A new vocal removal method for unreleased tracks (lossless rips)
Moderators: General Help, Kanzenshuu Staff
- Attitudefan
- I Live Here
- Posts: 2963
- Joined: Tue Aug 03, 2010 9:51 pm
- Location: Canada
A new vocal removal method for unreleased tracks (lossless rips)
Hello everyone! Not that I know how to do the new method to rip vocals from the soundtrack, but some people have figured it out.
Below are some examples where reverb was nearly removed completely from the vocal track, but the music track remains intact:
DBZ closeing theme (USA - Jeice/Burter/Goku variant): https://youtu.be/0TTEm8GSc-0
Ginyu Powers up in Goku's body & Gohan & Krillin vs Ginyu: https://youtu.be/NduXlbMJvZg
Goku vs Ginyu: https://youtu.be/NBYoXdpH3YU
Somehow they were able to rip these with nearly complete lossless audio. Apparently there is a technique to this new way to rip the music without (almost 100% but not quite) any vocals (and reverb) bleeding into the soundtrack.
Below are some examples where reverb was nearly removed completely from the vocal track, but the music track remains intact:
DBZ closeing theme (USA - Jeice/Burter/Goku variant): https://youtu.be/0TTEm8GSc-0
Ginyu Powers up in Goku's body & Gohan & Krillin vs Ginyu: https://youtu.be/NduXlbMJvZg
Goku vs Ginyu: https://youtu.be/NBYoXdpH3YU
Somehow they were able to rip these with nearly complete lossless audio. Apparently there is a technique to this new way to rip the music without (almost 100% but not quite) any vocals (and reverb) bleeding into the soundtrack.
My favourite art style (and animation) outside Toriyama who worked on Dragon Ball: Katsuyoshi Nakatsuru, Masaki Satō, Minoru Maeda, Takeo Ide, Hisashi Eguchi, Katsumi Aoshima, Tomekichi Takeuchi, Masahiro Shimanuki, Kazuya Hisada
- Naruto6583
- OMG CRAZY REGEN
- Posts: 946
- Joined: Sun Oct 13, 2013 3:08 pm
- Location: King, NC
Re: A new vocal removal method for unreleased tracks (lossless rips)
They've also been able to do it with kai's boo arc ost as well
PSN SonGoku325
DeviantArt http://narut6228.deviantart.com/
DeviantArt http://narut6228.deviantart.com/
Re: A new vocal removal method for unreleased tracks (lossless rips)
My rough guess would be that the way they did it was by inverting the audio that had kikuchi ost with the one that had falcouner ost so that the result would leave only music thus cancelling the sfx and voices. Though I'm amazed with the lack of sound artifacts, they did a really good job.
- Naruto6583
- OMG CRAZY REGEN
- Posts: 946
- Joined: Sun Oct 13, 2013 3:08 pm
- Location: King, NC
Re: A new vocal removal method for unreleased tracks (lossless rips)
How would they have done it with kai though?yugidmx5 wrote:My rough guess would be that the way they did it was by inverting the audio that had kikuchi ost with the one that had falcouner ost so that the result would leave only music thus cancelling the sfx and voices. Though I'm amazed with the lack of sound artifacts, they did a really good job.
These are a few good examples
Friends Or Rivals: https://www.youtube.com/watch?v=7akDFitsknU
Large Gap: https://www.youtube.com/watch?v=kCwhNnVqmUA
No More: https://www.youtube.com/watch?v=AzF9cRWxstM
Impatience: https://www.youtube.com/watch?v=AlttmFQMHgE
Misplaced Determination: https://www.youtube.com/watch?v=QwVS-MN82Ak
Unnatural Power: https://www.youtube.com/watch?v=rz3ea0QpWqc
And he's also gonna be releasing clear rips of the future trunks arc next year
PSN SonGoku325
DeviantArt http://narut6228.deviantart.com/
DeviantArt http://narut6228.deviantart.com/
Re: A new vocal removal method for unreleased tracks (lossless rips)
TBH, I am not quite sure for these ones since the rips are without any sound artifact remains from voices and sfx as far I could hear it. I am now really interested in the method myself.
