uploading it and retrieving via download the mono track that has been converted to stereo (two track). thanks. manu
@@BradDollar Brad: do you have tutorial that just says the basics of configuring it for notebookml vocal conversation (mono track) .mp3 format file and uploading it and retrieving via download the mono track that has been converted to stereo (two track). thanks. manu
What up man! Great question! I'm pretty sure that will work. It won't subdivide the choir into sections, but it will give you a lead vocal and the choir as backing vocals. You'll have to play with it a bit and I might experiment with different models (they're all free), but it'll work. I just did a quickie on a Kirk Franklin song and it worked.
@@BradDollar thank u bro for ur time I have one last question 🙏🏻🙏🏻 Is it also possible to separate the voices of two singers singing at the same moment?
Hey thank you for watching! There's an MDX-net model set available that you can use to extract what you want. Under download center, look for the ones below. Then you run the program, it will spit out the files and you can bring them right into Logic: MDX-Net Model: kuielab a vocals MDX-Net Model: kuielab_a_other MDX-Net Model: kuielab a bass MDX-Net Model: kuielab a drums MDX-Net Model: kuielab_ b vocals MDX-Net Model: kuielab_b_other MDX-Net Model: kuielab_b_bass MDX-Net Model: kuielab_b_drums
Hey! Yea! Try UVR-MDX-NET-Karaoke 1 and 2. Once you get them from the download center, go to additional settings > vocal splitter options > click "enable vocal split mode" > pick the model from the drop down. Each one will give you varying levels of positive success / refinement. Also, the model I used in the video InstHQ-3 does a great all around job too. Let me know if that works and if you need any other help at all! Thank you for watching!
Hi there! I think your best bet might be something like Izotope RX. You can split files apart with UVR, but it's really hard to take individual elements out of the stems that have already been separated. There are a few free models they have in their download center that might work though (VR Architecture Single Model or MDX MDX23c). It's free so it might be worth a shot. Hope this helps! And thank you a ton for watching!
Ah dang! I’ll see if there another model on here. Give iZotope RX10 a shot. You should be able to do it with the free demo. I haven’t gotten rid of dog barks before but I have removed things like piano pedals squeaking and background talking. www.izotope.com/en/products/downloads.html
ok, Let me try with this also. I tried with audo ai web site. They are giving demo example also before and after for dog barking. but not worked with my uploaded audio.
Hey thanks for watching! HQ4 is probably solid! There wasn’t a HQ4 when I made the video. My guess is it’s just ever so slightly improved. Give it a shot! Worst case you just make some extra files and run HQ3. Let me know how it goes! Appreciate you being here!
How do we know you didn’t play us the actual studio stems since you have those available? I would rather you used something that you don’t have the stems for.
Hey how's going! Firstly, thanks for watching! The good thing about this is you can try it yourself and see if I'm lying or not. It's free and you can give it a shot right now. Secondly, there's a bunch of people in the comment section saying they had success, so I think it works? Third, I can't use something you know because it would violate TH-cam's copyright laws and this video would get removed. It's happened to me before and it's not worth the risk. Lastly, what do I have to gain to from lying to you and making this fake? It's a free video, on a free platform and I'm not getting paid by anyone. I'm genuinely excited that this software is out there and just want everyone who needs it to know about it. Again - you can test all of this for yourself to know. Thanks for being here. I know the world is full of scammers, but I'm just a dude trying to share some knowledge. Be well!
I’ll even take it a step further, I’ll make you a private video that I can share and then you’ll know for sure. Pick the song, paste the TH-cam link for it here and I’ll send it back.
*Want to remove reverb from your vocals? Check out this video I made on how to do it with UVR!* th-cam.com/video/9Bj_bCqc5WY/w-d-xo.html
This video is awesome! I just used Uvr for the first time using this video, and everything worked well’
Let's go! Thanks for giving it a shot and watching the video. Super glad it worked!
that's wild this thing is free. thanks for the vid
Right?! Thank you for being here and watching.
nice work man, good job.
Appreciate you! Thank you for being here and watching.
Great video! Straight forward. That song is decent too
Thnx
Hey thank you! I’m glad it was helpful!
Hi. Can you tell what the best model to separate only the vocal(s) from a song. Thank you for your information and for share your videos.
Hey how are you! If you use the mdx-net inst HQ 3 model that should work. Then select instrumental or vocals only and you’re good to go. If you get stuck, send it to me at mgmt@braddollar.com and I’ll do it for you.
@@BradDollar Thank you. I"ll give a try.
