SoundOut 1.1.1 - Sound Output Plugin [10/15/07]

[G_M_C]

Probably allready made, but i'd like to request a feature to output 6 (7/8) separate WAVs (one for every channel). Is this possible ?

[sh0dan]

@All: I hope to be able to put out 1.2.0 beta, with a huge amount of bugfixes, and probably some new ones. I'm looking at FAAC native output, but I was in doubt whether it was worth it, or if the quality of FAAC was too bad.

@G_M_C: I will have a look at it, sounds like a nice feature!

[G_M_C]

Probably allready made, but i'd like to request a feature to output 6 (7/8) separate WAVs (one for every channel). Is this possible ?

@G_M_C: I will have a look at it, sounds like a nice feature!

That would be great !

:)

[tebasuna51]

Probably allready made, but i'd like to request a feature to output 6 (7/8) separate WAVs (one for every channel). Is this possible ?

You can use WavSplit.exe (in BeHappy package) and:

SoundOut(output="cmd", filename="e:\xxx.wav", autoclose=true, showprogress=true, type=0, format=0, executable="d:\Program\wavsplit.exe", prefilename=" - ")

You obtain xxx_FL.wav, xxx_FR.wav, ...

[G_M_C]

You can use WavSplit.exe (in BeHappy package) and:

SoundOut(output="cmd", filename="e:\xxx.wav", autoclose=true, showprogress=true, type=0, format=0, executable="d:\Program\wavsplit.exe", prefilename=" - ")

You obtain xxx_FL.wav, xxx_FR.wav, ...


:)

Thx, didn't know you could pipe soundout's results to another programm like that. I'll give it a try when i do my next project later this week !

[tebasuna51]

:)

Thx, didn't know you could pipe soundout's results to another programm like that. I'll give it a try when i do my next project later this week !

You can send the output to encoders like NeroAacEnc (http://forum.doom9.org/showthread.php?p=966892#post966892) or others (http://forum.doom9.org/showthread.php?p=942513#post942513) with STDIN support.

[G_M_C]

You can send the output to encoders like NeroAacEnc (http://forum.doom9.org/showthread.php?p=966892#post966892) or others (http://forum.doom9.org/showthread.php?p=942513#post942513) with STDIN support.

Didnt know that :o

Thats that is a nice feature, that could save some time on some projects. I asked Sh0dan aubout the WAVs-output because i was looking for a a programm that could convert my audio (and video) to 24.000 fps. EAC3To doesnt speedup/slowdown to 24.000 as yet, so i looked at SoundOut, but it had no "seperate channel WAV-output". I needed that because i use Surcode's encoders, that accept separate channel input.

So in reality: The only thing missing for my situation is a hook to pipe soudouts output straight to Surcode's encoders ;)

[DrP]

I haven't seen this mention before, but when using ver 1.1.1 and MP2 output when ever I select bitrate=384 the opening application crashes after a short time and a truncated MP2 file of 256kb length is created. Any other bitrate works correctly. The source audio is from a type 2 DV avi file, 48k stereo.

But I have found an oddity. When using hcenc 023 to encode the avs video output to MPEG2 the audio saved by soundout has frequent bursts of noise. This doesn't happen if I use virtualdub to create an avi but still use soundout to create an audio file.

[Gavino]

When using hcenc 023 to encode the avs video output to MPEG2 the audio saved by soundout has frequent bursts of noise. This doesn't happen if I use virtualdub to create an avi but still use soundout to create an audio file.
Have you tried setting silentblock=false ?

[tebasuna51]

So in reality: The only thing missing for my situation is a hook to pipe soudouts output straight to Surcode's encoders ;)

Sorry but SurCode need physical monowavs files in disk.
You can use SoundOut with WavSplit, configure SurCode with Registry and launch SurCode. After you can delete the monowav files.

WaveWizard and eac3to can launch SurCode (I don't know how).

[G_M_C]

Sorry but SurCode need physical monowavs files in disk.
You can use SoundOut with WavSplit, configure SurCode with Registry and launch SurCode. After you can delete the monowav files.

WaveWizard and eac3to can launch SurCode (I don't know how).

Seems i have made an error while testing my DR-players capabilities. After my initial tests i concluded that i had to convert audio and video to 24fps, and because of that i really needed soundout to output monowav's/separate wavs per channel.

But it seems i've made an error while expirimenting; There seems to be less need to covert framerates, so i have less use for the separate wav-option i asked for. Sorry about that anyway.

But that doesn't mean it's not a good feature to have. There are may cases where it would be very nice to have such a feature; For instance where you have to convert from NTSC to PAL or vise versa, and you really want to keep the audio as DTS !