- Naruto6583
- OMG CRAZY REGEN
- Posts: 946
- Joined: Sun Oct 13, 2013 3:08 pm
- Location: King, NC
Re: A new vocal removal method for unreleased tracks (lossless rips)
So am I. I'm curious if its audacity or something else.yugidmx5 wrote:TBH, I am not quite sure for these ones since the rips are without any sound artifact remains from voices and sfx as far I could hear it. I am now really interested in the method myself.
PSN SonGoku325
DeviantArt http://narut6228.deviantart.com/
DeviantArt http://narut6228.deviantart.com/
-
- Beyond Newbie
- Posts: 136
- Joined: Mon Jun 12, 2017 6:06 pm
Re: A new vocal removal method for unreleased tracks (lossless rips)
Cool. Shame we'll probably never know the method so it'll be exclusive to relying on these guys, and shame we probably won't get another official ost release for 100+ years..if at all
Re: A new vocal removal method for unreleased tracks (lossless rips)
I asked him about the vocal isolation he did on Kai and it seems the method is just removing or inverting the center channel from the 5.1 audio.
-
- Beyond Newbie
- Posts: 136
- Joined: Mon Jun 12, 2017 6:06 pm
Re: A new vocal removal method for unreleased tracks (lossless rips)
So then in theory, this should be doable with the DBS episodes(with the required blu ray/DVD)? Meaning that a rip of all the songs from the Goku Black Arc, with them all being clean, is possible?yugidmx5 wrote:I asked him about the vocal isolation he did on Kai and it seems the method is just removing or inverting the center channel from the 5.1 audio.
- Naruto6583
- OMG CRAZY REGEN
- Posts: 946
- Joined: Sun Oct 13, 2013 3:08 pm
- Location: King, NC
Re: A new vocal removal method for unreleased tracks (lossless rips)
It should be. He's already planning to do it next yearRichardKing2 wrote:So then in theory, this should be doable with the DBS episodes(with the required blu ray/DVD)? Meaning that a rip of all the songs from the Goku Black Arc, with them all being clean, is possible?yugidmx5 wrote:I asked him about the vocal isolation he did on Kai and it seems the method is just removing or inverting the center channel from the 5.1 audio.
PSN SonGoku325
DeviantArt http://narut6228.deviantart.com/
DeviantArt http://narut6228.deviantart.com/
Re: A new vocal removal method for unreleased tracks (lossless rips)
We use the blu-ray audio to rip from. I'll just copy and past this.... It's how I learned how to do it.
The .flac file for the episode is the US 2.0 stereo track (has dub voices + sound effects + Faulconer music) while the .thd file is the 5.1 surround sound track (has dub voices + sound effects + Japanese music). The basic concept behind what we want to do is take the US 2.0 track and subtract the voices and sound effects to leave only the music. In order to do that we first need to isolate the voices alone as well as the sound effects alone from the 5.1 mix.
Instructions:
1.) Import the .thd file into Audacity.
2.) Remove the channel that only has rumbling (the subwoofer, aka ".1" part of the 5.1).
3.) Take the first two channels, and make them into a stereo channel. Take the last two channels and make them into a stereo channel. Duplicate each one, and mute the duplicates (we'll be using them later).
4.) Now, take your two stereo channels (the originals that are not muted), and split them back to mono.
5.) Take your center channel and invert it. Now raise its volume until the Japanese music cancels out. Now select these 5 tracks, click on "Tracks" from the top, and click "Mix and Render". The resulting mono track has dub voices + sound effects.
Now we need to create an isolated sound effects track.
6.) Mute your "dub voices + sound effects" track for now. Unmute those two stereo tracks (the duplicates) from before.