@@BradDollar Hi. Hope you're doing well. I'm very satisfied with the result after the first two tests, with the vocals and other with instrumental. However I'm wondering if there is an other option where I could separate vocals and several instruments (keyboard, guitar, drums, etc.) in a one shot or is just possible in several proccesses. Thank you in advance.
Great vid! Do you know if there is a way to get separate stems of guitars and keyboards?
Hi there! Thank you so much for watching! You may be able to do it with MDX-net instrumental model (see my video here), but the surefire way is to go under the "choose mdx-net model" drop down and then select "download more models." Under mdx-net, look for these:
MDX-Net Model: kuielab a vocals
MDX-Net Model: kuielab a other
MDX-Net Model: kuielab a bass
MDX-Net Model: kuielab a drums
MDX-Net Model: kuielab_ b vocals
MDX-Net Model: kuielab b other
MDX-Net Model: kuielab b bass
MDX-Net Model: kuielab b drums
You can run those individual models on a track to get the stems you're looking for. If you get stuck, send me the song to mgmt@braddollar.com and I'll help you out!
@@BradDollar Hi. Thansk a lot for the deteailed answer. I'm gonna try the models you suggested. Thanks a lot and have a nice day!
How to remove the singer's voice and also the audience's voice from a live show?
Hey thank you for watching! You might be able to do that in one shot, just by making an instrumental the way I show you here. If that doesn't work, try breaking it up. Remove the vocal first, and then try the MDX model called MDX-NET Crowd HQ 1 (you'll have to download it from their built in download center, real easy). That should help manage some of the audience sound.
If that doesn't work, you could potentially try a gate or automating up and down. I've mixed a lot of Grateful Dead live recordings and one of the big things to manage is audience noise. We had a lot of luck automating any sources of audience noise up and down throughout. It's tedious but made a big difference.
Hope this works!
@@BradDollarThanks for the help!
thx !! it works !!!
Hell yea! Let's go! Good work. Thanks for being here!
I did this but the parts where the vocals are removed end up quieter than the rest of the song :( so the volume appears to fluctuate throughout. is that just my song choice
Hey what’s up! Yea there’s bound to be a few artifacts here and there. Especially if your song is super compressed. You can also look for the mdx-net model called Inst HQ 5, which is a newer version than what was available when I made this video. You can also try increasing the segment size to something higher like 2048 or above. It will make it run a lot slower, but you will get better results. I ran into the same issue you’re having with a country song this week that was HYPER compressed and I did these workarounds to get to it. It’ll never be perfect, but it will be damn close.
Thank you for watching!
@@BradDollar cheers man, I'll try that. appreciate ya.
Appreciate you! There are other models coming out all the time so I’m sure it will only get better.
how do you know if this Ultimate Vocal Remover .exe does not have virus binded to it?
Hi! As long as you’re downloading directly from ultimatevocalremover.com there shouldn’t be any issues. It’s worth noting I’m on Mac, but there’s nothing suspicious about the UVR files or downloads. I bet there is a tool you can use to scan the file though.
I hope you try it! This thing is super helpful.
can you do stem separation with this?
Great question! You sure can. There's an MDX model in their download center called KuieLab. That you can run on your track to get vocals, drums, bass and guitars/keys. Totally free as well!
Thank you so much for watching and being here! Hope this helps.
@@BradDollar Thanks for the info, I'll try it. I have to do a sample solo from a transcription and wanted to use the original track
Oh yea this should definitely work on that.
Were you able to do it with one click?
My first tool was StemPlayer but since the crack does not work, I have tried to use one of the songs I have splitted with StemPlayer on Spleeter, AudioLab Spleeter and I have found some faults on them. The seperation quality is not like Stem Player.
UVR is a great program but still can not find to seperate them with one click.
@@tromick results were abit mixed so I ended up scratching it. But the mix was very busy and the sax and piano have similar transients
Is there a way to separate multiple vocal tracks? Like if two people are singing together, harmonizing or speaking, that would make it so you'd have the individual tracks for their specific vocals?
Hey there! There is actually! It's not super perfect, but it can definitely separate out lead vocals from backing vocals. There's a "Vocal Splitter" option in the main menu you can play with and it works with the UVR-MDX-NET model. Hope this helps! Let me know how it goes.
Thank you so much for watching!
I'm sad that it doesn't work for the MacBook M3 chip at least not on my M3 Max :/
Dang that is a bummer! Really good to know though, thanks. I bet it will catch up soon. It seems like their team updates it frequently. Also, just in case, make sure you turn off "GPU Conversion." That also doesn't work with my Mac (M1).
I hope this helps! Thank you for watching.
Can URV separate two vocals (referring to Google Notebook LM) from a MONO track to Two invidual track?