So, even isf there is less need for me to have this feature, i would still like to "have it handy", cause it is a good feature to have :)

[DrP]

Have you tried setting silentblock=false ?

Yes. It didn't make any apparent difference.

[sidewinder711]

Thanks for the upgrade of Soundout... it is a very neat tool!

So far I used the following command line:

SoundOut(output="mp3",filename="e:\test\audio-1.mp3", autoclose=true,showprogress=true,mode=2,cbrrate=128)

But now I would like to use Soundout for some videos having two different language/audio channels. I read the documentation but it is still unclear for me how to change the command line to get these 2 channels extracted (preferrable to mp3 or ac3). Any help appreciated.

[vlada]

@All: I hope to be able to put out 1.2.0 beta, with a huge amount of bugfixes, and probably some new ones. I'm looking at FAAC native output, but I was in doubt whether it was worth it, or if the quality of FAAC was too bad.

Sorry for my late answer to this point. I've been on holidays.

I think it makes sense to incorporate FAAC into SoundOut. I'm working on a GPL application which converts video to many formats including MP4s for cellphones. So I need a free AAC-LC encoder and I don't care about quality that much. I will leave the current option to use external NeroAAC encoder. So yes please, if you can incorporate FAAC, do it.

Also for the 1.2 version please try to change settings for MP3 encoding to those suggested by lame developers (http://lame.sourceforge.net/lame_ui_example.php). The vbrpresets are depricated and quality setting in range from 0 to 9 should be used.

A very good source for suggested lame settings is at HydrogenAudio (http://wiki.hydrogenaudio.org/index.php?title=LAME#Recommended_encoder_settings).

Also I'm not sure whether the settings for Ogg Vorbis are correct. The correct parameter should be a quality index, which is a real value from -2 to 10. It is described here (http://wiki.hydrogenaudio.org/index.php?title=Recommended_Ogg_Vorbis).

[smok3]

i have stereo in, and i would like to mix to center, but remain stereo, how would i do that in avisynth?

edit, nm, got it;

#input stereo audio, ouput stereo center-mixed audio
video = (somefile.avi) # avi has sound as well
audio = ConvertToMono(video)
l_ch = GetChannel(audio, 1)
r_ch = GetChannel(audio, 1)
stereo = MergeChannels(l_ch, r_ch)
audiodub(video, stereo)

probably
stereo = MergeChannels(audio, audio) would work as well.

[wuziq]

I just wanted to say THANKS for this awesome plugin.

FYI, and I'm sure everyone knows this, but I just realized that if you want to simply open an AVS script to allow SoundOut() to do its thing without actually processing the video, just use vdub.exe:

vdub.exe yourFile.avs

:)

[wuziq]

What are the exact PC requirements for using SoundOut?

The reason I ask is because I am getting inconsistent results on one of my machines when I use specified format output (not command line, not GUI), eg., MP3 or WAV. On that machine, sometimes the file is perfect, but most times there are one or two bursts of loud noise at the beginning. I can't seem to get any consistent, reproducible behavior.

That machine is running Windows XP Pro SP2, AviSynth v2.57, SoundOut v1.1.1.. dual Intel Xeon 2GHz.. 2GB RAM.. not sure what other details to provide (maybe someone can ask the right questions).

When I use the SoundOut GUI, the audio is always fine.

When I demux the WAV audio in vdubmod, the audio is always fine.

Not sure what else to say. Anyone ever experience something like this?

[vlada]

@wuziq
I wouldn't suspect your hardware for this. If AviSynth's output is correct and SoundOut output is corrupted, it seems like a bug in SoundOut.

@sh0dan
Any news regarding SoundOut 1.2?

[wuziq]

Question about AC3 channel ordering on the SoundOut wiki page:

Does acmod describe how SoundOut expects the input source channels to be ordered? If I use acmod=7, would I need to put the LFE channel as the last channel in my input source?

[tebasuna51]

Question about AC3 channel ordering on the SoundOut wiki page:

Does acmod describe how SoundOut expects the input source channels to be ordered? If I use acmod=7, would I need to put the LFE channel as the last channel in my input source?

Nope, you must preserve the standard uncompressed channel order:
FL.FR,FC,LF,BL,BR
or
FL.FR,FC,LF,SL,SR
is the same for 5.1

The encoders must accept uncompressed input at their default order and Aften encoder is compliant with this requirement.

[AlanHK]

I was looking at the SoundOut wiki page (http://avisynth.org/mediawiki/SoundOut) and saw

Installation and Usage
The filter is implemented as a plugin (included in AviSynth distribution since v2.6).

And no link for those of us living in 2009 and still using 2.58.