7.) Take the one that has smaller audio waves and invert it. Now raise its volume until the Japanese music cancels out.
8.) Now select these two stereo tracks, click "Tracks" from the top, and click "Mix and Render". The resulting stereo track has isolated sound effects.
9.) Now duplicate your isolated sound effects track. Mute the duplicate for now. Now take the original, and split it to mono. Select the resulting two mono tracks, and "Mix and Render" them into a single mono track. The result is an isolated sound effects track in mono.
10.) Now unmute your "dub voices + sound effects track". You may need to invert your mono sound effects track (it's hard to keep track of what needs inverting and what doesn't). Raise the volume on your mono sound effects track until the sound effects cancel out (as best as you can). Now select these two tracks to "Mix and Render" them. The result is a mono isolated voices track. Now mute your isolated voices track for now.
11.) Import the US 2.0 stereo .flac file. Unmute your stereo sound effects track. You may need to invert it. Now raise the sound effect track's volume until the sound effects cancel out.
12.) Now unmute your isolated voices track. You may need to invert it. It might be very loud so be careful. Adjust its volume until the voices cancel out (you might actually have to lower its volume until the voices cancel out....since its so loud). The result should be the US 2.0 track that only has music!........and some sound effects that (very loudly) leak through here and there.
Now remember, this isn't foolproof and won't work on every single episode, just the ones whose audio is mixed like this. For some episodes, it may just not be possible to remove the Japanese music or sound effects. For those instances, you may only be able to obtain music for those areas that don't have JPN music playing (just invert the center, front stereo, & rear stereo channels against the US 2.0 whenever there's no JPN music and you should still get something decent). Or in other instances, the center channel may already have no music. Just use your intuition and you should be able to figure out what to do in these kinds of cases.
The .flac file for the episode is the US 2.0 stereo track (has dub voices + sound effects + Faulconer music) while the .thd file is the 5.1 surround sound track (has dub voices + sound effects + Japanese music). The basic concept behind what we want to do is take the US 2.0 track and subtract the voices and sound effects to leave only the music. In order to do that we first need to isolate the voices alone as well as the sound effects alone from the 5.1 mix.
Instructions:
1.) Import the .thd file into Audacity.
2.) Remove the channel that only has rumbling (the subwoofer, aka ".1" part of the 5.1).
3.) Take the first two channels, and make them into a stereo channel. Take the last two channels and make them into a stereo channel. Duplicate each one, and mute the duplicates (we'll be using them later).
4.) Now, take your two stereo channels (the originals that are not muted), and split them back to mono.
5.) Take your center channel and invert it. Now raise its volume until the Japanese music cancels out. Now select these 5 tracks, click on "Tracks" from the top, and click "Mix and Render". The resulting mono track has dub voices + sound effects.
Now we need to create an isolated sound effects track.
6.) Mute your "dub voices + sound effects" track for now. Unmute those two stereo tracks (the duplicates) from before.
7.) Take the one that has smaller audio waves and invert it. Now raise its volume until the Japanese music cancels out.
8.) Now select these two stereo tracks, click "Tracks" from the top, and click "Mix and Render". The resulting stereo track has isolated sound effects.
9.) Now duplicate your isolated sound effects track. Mute the duplicate for now. Now take the original, and split it to mono. Select the resulting two mono tracks, and "Mix and Render" them into a single mono track. The result is an isolated sound effects track in mono.
10.) Now unmute your "dub voices + sound effects track". You may need to invert your mono sound effects track (it's hard to keep track of what needs inverting and what doesn't). Raise the volume on your mono sound effects track until the sound effects cancel out (as best as you can). Now select these two tracks to "Mix and Render" them. The result is a mono isolated voices track. Now mute your isolated voices track for now.
11.) Import the US 2.0 stereo .flac file. Unmute your stereo sound effects track. You may need to invert it. Now raise the sound effect track's volume until the sound effects cancel out.
