Extracting channels from a vma?

[Foofaraw]

Say you have a 5 channel wma pro - is there some tool which can extract each channel as a seperate sound file?

[tebasuna51]

You can try:
- BeHappy, AviSynth with BassAudio decoder and WavSplit mono output
- Foobar2000, configuring WavSplit like external decoder or using WaveWizard to extract the channels of the multichannel wav output.

[Foofaraw]

You can try:


The phrase suggested you don't know if its going to work?


- BeHappy, AviSynth with BassAudio decoder and WavSplit mono output
- Foobar2000, configuring WavSplit like external decoder or using WaveWizard to extract the channels of the multichannel wav output.

All of that sounds like decoding - I just want to preserve the original audio.

[Midzuki]

All of that sounds like decoding - I just want to preserve the original audio.

Do you mean, convert a multichannel .WMA into a set of mono .WMAs,
losslessly ??? Even if you accepted a Lossless WMA output,
I think you could not do that without the help from non-Microsoft tools.

[tebasuna51]

The phrase suggested you don't know if its going to work?
Work with my samples.

All of that sounds like decoding - I just want to preserve the original audio.
Is not possible.

[Foofaraw]

Do you mean, convert a multichannel .WMA into a set of mono .WMAs,
losslessly ??? Even of you accepted a Lossless WMA output,
I think you could not do that without the help from non-Microsoft tools.

Yes losslessly (not necessary the same bitpatten) - but you don't think that is possible, right?

[Foofaraw]

Work with my samples.


Is not possible.

So you managed to use some of those programs to reencode each of the 5 channels to a separate file only containing that channel?
By which software? All of the ones you mention?
(and why isn't it possible - seemed the old cooledit pro could almost do it, but it has a problem with too large file sizes it seems)

[Midzuki]

Yes losslessly (not necessary the same bitpatten) - but you don't think that is possible, right?

Possible it is, indeed, but perhaps not in the way you may be thinking of.

I mean, WMA Lossless is not the same as WMA Pro,
it's a different codec. So yes, one can go from a 5.1 WMA Pro file to 6 mono WMAL files. But, is this what you really want?

Besides, what would be the point of such conversion ? :confused:

[tebasuna51]

So you managed to use some of those programs to reencode each of the 5 channels to a separate file only containing that channel?
By which software? All of the ones you mention?
(and why isn't it possible - seemed the old cooledit pro could almost do it, but it has a problem with too large file sizes it seems)

If you want edit an audio with cooedit (or other audio editor) always need decode to uncompressed format (wav file or any internal format decoded by the editor) and after recode to the desired compressed format. You can't edit a compressed format (only a few parameters or cuts can be done without decompress).

If you are an AviSynth user I recommend you BeHappy, with the last audio decoder versions of NicAudio/BassAudio plugins.

If not maybe Foobar2000 is more easy to use.

[Foofaraw]

Possible it is, indeed, but perhaps not in the way you may be thinking of.

I mean, WMA Lossless is not the same as WMA Pro,
it's a different codec. So yes, one can go from a 5.1 WMA Pro file to 6 mono WMAL files. But, is this what you really want?

Besides, what would be the point of such conversion ? :confused:

What I really want is to take a 6 channel wma pro file and downcode it to a well balanced stereo image. If it was an ac3 file thats quite easy with the ac3tool - but with wma i thought i had to do some multitrack mixing to get it to a stereo image.

[Foofaraw]

If you want edit an audio with cooedit (or other audio editor) always need decode to uncompressed format (wav file or any internal format decoded by the editor) and after recode to the desired compressed format. You can't edit a compressed format (only a few parameters or cuts can be done without decompress).


I think there might be language barrier problems here.
I just want to go through as few decompression recompressions to avoid quality suffering.


If you are an AviSynth user I recommend you BeHappy, with the last audio decoder versions of NicAudio/BassAudio plugins.

If not maybe Foobar2000 is more easy to use.

I've tried to avoid avisynth like the plague, but perhaps i'll have to. Are those plugins free or commercial?

[tebasuna51]

What I really want is to take a 6 channel wma pro file and downcode it to a well balanced stereo image.
This is the first info you must say. You need decode to downmix. You can decode and downmix to a stereo wav file (uncompressed audio data) with BeHappy or Foobar2000.
After you need a WMA encoder (I can't help you with this).

I think there might be language barrier problems here.Sorry for my bad english, but you need learn some audio concepts also.

I've tried to avoid avisynth like the plague, but perhaps i'll have to. Are those plugins free or commercial?
Then use Foobar2000. There are very good downmixer plugins.
All software I recommend is free.

[Foofaraw]

This is the first info you must say. You need decode to downmix. You can decode and downmix to a stereo wav file (uncompressed audio data) with BeHappy or Foobar2000.


But can you control the mix of the individual channels?


After you need a WMA encoder (I can't help you with this).


Got that covered thanks.



Sorry for my bad english, but you need learn some audio concepts also.


Sure, but I was referring to me - it seemed I couldn't communicate what I was looking for.


Then use Foobar2000. There are very good downmixer plugins.
All software I recommend is free.