So after Googling my way here I added a link to the beginning of this thread, which has a download link.

I hope no one is annoyed, but it seemed a bit like vapourware as it was.


And a question, or perhaps a request:

Is there a way to script "analyze"?

If not, how about something like
SoundOut(output = "Statistics", filename="c:\Statistics.txt")

Also, the "Analyze" button now gives results like:

[Ch 0] Maximum:-14.58dB. Average:-34.33dB. RMS:-30.98dB. ReplayGain:8.34dB
[Ch 1] Maximum:-14.73dB. Average:-34.22dB. RMS:-30.88dB. ReplayGain:8.38dB

[All channels] Maximum:-14.58dB. Average:-34.28dB. RMS:-30.93dB

Why doesn't the "All channels" line include an average for ReplayGain?
Yes, it's trivial to work it out, but seems an odd omission.

[patrick_]

I'm getting what seems to be a deadlock when using showprogress=false:

this is my entire script:
Soundout(output="cmd", filename="E:\test\VTS_01_1.aac", overwritefile="Yes", showprogress=false, type=0, executable="E:\App\neroaacenc.exe", prefilename="-q 0.3 -ignorelength -if - -of")

The CommandLine Output window shows up (BTW is it possible to make that not happen?), but it will never close. It does respond to input (it's possible to scroll) but it's imposible to close the window. I have to close the calling application to close the output window. If I remove showprogress=false, it does work well (including using autoclose).
I tried using avs2avi and using custom code.

Thanks in advance.

[Blue_MiSfit]

This is such a tremendously useful plugin!!!

~MiSfit

[zee944]

I'd like to use SoundOut to analyze 6ch AC3 files, no encoding yet.

LoadPlugin("NicAudio[2.0.4].dll")
LoadPlugin("SoundOut[1.1.1].dll")
NicAC3Source("audio3.ac3")
ConvertAudioToFloat()
SoundOut()

Is this a reliable way to do it?

And is this the right channels order: FL, FR, C, LFE, SL, SR?

[tebasuna51]

Yes and yes.

BTW, you don't need ConvertAudioToFloat() because the NicAc3Source output is already 32 bit float.

[wuziq]

OK, this is a little random, but I just found a fix for a problem I was having with SoundOut(), and thought I'd post it here in case someone is having the same problem.

The problem: loud glitches near the beginning of SoundOut output.

The fix: Use KillVideo() before calling SoundOut(), and include "addvideo=true" in SoundOut() arguments.

[tebasuna51]

The problem: loud glitches near the beginning of SoundOut output.

I never see this problem, can you provide a sample?

[MadRat]

I've been using AVISynth to split, normalize and recombine the channels, then use SoundOut to save it for further processing. However, from time to time I notice I get echo when saving as a WAV file. Any idea what might be causing this?

[AlanHK]

I notice that if I do "Analyze Sound" that SoundOut gives a report on each channel, starting from zero:
Channel[0]
Channel[1]
etc.

However, Avisynth's GetChannel() numbers from 1. So the corresponding channels are off by one.

Shouldn't SoundOut conform to GetChannel's numbering?

[tebasuna51]

I've been using AVISynth to split, normalize and recombine the channels, then use SoundOut to save it for further processing. However, from time to time I notice I get echo when saving as a WAV file. Any idea what might be causing this?
Can't be a SoundOut effect.

¿How do you recombine the channels?
Sometimes the surround channels are the front channels with a delay (and others effects) and maybe you listen a echo.

[AlanHK]

I wrote this a year ago, no response.

I notice that if I do "Analyze Sound" that SoundOut gives a report on each channel, starting from zero:
Channel[0]
Channel[1]
etc.

However, Avisynth's GetChannel() numbers from 1. So the corresponding channels are off by one.

Shouldn't SoundOut conform to GetChannel's numbering?

Since Soundout is going to become an internal function, surely it should use the same channel numbering as other Avisynth functions?

If there is a reason for this not to be so, please explain.

EDIT:
Checking the "current" download, it hasn't been updated for three years. So I guess there isn't much chance of any response from the author, let alone an update.

[a451guy451]

I'd love to see this thing get updated too. It seems, if it really is going to be included in v2.6, that there must be some kind of recent work done on it.

[a451guy451]

When I try to make MP2 files at CBR 384kbps, they all seem to cut short and produce 256kb files. VBR gives me 0kb files. Anybody else experience this at all? I've tried it on a few different machines. Seems to be the only "broken" part of this plugin I can find (but it could just be me).

[Raptus]

Thank you for this plugin.

I'm also experiencing hangs on exit when using showprogress = false. Thats with an AVS script fed into x264, writing the audio to WAV. With showprogress = true and autoclose = true it works as expected.