12.) Now unmute your isolated voices track. You may need to invert it. It might be very loud so be careful. Adjust its volume until the voices cancel out (you might actually have to lower its volume until the voices cancel out....since its so loud). The result should be the US 2.0 track that only has music!........and some sound effects that (very loudly) leak through here and there.
Now remember, this isn't foolproof and won't work on every single episode, just the ones whose audio is mixed like this. For some episodes, it may just not be possible to remove the Japanese music or sound effects. For those instances, you may only be able to obtain music for those areas that don't have JPN music playing (just invert the center, front stereo, & rear stereo channels against the US 2.0 whenever there's no JPN music and you should still get something decent). Or in other instances, the center channel may already have no music. Just use your intuition and you should be able to figure out what to do in these kinds of cases.
My Youtube account (Coycoy88)
I primarily rip and upload unreleased Faulconer Productions music, as well as some other American dub music.
I primarily rip and upload unreleased Faulconer Productions music, as well as some other American dub music.
Re: A new vocal removal method for unreleased tracks (lossless rips)
I wonder if we might be able to do this for the original Dragon Ball IF that series were to get a Blu-Ray?
- Attitudefan
- I Live Here
- Posts: 2963
- Joined: Tue Aug 03, 2010 9:51 pm
- Location: Canada
Re: A new vocal removal method for unreleased tracks (lossless rips)
Interesting! I had a feeling it was inverting the tracks. You can find threads of mine where I have mentioned such from the past a looong time ago. I guess with the higher fidelity of the Blu Ray soundtrack, the rips come out better... but then again, the examples I shown were from the original FUNi singles. I am guessing since the track we are using for the inversion is from the blu ray, and thus has clearer sound for each track per each mix, and thus cutting sounds from lower quality sources are still possible? I wonder how they cut Dale Kelly's voice since he his tacks are not available on the 5.1.Naruto6583 wrote:How would they have done it with kai though?yugidmx5 wrote:My rough guess would be that the way they did it was by inverting the audio that had kikuchi ost with the one that had falcouner ost so that the result would leave only music thus cancelling the sfx and voices. Though I'm amazed with the lack of sound artifacts, they did a really good job.
These are a few good examples
Friends Or Rivals: https://www.youtube.com/watch?v=7akDFitsknU
Large Gap: https://www.youtube.com/watch?v=kCwhNnVqmUA
No More: https://www.youtube.com/watch?v=AzF9cRWxstM
Impatience: https://www.youtube.com/watch?v=AlttmFQMHgE
Misplaced Determination: https://www.youtube.com/watch?v=QwVS-MN82Ak
Unnatural Power: https://www.youtube.com/watch?v=rz3ea0QpWqc
And he's also gonna be releasing clear rips of the future trunks arc next year
Thanks for the steps. It is really helpful! I kind of had an idea all this time, but thanks for the clarificationCoycoy88 wrote:We use the blu-ray audio to rip from. I'll just copy and past this.... It's how I learned how to do it.
The .flac file for the episode is the US 2.0 stereo track (has dub voices + sound effects + Faulconer music) while the .thd file is the 5.1 surround sound track (has dub voices + sound effects + Japanese music). The basic concept behind what we want to do is take the US 2.0 track and subtract the voices and sound effects to leave only the music. In order to do that we first need to isolate the voices alone as well as the sound effects alone from the 5.1 mix.
[quote="Coycoy88"Now remember, this isn't foolproof and won't work on every single episode, just the ones whose audio is mixed like this. For some episodes, it may just not be possible to remove the Japanese music or sound effects. For those instances, you may only be able to obtain music for those areas that don't have JPN music playing (just invert the center, front stereo, & rear stereo channels against the US 2.0 whenever there's no JPN music and you should still get something decent). Or in other instances, the center channel may already have no music. Just use your intuition and you should be able to figure out what to do in these kinds of cases.[/quote]
Exactly, so how were they able to cut Dale Kelly's voice, for example?
