Help Wanted: The Prequel Edit Problem - remove the music & sound effects without removing the dialogue...

One of the biggest, and seemingly insurmountable issues all fan editors of the Prequel Trilogy seem to face, is the addition of new music and dialogue to the films. Since the sound has obviously already been mixed, the creative direction that prequel fan edits can go to is very limited.

If this seemingly insurmountable problem was somehow corrected, however, hundreds of possibilities for altering the storyline of the prequels would be introduced.

I know this subject has been brought up before. Many have claimed it’s impossible, to remove the music and/or sound effects from the films without removing the dialogue as well. But through my perusing online, I’ve found several instances where aspiring editors have done just that.

http://www.youtube.com/watch?v=W8ycYOpltqM

I contacted the editor of this video, but he never got back to me. HOWEVER: this video proves that the removal of music can be done. Now, I suppose the question is: HOW can it be done?

This, to me, is the most important question that fan editors need to answer. It’s imperative if we wish to take this medium to another level. In this thread, I hope we can discover how to accomplish said task.

I've not worked with The Clone Wars, but I believe they have fully separated dialogue that can be extracted, unlike the prequel films.

Even so, there has to be a way around this problem.

If science can give us the power of flight, and the security of longer life spans, it can give us the ability to dub over Jar-Jar Binks. 

You cannot remove dialog if it's mixed in the same channel with the music and sound effects.  I'm assuming Clone Wars has a separate channel for dialog hence the easy extraction.

The same thing goes for music.  You can use a phase inversion technique to remove everything but the center channel, which is usually the vocals, but traces still remain of the music in most instances.

You can't just magically extract something after its been mixed.  It's like trying to remove the egg from a baked cake.  :)

I can not fly or live longer than a few thousand years.

Aalenfae and I made some experiments on this matter. Thing is you can't fully isolate the voice if mixed,  but it is possible to reduce the music to very low levels and to mask it with the new score

All you need is the complete score exactly as it sounds in the movie, audacity and patience using the noise reduction filter in different levels.

Oh, I gave up the recipe ;)

If only it were that easy.

I too would love to figure this one out but one thing is for sure there is no magic wondertool for the job.

To get any results at all must take loads of hard work so hats off to anyone who gets anywhere with this.

Well, it's not easy, but it's possible to get a decent result if you choose the accurate track to do the masking

So, if it's somewhat possible...maybe we should get a team of aspiring fan editors to begin working on this issue? 

And then release the altered audio track for all other fan editors to use freely? Because I think we should begin working towards that. 

So, whoever has the ability collaborate on a project like this, please come hither to this thread with suggestions and progress. Perhaps we can begin to take this idea in a positive direction. 

It is a shame that groups of people haven't broken the films into elements.

It would take ages but one done we could have loads of interesting fan projects because there would be less duplication of effort.

Editors or fan film makers could pull one dialogue only track down or pull a clean background with no model shots or clean model shot and just mix away.

You might be able to use spectral sound removal which will be built in to the next version of Adobe Audition.  It might take a lot of time, but maybe with this it could become possible.

The feature is shown in the first 6 minutes of the video.

http://tv.adobe.com/watch/adobe-at-nab-2013/adobe-audition-overview/

I know there are also some plugins that can do a similar thing at the moment, but if anyone has Creative Cloud (like me) you wouldn't have to pay anything more.

This Removal of frequencies will only work providing what you are trying to remove does not clash with what you are trying to keep.

If the Dialogue is at the same frequency as a piece of background music then you have no chance of removing it.

The example shown in that video is fine because the police siren is nowhere near the frequency of the dialogue, but a full orchestral score elements of it will be at the same frequency as the dialogue so this won't fix the problem unless you have separate audio tracks. 

This is what I have found.

Ronster said:

This Removal of frequencies will only work providing what you are trying to remove does not clash with what you are trying to keep.

If the Dialogue is at the same frequency as a piece of background music then you have no chance of removing it.

The example shown in that video is fine because the police siren is nowhere near the frequency of the dialogue, but a full orchestral score elements of it will be at the same frequency as the dialogue so this won't fix the problem unless you have separate audio tracks. 

I know that is an issue, but generally using the spectral analyze you can make more detailed selections.  Unless, there something I don't know of, frequency removal is the only way through mono content assuming you are working only with the center channel.

It should be possible though since our brains can seperate the dialogue, music and sound effects from eachother.  We just need to figure out how that works software wise.

emanswfan said:

Ronster said:

This Removal of frequencies will only work providing what you are trying to remove does not clash with what you are trying to keep.

