Very well. I shall be waiiiiting.

If you are re-releasing rips with a new high quality algorithm, that would be fantastic! I wish you the best of luck. If possible, my recommendations include:

-Ep1 - When Kaiba leaves the game shop
-Ep3 - When Yugi receives the invitation, when Yugi and Joey discuss their friendship on the roof, and the fanfare as the boat departs
-Ep4 - When Yugi and Weevil begin their duel (its squeezed between the Millennium Duel cue)
-Ep8 - When Yugi duels Mokuba
-Ep9 - When Kaiba first enters the computer room
-Ep10 - When Yugi duels the now revealed 'ghost' Kaiba (Sounds similar to the first part of that rip from the BBT Movie)
-Ep11 - When Joey duels Rex (Sounds similar to the second part of that rip from the BBT Movie)
-Ep21 - When the Paradox Brothers play Riyoku (Its a remix of when Rex and Weevil duel in Ep2)

Hey guys, just some cool news I guess. I decided to try something I never did before. Make a GUI!
This will be very interesting for me to learn and also for you guys to work on your audio files. :)

I started already and have a good idea how to do it. I will update you guys at the end of May with what I have,
and if it works nicely, I can then develop tons of algorithms so you guys can enjoy what I have and always
have the latest state-of-the-art tools at your disposal :)

Let me know if this is something that is interesting to you guys. Thank you! :)

I would be greatly interested, good luck.

Just a small edit I made, it won't be a VST, but a Graphical User Interface (GUI). The reason being I tried with VST and for the algorithms I am using, it is not possible in real-time. A GUI makes more sense, you load wav file, process it, preview it, if happy, save results, if not, change parameters, then process and save, etc. It is actually easier for me than a VST anyway. Will let you know soon! :)

I am improving v1.0 of my stereo speech separator, but here is quick sample from Yugioh, not the music,
but the speech extracted. It is 1 minute long cue and processed in 3-4 seconds.

Look at that isolation quality magic : isolated speech sample (http://www.mediafire.com/file/h6j1wdzz813y4cc/magic1.wav)

More demos before release soon! :)

I am improving v1.0 of my stereo speech separator, but here is quick sample from Yugioh, not the music,
but the speech extracted. It is 1 minute long cue and processed in 3-4 seconds.

Look at that isolation quality magic : isolated speech sample (http://www.mediafire.com/file/h6j1wdzz813y4cc/magic1.wav)

More demos before release soon! :)

Nice extraction im looking for the exact score in that exact clip you just posted. Great speech extraction I have high hopes with your extractions brother keep up the good work and I look forward to hearing some actual scores especially the one from this clip you shared. :)

Nice extraction im looking for the exact cost in that exact clip you just posted. Great speech extraction I have high hopes with your extractions brother keep up the good work and I look forward to hearing some actual scores especially the one from this clip you shared. :)

Hey, thank you for feedback. There is always room for improvement and actually Kharaxel sent me some challenging ones that I still need to tackle, so I will try to do that before releasing the application.

Hey, that's pretty great! Extracting the voices themselves in such quality is something I also desired. Now I'm super interested.

Hey, thank you for feedback. There is always room for improvement and actually Kharaxel sent me some challenging ones that I still need to tackle, so I will try to do that before releasing the application.

no problem only nitpick id have in relation to the voice extractions is the abrupt end of some dialogue. It is complete but ends abruptly almost unnaturally. I don't think theres a way to fix that unless you fade out the end of the voice but even that I don't think will achieve the natural flow of speech. besides that its all good. I doubt the isolated score will have such abrupt endings as it messes with listening experience.

Oh I know exactly what you mean and I agree, those abrupt ends are because of an additional post-processor that removes additional energy in the isolated audio. It is also a field I have to improve on, either with the algorithm itself or implement another. There is no need to worry about that, by the time I release this application, the isolated speech will feel natural, like you say. This was just a very quick demo so you can see the idea or direction of where this is going. The application itself will be released around end of May, still experimenting with newer algorithms. Will keep you guys up to date on this, because this is just going to be the first app, I plan on releasing a lot more, for maybe different purposes, etc. stay tuned :)

Oh I know exactly what you mean and I agree, those abrupt ends are because of an additional post-processor that removes additional energy in the isolated audio. It is also a field I have to improve on, either with the algorithm itself or implement another. There is no need to worry about that, by the time I release this application, the isolated speech will feel natural, like you say. This was just a very quick demo so you can see the idea or direction of where this is going. The application itself will be released around end of May, still experimenting with newer algorithms. Will keep you guys up to date on this, because this is just going to be the first app, I plan on releasing a lot more, for maybe different purposes, etc. stay tuned :)


Awesome I look forward to it, applications aside though hopefully you'll be releasing some of your edited unreleased cues swell. :) I have a Mac and wonder how said application would even work without additional software?