[Wilbert]

I compiled it for the latest AviSynth 2.6 (with VC2005 express): http://www.wilbertdijkhof.com/SoundOut26-1.1.1.zip

I will commit the changes soon and I hope i didn't break anything ;) So please let me know if something is not working anymore which used to work before.

A final note. Getting this to compile was one thing, but to fix bugs or add new features to it (i like to see an aac encoder for example) is way over my head.

[Keiyakusha]

A final note. Getting this to compile was one thing, but to fix bugs or add new features to it (i like to see an aac encoder for example) is way over my head.
Thanks for the update!
Not that important, but while you at it, maybe you can at least update some libraries? FLAC and wavpack could have a nice update. Not sure if there is any point to touch other formats. (personally I don't see the reason to use this plugin for anything other than creating intermediate files, where lossless/uncompressed is the way to go)
Also no one needs AAC encoder. The most used ones that are topping listening tests can't be included in this plugin anyway.

[tebasuna51]

Also no one needs AAC encoder. The most used ones that are topping listening tests can't be included in this plugin anyway.
Can't be included but we always can use the command line output syntax:

SoundOut(type=1, format=1, executable="D:\Program\qaac.exe", prefilename="-V 99 --ignorelength -o", postfilename="-")

[Wilbert]

Thanks for the update!
Not that important, but while you at it, maybe you can at least update some libraries? FLAC and wavpack could have a nice update.
Yes I will try so. May take a while though, i want to do other things first. Btw, Sh0dan downgraded FLAC tot v1.2.0 on purpose. What's the deal with that?

Also no one needs AAC encoder. The most used ones that are topping listening tests can't be included in this plugin anyway.
You mean because of licensing issues? I was thinking about fdkaac. Is that one not good enough?

[Keiyakusha]

Can't be included but we always can use the command line output syntax:

SoundOut(type=1, format=1, executable="D:\Program\qaac.exe", prefilename="-V 99 --ignorelength -o", postfilename="-")

Yes, of course... but I don't really see why would you want to go to such extreme instead of using this encoder without avisynth...
Edit: besides how will it really work? There will be some note in the readme that says: here is a new feature, aac encoder. But for that you absolutely have to use commandline! Also QuickTime or winamp (in case of fgh) are required for it to function... Meh

You mean because of licensing issues? I was thinking about fdkaac. Is that one not good enough?
Well, yes and no. If we are to trust guys from hydrogenaudio, the most popular encoders are the one from apple (for which qaac or qtaacenc are used to get access to) or fghaac from winamp. Neither is opensource. If not these two, FDK is probably what one would use, but here some licensing issues may arise. I don't really follow this project, but from what I understand there are some serious issues so that people are staying away from distributing it in binary form. Which is kinda required for the avisynth plugin... People are even building automated scripts for users so that it will fetch the source code, mingw and will automatically build the encoder for them. There have to be a good reason for that.

[Wilbert]

If not these two, FDK is probably what one would use, but here some licensing issues may arise. I don't really follow this project, but from what I understand there are some serious issues so that people are staying away from distributing it in binary form. Which is kinda required for the avisynth plugin...
I stumbled on the reason, so i though i will post it here. I guess the consensus is point 3 of their license: https://launchpad.net/ubuntu/trusty/+source/fdk-aac/+copyright makes it (L)GPL incompatible:

"You may use this FDK AAC Codec software or modifications thereto only
for purposes that are authorized by appropriate patent licenses."

See for example http://patches.libav.org/patch/24437/ or http://comments.gmane.org/gmane.linux.gentoo.devel/79027.

[Emulgator]

SoundOut 1.1.1 commandline usage format is documented wrong !

format=-2 throws an error..
format=-1 delivers a 32-bit float header only.
format=0 delivers 8bit Unsigned Integer
format=1 delivers 16bit Little Endian Signed Integer
format=2 delivers 24bit Little Endian Signed Integer
format=3 delivers 32bit Little Endian Signed Integer
format=4 delivers 32bit float
format=5 makes Avisynth hang forever.

If format is unspecified it delivers source format unchanged, fine (in my test case 24bit Little Endian Signed Integer)

Who has access to the documentation may correct that.

[tebasuna51]

format=0 delivers 8bit Unsigned Integer
format=1 delivers 16bit Little Endian Signed Integer
format=2 delivers 24bit Little Endian Signed Integer
format=3 delivers 32bit Little Endian Signed Integer
format=4 delivers 32bit float

I can confirm that, but for me:

format=5 delivers 32bit float also.
format = 6 -> error.

Using AviSynth+ r2728