It might be able to! I’m constantly surprised by it. It doesn’t seem to care if it’s stereo or mono. Let me know if that works! Thank you a ton for watching.
Also shout out KCXU! I’m from Richmond originally. Great to have some Bay Area love on the channel. Thanks for being here!
Brad: do you have tutorial that just says the basics of configuring it for notebookml vocal conversation (mono track) .mp3 format file and uploading it and retrieving via download the mono track that has been converted to stereo (two track). thanks. manu
@@BradDollar
Brad: do you have tutorial that just says the basics of configuring it for notebookml vocal conversation (mono track) .mp3 format file and uploading it and retrieving via download the mono track that has been converted to stereo (two track). thanks. manu
Bro Can I remove the singer's voice mixed with the choir's voice?
What up man! Great question! I'm pretty sure that will work. It won't subdivide the choir into sections, but it will give you a lead vocal and the choir as backing vocals. You'll have to play with it a bit and I might experiment with different models (they're all free), but it'll work. I just did a quickie on a Kirk Franklin song and it worked.
@@BradDollar thank u bro for ur time I have one last question 🙏🏻🙏🏻 Is it also possible to separate the voices of two singers singing at the same moment?
@@HishamTawfikall questions are welcome! Something like a duet might be a little tricky but it would definitely be worth trying.
How to do stem separation and import to logic pro ?
Hey thank you for watching! There's an MDX-net model set available that you can use to extract what you want. Under download center, look for the ones below. Then you run the program, it will spit out the files and you can bring them right into Logic:
MDX-Net Model: kuielab a vocals
MDX-Net Model: kuielab_a_other
MDX-Net Model: kuielab a bass
MDX-Net Model: kuielab a drums
MDX-Net Model: kuielab_ b vocals
MDX-Net Model: kuielab_b_other
MDX-Net Model: kuielab_b_bass
MDX-Net Model: kuielab_b_drums
hi,
do you know which model can be used to extract individual vocals in a song with multiple vocalist doing harmony ?
Hey! Yea! Try UVR-MDX-NET-Karaoke 1 and 2. Once you get them from the download center, go to additional settings > vocal splitter options > click "enable vocal split mode" > pick the model from the drop down. Each one will give you varying levels of positive success / refinement. Also, the model I used in the video InstHQ-3 does a great all around job too.
Let me know if that works and if you need any other help at all! Thank you for watching!
How can I remove dog bark sound from vocal in UVR ?
Hi there! I think your best bet might be something like Izotope RX. You can split files apart with UVR, but it's really hard to take individual elements out of the stems that have already been separated. There are a few free models they have in their download center that might work though (VR Architecture Single Model or MDX MDX23c). It's free so it might be worth a shot.
Hope this helps! And thank you a ton for watching!
Thank you for reply fast, I tried both model now and yes it's not remove barking part from already separated vocal
Ah dang! I’ll see if there another model on here.
Give iZotope RX10 a shot. You should be able to do it with the free demo. I haven’t gotten rid of dog barks before but I have removed things like piano pedals squeaking and background talking.
www.izotope.com/en/products/downloads.html
ok, Let me try with this also.
I tried with audo ai web site. They are giving demo example also before and after for dog barking. but not worked with my uploaded audio.
Yea give RX10 a shot. That'll give you the most control anyways.
How about UVR-MDX-NET Inst HQ 4?
Hey thanks for watching! HQ4 is probably solid! There wasn’t a HQ4 when I made the video. My guess is it’s just ever so slightly improved. Give it a shot! Worst case you just make some extra files and run HQ3. Let me know how it goes!
Appreciate you being here!
How do we know you didn’t play us the actual studio stems since you have those available? I would rather you used something that you don’t have the stems for.
Hey how's going! Firstly, thanks for watching! The good thing about this is you can try it yourself and see if I'm lying or not. It's free and you can give it a shot right now.
Secondly, there's a bunch of people in the comment section saying they had success, so I think it works?
Third, I can't use something you know because it would violate TH-cam's copyright laws and this video would get removed. It's happened to me before and it's not worth the risk.
Lastly, what do I have to gain to from lying to you and making this fake? It's a free video, on a free platform and I'm not getting paid by anyone. I'm genuinely excited that this software is out there and just want everyone who needs it to know about it. Again - you can test all of this for yourself to know.
Thanks for being here. I know the world is full of scammers, but I'm just a dude trying to share some knowledge. Be well!
I’ll even take it a step further, I’ll make you a private video that I can share and then you’ll know for sure. Pick the song, paste the TH-cam link for it here and I’ll send it back.
you rule!