I'll have a look at it. Thank you for your suggestions.

[tebasuna51]

But can you control the mix of the individual channels?

With AviSynth you can write a script with the desired mix.

With Foobar2000 plugin foo_dsp_downmix you can control the channel volume mix and apply, or not, a phase shift for back channels.

[Foofaraw]

With Foobar2000 plugin foo_dsp_downmix you can control the channel volume mix and apply, or not, a phase shift for back channels.

I'm trying foobar2000.

Under "Available DPSs", the one I should use is called "Downmix AC3/DTS" yes?

[tebasuna51]

Under "Available DPSs", the one I should use is called "Downmix AC3/DTS" yes?

Yes, and configure it.
The downmix formula used is:

fl' = cf x FL + cc x FC + cl x LF + ca x SL + cb x SR
fr' = cf x FR + cc x FC + cl x LF - cb x SL - ca x SR

Where
cf = Front Level
cc = Center Level
cl = LFE Level
ca = Surround Level A
cb = Surround Level B

For instance the DPL II downmix coeficients:
cf = 1.0000
cc = 0.7071
cl = 0.0000
ca = 0.8660
cb = 0.5000
------
3.0731

Master Gain between:
0.325 (1/3.0731, max volume to ensure no clips)
1.0 (High volume with high clip risk)

[Foofaraw]

Yes, and configure it.
The downmix formula used is:

fl' = cf x FL + cc x FC + cl x LF + ca x SL + cb x SR
fr' = cf x FR + cc x FC + cl x LF - cb x SL - ca x SR

Where
cf = Front Level
cc = Center Level
cl = LFE Level
ca = Surround Level A
cb = Surround Level B

For instance the DPL II downmix coeficients:
cf = 1.0000
cc = 0.7071
cl = 0.0000
ca = 0.8660
cb = 0.5000
------
3.0731

Master Gain between:
0.325 (1/3.0731, max volume to ensure no clips)
1.0 (High volume with high clip risk)

Oh dear now I get all confused with this math.

It seems the plugin was preconfigured to:

cf = 1.0000
cc = 0.7070
cl = 0.7070
ca = 0.8165
cb = 0.5774
--------
3.8079

So that is wrong?

Master gain was set all the way to 1.

Then there is the Phase shifting.
Default was set to "enabled DTS" - is that ok or shoudl it be "disable ac3"?

[tebasuna51]

So that is wrong?
My short answer is yes. But is your choice. In the same thread you can read my opinion. Resumed reasons:

1) The cl coeficient:
Dolby recommend don't use LFE channel in dpl downmix.

2) The pair ca, cb:
A old study in this forum prove than the modified cofecients obtain a better channel separation when decode. Was assumed by Besweet-Azid dpl II downmix. Tested by me and used also in BeHappy.

3) The master Gain:
If all the channel have a peak at same time the signal is clip. You can select between low volume or many clips. Maybe values between 0.5 and 0.6 can be recommended.

4) The phase shift for surround channels:
If the source is a Dolby compliant ac3 you don't need a new phase shift (already shifted) but with other sources a phase shift is recommended.

[Foofaraw]

My short answer is yes. But is your choice. In the same thread you can read my opinion. Resumed reasons:

Yeah, its just i don't quite understand it :) But I'll try your numbers.


3) The master Gain:
If all the channel have a peak at same time the signal is clip.


The peeks of the sound is clipped? Ie, loud sounds are distorted?

[Tiziano]

Hi tebasuna51

I followed your soggestions in the other topic and I'm here to ask some question about, so that no other topic with the same arguments are opened.

I would like to understand one thing about AC3 compliant file to be imported in Foobar2000.

I produce an AC3 file with Nuendo ( 6 channels = 5.1 ) and I import it as well in Foobar, with the plug in "Downmix AC3/DTS", and I try to listen to this file, what I obtain its a pulsy noise very nasty ... I tried also with .DTS file obtaining the same thing.
I tried to downmix those files into a DPLII and the result is very nasty ... because there is not music but only the "ts-ts-ts-ts" noise ...

In order to be able to obtain something, I had to save the file NOT as AC3 or DTS but as normal interleaved 6 channels .WAV file

My question is: In this case, do I have to set-up phase shifting in the plug-in "Downmix AC3/DTS" or not ?

I thank you very much

[tebasuna51]

I don't know how was make your source in Nuendo.

But:
If the SL channel is independent from the SR channel don't need phase shift
If the SL channel is dependent (and without phase shift already applied) from the SR channel need phase shift

[Tiziano]

I don't know how was make your source in Nuendo.

But:
If the SL channel is independent from the SR channel don't need phase shift
If the SL channel is dependent (and without phase shift already applied) from the SR channel need phase shift

What do you mean with "dependency" please ?

Because both channels are recorded in separate tracks ...
Sometimes I have effects in which the SL is panned to SR and viceversa ... but they are discrete tracks ... I treat SL and SR in the same manner i treat FL and FR ...

[tebasuna51]

Because both channels are recorded in separate tracks ...
If was recorded with different microphones in oposite position you don't need phase shift.

If you have a doubt try the two options.