That is a good question. As of right now, I am developing for 64-bit Windows. So this is one requirement, the other is just a library, but that library is on all platforms. The thing here is that you would have to just use vmware or dual boot windows or something, the applications by themselves are not crazy, they are small, but sometimes, they might require lot of RAM to operate. This is something I am also testing while working, checking the memory usage, and if I can optimize code or/and change to another algorithm that does very similar job, but with way less memory needed. So will see about that.

and as for the cues, ye, as soon as I am happy with the app, I will start doing the cues :)

This level of voice isolation would be great for those montages with the anime characters "narrating" Yu-Gi-Oh Pro moves.

Hey guys, I have something that might interest the participants of this thread.

Namely, I have managed to acquire the 4Kids Music To Duel By album in lossless FLAC format, plus the japanese Dark Side of Dimensions OST (in regular mp3 format though).

The OP can contact me whenever possible, to add them to the first post's MEGA folder, if he so wishes.

Hey guys, I have something that might interest the participants of this thread.

Namely, I have managed to acquire the 4Kids Music To Duel By album in lossless FLAC format, plus the japanese Dark Side of Dimensions OST (in regular mp3 format though).

The OP can contact me whenever possible, to add them to the first post's MEGA folder, if he so wishes.

Sure, just send me a PM, however this is for dub, I am not interested in the Japanese OST for the movie tho, but send anyway :)

Sure, just send me a PM, however this is for dub, I am not interested in the Japanese OST for the movie tho, but send anyway :)

Hey. I'm afraid I'm not allowed to send PMs (I think my acc is too new) so I guess I can drop the MEGA link here.

https://mega.nz/#F!xXhDzQqY!Sa67q1IRiFvi8EvwxtKmhA

Thank you very much, I will include it eventually, just really busy these days, but thanks a lot, I always wanted the FLAC to this OST :)

<3

also: I have good and bad news.
Bad news, will release my app late June, still testing and trying new methods and tricks.
Good news is I am working on some new method that improves certain aspects of the separation,
especially on the music side. Will show some demos soon! stay tuned guys :)

And there goes June.

And there goes June.

Haha, I know.

I remind myself of those game developers when they announce a game way ahead of development and then people have to wait forever for the game to come out, cause it was announced too early. I mean, I guess I was just excited that I didn't judge well when to release this in a time frame. I am literally everyday spending a lot of time with the cues you sent me and some others while at the same time trying all sorts of things. btw fantastic Megaman music on some of those, it's a shame there is no OST for them :(

I am very confident that I will release the first version of the app by this next weekend (somewhere around July 8).

I am still trying to figure out some stuff regarding reverberation effects and sound effects, two of the biggest problems here.

I can deal with background music, but what's left is usually either the reverb or/and some sound effect or/and some small trace
of background music still not suppressed. Separating speech from complex background is one of the hardest challenges, but I am
very happy with the progress I made.

I remember the old days when I used software to do voice isolation, even stuff like Audacity...
my god.... those were the bad days.... that bad quality isolation... but hey, everything can improve, I feel like I have a PhD compared
to those days in terms of audio source separation knowledge that I gathered from doing this kind of hobby.

It's really fun for me, but again, it's a hobby, it's not like anyone is willing to pay me for this to keep me motivated or even for this to
become a professional business. But then again, who knows, I am still young, I might end up working somewhere or making my own
company that deals with these audio separation problems.

Sorry for long text :(

Aside from this hobby, I am busy with other things and real work, so I need time for this.
Don't worry, it will be released soon in some form, and then the app will improve from there.

Thank you for understanding!

*anxiously awaits some kind of news*

The app will be released tomorrow guys! Just need to add some small documentation to it and the website thingy and all good :)

Greatness is almost here!