If the Dialogue is at the same frequency as a piece of background music then you have no chance of removing it.

The example shown in that video is fine because the police siren is nowhere near the frequency of the dialogue, but a full orchestral score elements of it will be at the same frequency as the dialogue so this won't fix the problem unless you have separate audio tracks. 

I know that is an issue, but generally using the spectral analyze you can make more detailed selections.  Unless, there something I don't know of, frequency removal is the only way through mono content assuming you are working only with the center channel.

It should be possible though since our brains can seperate the dialogue, music and sound effects from eachother.  We just need to figure out how that works software wise.

This software comes out soon?

MrInsaneA said:

This software comes out soon?

Do you mean that sarcastically?

If not, then yeah the new Adobe stuff comes out May or June I think.

Otherwise, I do remember the other tool.  I've tried the trial version (which is fully featured, you just can't export) and it can work miracles.  I or someone else could test it out on a scene or two.  http://www.izotope.com/products/audio/rx/

emanswfan said:

MrInsaneA said:

This software comes out soon?

Do you mean that sarcastically?

If not, then yeah the new Adobe stuff comes out May or June I think.

Otherwise, I do remember the other tool.  I've tried the trial version (which is fully featured, you just can't export) and it can work miracles.  I or someone else could test it out on a scene or two.  http://www.izotope.com/products/audio/rx/

No sarcasm here, friend. And I would be eternally grateful if you could test it out. 

If you have access to multiple language tracks with identical music then it should theoretically be possible to filter out what matches between the tracks -- which would be the music. However, some speech may be lost where the frequency of the music overlaps with the speech and some of the music may remain because of audio compression artifacts. The process should improve the more language tracks you use. I think that this would be better than a typical voice-removal filter.

I don't know if there is a software package that allows you to do this ...

Analysis should be done in some frequency domain, and preferably according to the transform used by the codec ... if the tracks used the same codec, which I don't think that they do.

Am I correct in that for all movies, the English track is in DTS-HD (lossless) and that the others are either in DTS 5.1 or in Dolby Digital 5.1 ? The DTS-HD could be transformed into the frequency domain of any of the other. I have not checked, but I doubt that DTS and Dolby Digital would use the same transforms.

Darth Lars said:

If you have access to multiple language tracks with identical music then it should theoretically be possible to filter out what matches between the tracks -- which would be the music. However, some speech may be lost where the frequency of the music overlaps with the speech and some of the music may remain because of audio compression artifacts. The process should improve the more language tracks you use. I think that this would be better than a typical voice-removal filter.

I don't know if there is a software package that allows you to do this ...

Analysis should be done in some frequency domain, and preferably according to the transform used by the codec ... if the tracks used the same codec, which I don't think that they do.

Am I correct in that for all movies, the English track is in DTS-HD (lossless) and that the others are either in DTS 5.1 or in Dolby Digital 5.1 ? The DTS-HD could be transformed into the frequency domain of any of the other. I have not checked, but I doubt that DTS and Dolby Digital would use the same transforms.

Hmm...that's something I didn't think of there.  Theoretically it sounds completely possible, I'm just not sure where there would be an analysis tool to do that.  I'll look into that some more.

Yes, 99% of blurays all have lossless main tracks, which are either DTS-HD, Dolby True HD, or LPCM.  Today there are more lossless sub tracks for spanish and french though too, but mostly they are DTS 5.1 or Dolby Digital 5.1.  You can use the tracks as frequency domain no matter if lossless or not.

But here's the issue: New English 6.1 mix, Old Spanish, etc. 5.1 mixes

you might get lucky on a 5.1 audio channel mix for removal... 

But it's not like picking sultanas out of a cake that's baked.

The mix is all a whole and your not actually seperating anything whatever frequency you remove from the sound will remove everything on that frequency.

Even if you have a partial clash across frequencies with what you delete the voice may become hollow sounding or Boomey depending if you slice the top or the bottom end of respectivly

so you'll find sometimes you can do it and other times you will have a total clash or partial clash of frequeies and won't be able to remove the soundtrack score.

Actually I can fly but I don't like showing off.

I'd love to see some test videos from any of you who have ideas on how to do this. 

I don't have an account or time to prepare and upload the file, It's 23 min long.

Talked to Aalenfae tho, and he said that if he has time will trim it and upload it somewhere.

Mithrandir said:

I don't have an account or time to prepare and upload the file, It's 23 min long.

Talked to Aalenfae tho, and he said that if he has time will trim it and upload it somewhere.

great, would love to see what he's done.