Looks like it would be possible on the Blu Bricks if the mix in the 5.1 channels allow us to do so. I noticed FUNi mixes the early DBZ arcs differently than the later arcs (at least on the original remastered bricks). The Saiyan arc had more isolation per each track in the 5.1 audio channels than the Cell arc. For example, the Saiyan arc had the sound fx and voice tracks completely separate whereas the later arcs had them mixed together.DB1984 wrote:I wonder if we might be able to do this for the original Dragon Ball IF that series were to get a Blu-Ray?
My favourite art style (and animation) outside Toriyama who worked on Dragon Ball: Katsuyoshi Nakatsuru, Masaki Satō, Minoru Maeda, Takeo Ide, Hisashi Eguchi, Katsumi Aoshima, Tomekichi Takeuchi, Masahiro Shimanuki, Kazuya Hisada
-
- Banned
- Posts: 551
- Joined: Fri Jul 07, 2017 3:04 pm
Re: A new vocal removal method for unreleased tracks (lossless rips)
Any tutorial or possibilité for japanese soundtrack? Maybe this méthode can isolate japanese vocals and bgm for j'ai 96 97 and 98 to create Yamamoto for them
- Attitudefan
- I Live Here
- Posts: 2963
- Joined: Tue Aug 03, 2010 9:51 pm
- Location: Canada
Re: A new vocal removal method for unreleased tracks (lossless rips)
It should work the same way, although isolating Japanese voices may be very difficult without a surround sound set-up. You cannot isolate the voice track in stereo very clearly, if at all.HakkaiBills93 wrote:Any tutorial or possibilité for japanese soundtrack? Maybe this méthode can isolate japanese vocals and bgm for j'ai 96 97 and 98 to create Yamamoto for them
My favourite art style (and animation) outside Toriyama who worked on Dragon Ball: Katsuyoshi Nakatsuru, Masaki Satō, Minoru Maeda, Takeo Ide, Hisashi Eguchi, Katsumi Aoshima, Tomekichi Takeuchi, Masahiro Shimanuki, Kazuya Hisada
- Attitudefan
- I Live Here
- Posts: 2963
- Joined: Tue Aug 03, 2010 9:51 pm
- Location: Canada
Re: A new vocal removal method for unreleased tracks (lossless rips)
Hakkai, you may want to take a look at this regarding the Japanese track. They seem to be able to isolate the Japanese audio decently.
https://youtu.be/w4so6sKzifw
But further discussion regarding stereo Japanese audio should be talked about here
https://youtu.be/w4so6sKzifw
But further discussion regarding stereo Japanese audio should be talked about here
My favourite art style (and animation) outside Toriyama who worked on Dragon Ball: Katsuyoshi Nakatsuru, Masaki Satō, Minoru Maeda, Takeo Ide, Hisashi Eguchi, Katsumi Aoshima, Tomekichi Takeuchi, Masahiro Shimanuki, Kazuya Hisada
Re: A new vocal removal method for unreleased tracks (lossless rips)
This is really interesting, mainly because if I'm understanding this right, a 5.1 version of the Faulconer track could be constructed this way. For anyone interested in such a project, I would love to contribute in any way I could.
This is the episode of when Gokuh enrages himself after Freezer talk shit about Kuririn
-
- Banned
- Posts: 551
- Joined: Fri Jul 07, 2017 3:04 pm
Re: A new vocal removal method for unreleased tracks (lossless rips)
he used the japanese broadcast audio and the 5.1 with japanese score it's not the same , the main pb is that the 5.1 with jap score is that there is some voice remaining, you can't allways keep only the bgm etc...Attitudefan wrote:Hakkai, you may want to take a look at this regarding the Japanese track. They seem to be able to isolate the Japanese audio decently.
https://youtu.be/w4so6sKzifw
But further discussion regarding stereo Japanese audio should be talked about here