The app will be released tomorrow guys! Just need to add some small documentation to it and the website thingy and all good :)

Greatness is almost here!

Hopefully with the release of the app (for PC users at least) I'm really hoping we get some new cues or cues which show the apps true nature in isolation., One can hope. :)

Yayz! Cannot wait to try out.

July 9 2018:

v1.0 of my speech/music separation app is finally released! More apps coming
soon and I will improve current apps also! Check it out here:
https://danielkorg.github.io/ :)

Its a shame being a Mac OS user I cannot even experience or use this app, I do wish that some tracks from yugioh will be shared by those who have used this app or even if anyone is taking suggestions on what to test it with I can give certain scene and episodes of other series like sherlock BBC series. I can't wait for this to be ported to Mac OS or at least more new yugioh tracks to see the capabilities of this app. :)

I will work on the Mac OS AND Linux versions soon! Might release them this coming weekend! :)

Atenzor! Atenzor!! ATENZOR!!!
This app is a fucking beauty and I love it to bits! Once I have some money (currently kinda tight this month) I'll gladly donate some to you! I will let you know when I can so you can send me your paypal or whatever.
For anyone wondering, this actually works as it should. Though please bear in mind the quality of the source you want to separate also matters!

I am happy the app helps you man, can only get better from here!

Yes, the higher quality source you have, the better results.

I will work on the Mac OS AND Linux versions soon! Might release them this coming weekend! :)

Awesome brother, I can't wait to use it and hear the magic. So If ripping from a blu ray I assume 2 channel to 5 channels are the best ways to go? Im positive its not good to use web rips of streamed shows for sources correct?

Awesome brother, I can't wait to use it and hear the magic. So If ripping from a blu ray I assume 2 channel to 5 channels are the best ways to go? Im positive its not good to use web rips of streamed shows for sources correct?

If you have access to a multichannel such as 5.1 or higher, I am pretty sure you don't need my algorithm, since most likely you can easily isolate the speech in Audacity or something by looking at each channel and remove the Center one. But if not, you would have to downmix to a stereo (2-channel). My algorithm doesn't work for 1 channel (mono) or higher than 2 (stereo). However, I could make this possible in the future if it's something useful for people, will see.

If you have access to a multichannel such as 5.1 or higher, I am pretty sure you don't need my algorithm, since most likely you can easily isolate the speech in Audacity or something by looking at each channel and remove the Center one. But if not, you would have to downmix to a stereo (2-channel). My algorithm doesn't work for 1 channel (mono) or higher than 2 (stereo). However, I could make this possible in the future if it's something useful for people, will see.

Oh that absolutely makes sense when it comes to the yugioh episodes I own its weird because the audio is in stereo with 2 channels but when you try to remove the audio it doesn't work so im hoping to use the algorithm for that but also a 2 channel blu ray rip of season 4 for abc sherlock series.

Oh that absolutely makes sense when it comes to the yugioh episodes I own its weird because the audio is in stereo with 2 channels but when you try to remove the audio it doesn't work so im hoping to use the algorithm for that but also a 2 channel blu ray rip of season 4 for abc sherlock series.

Hey, I finally released the Mac and Linux versions on my site, go grab it!
https://danielkorg.github.io/

:)

Hey, I finally released the Mac and Linux versions on my site, go grab it!
https://danielkorg.github.io/

:)

Awesome brother, I'll follow your instructions and hopefully I'm not too dumb to not figure it out hahahaha. Hopefully it supports Mp4, mkv, files

Awesome brother, I'll follow your instructions and hopefully I'm not too dumb to not figure it out hahahaha. Hopefully it supports Mp4, mkv, files

Input for now is ONLY wav files. Nothing else. I actually am not sure if I could make it work directly with a video file.

You would have to demultiplex the video into video and audio, process the audio and then remux back everything together, assuming this
is what you want. Only thing I could do in future versions for sure, is allow other audio formats, like mp3, flac, etc.

Idk, will see and investigate this..

Input for now is ONLY wav files. Nothing else. I actually am not sure if I could make it work directly with a video file.

You would have to demultiplex the video into video and audio, process the audio and then remux back everything together, assuming this
is what you want. Only thing I could do in future versions for sure, is allow other audio formats, like mp3, flac, etc.

Idk, will see and investigate this..
No need if it isn't necessary brother. Also any plans on updating that yugioh cues folder? Its empty hahahaha I hope others share their algorithms for some yugioh cues. Keep up the great work man its epic.

With the tool that I released, even tho it's not in the final phase, people can also rip cues now :)

I am so busy with developing apps now, and improving this one! And of course, busy with other things...

Maybe when I finish and am personally happy with some algorithm, I will then start doing rips again.
It will happen for sure, just not sure when :P

With the tool that I released, even tho it's not in the final phase, people can also rip cues now :)

I am so busy with developing apps now, and improving this one! And of course, busy with other things...

Maybe when I finish and am personally happy with some algorithm, I will then start doing rips again.
It will happen for sure, just not sure when :P

No problem brother its understandable and makes sense. I can't wait, hopefully though im not holding my breath we might get some more cues from the composers who knows, But I look forward to awesome shares from those who manage to rip cues from the series. :)

btw fantastic Megaman music on some of those, it's a shame there is no OST for them !

You said it... That was why I sent them to you. Pretty much my holy grail of "lost ost's".

I shall check out your proggy. (Crazy how matlab requires so much space... welp, time to lay that internet connection to waste xD)


Note: I didn't realize there existed such online services aswell: https://phonicmind.com heh. odd.



Its a shame being a Mac OS user I cannot even experience or use this app

Can't you use Bootcamp in conjunciton with a Windows x64 installation ?

can somebody extract the flute type sound? (the one where the transformation sequence starts).

That app sounds awesome, will try to rip the GX background music from the dub since its never been released.

It seems like this thread is dead now unfortunately :( With all the anniversaries Yu-Gi-Oh! has had you'd think they would release a genuine post from the series to celebrate it. But no dice, heck there isn't even any new cues released for a long time, I wish that there would be some more, I amongst many others, it'd be nice to complete this set. If I had the info of the right person to contact I'd reach out myself, sadly it doesn't seem to be the case. :(

Thanks for the music, the US score of Yu-Gi-Oh is great.

It seems like this thread is dead now unfortunately :( With all the anniversaries Yu-Gi-Oh! has had you'd think they would release a genuine post from the series to celebrate it. But no dice, heck there isn't even any new cues released for a long time, I wish that there would be some more, I amongst many others, it'd be nice to complete this set. If I had the info of the right person to contact I'd reach out myself, sadly it doesn't seem to be the case. :(

Well in terms of the right person to contact, you can try contacting 4K Media, or check the credits from here https://www.imdb.com/title/tt0249327/fullcredits/?ref_=tt_ov_st_sm
under the music/sound department at the bottom, and try to find those people (facebook/twitter/etc). But it's not like we didn't already try all that. Either they don't reply back or when they do, it's very minimal info or nothing at all really. We were lucky in the earlier days when some of the composers shared some of the cues, but it most probably won't happen again.

In the meantime, you will have to wait for high quality rips.

Well in terms of the right person to contact, you can try contacting 4K Media, or check the credits from here https://www.imdb.com/title/tt0249327/fullcredits/?ref_=tt_ov_st_sm
under the music/sound department at the bottom, and try to find those people (facebook/twitter/etc). But it's not like we didn't already try all that. Either they don't reply back or when they do, it's very minimal info or nothing at all really. We were lucky in the earlier days when some of the composers shared some of the cues, but it most probably won't happen again.

In the meantime, you will have to wait for high quality rips.

Hi Atenzor. When are you going to be releasing V2.0 of your Deep Vocal Attractor program? I noticed you have removed v1 from your website because of problems such as bugs and the like. I would like to know if/when you playing on releasing the next version as I would love to test it out. Do it seperate sound effects as well from a audio track? Thanks.

Hi Atenzor. When are you going to be releasing V2.0 of your Deep Vocal Attractor program? I noticed you have removed v1 from your website because of problems such as bugs and the like. I would like to know if/when you playing on releasing the next version as I would love to test it out. Do it seperate sound effects as well from a audio track? Thanks.

Yes, I plan on releasing v2 very soon, just so busy these days. Hopefully somewhere next month.

It has many improvements over the first, to tackle different scenarios, but of course, some mixtures
really are complicated to separate. Lucky for us, there are 224 episodes right? So, if one mixture seems
too hard to extract the music, I am pretty sure one can find another episode with the same music,
but with less complexity.

More info/updates soon!