Stereo Tool

Hi everyone, I just discovered this very interesting tool that may help to solve our problem.

We're an Italian cultural association devoted to help emerging artists' growth and since dec 07 we can legally share music (for free teaching purposes) on our website at "degraded quality".
The purpose is to "train" the emeging artists' hearing attitude, but we wanna respect musicians rights too.

So we choosed to put only 1 album (as 1 image file) per band, mono, 32 KHz, Vorbis (q0 setting) encoded files in sharing. Anyway to achieve the best possible quality from this parameters we started a couple of discussions about best downmix to mono and FlatStereo @ Hydrogenaudio Forums.

Now i've tried some presets with the command line tool, but the results is always 2 channel file.

How achieve the goal ?

Can someone construct a batch pipe encoding string in the correct quality-oriented order (with the following tools) ?

aotuv (should be the last)
Shibatch's fast and high quality sampling rate converter (before or after the mono mix ?)

Alternatively, a piping with SoX - Sound eXchange would be great too.

Thanks in advice,
Marco Radossevich

_________________
>>Forward Agency
In progress we (always) trust.

Welcome Marco,

The command line version indeed always produces output files of the same format as the input file (which means stereo). However, if the Phase and Width sliders are set to 0, the output signal itself should be mono (so you should have 2 identical channels). See http://help.stereotool.com/stereo_image ... ator.shtml for details on what the settings do.

I would assume that you can instruct the Vorbis encoder to convert to mono. Because the signal already is mono, you wouldn't loose any sounds as would normally be the case. Alternatively, you could add another program behind Stereo Tool that converts to mono (or throws away one channel).

I don't know any of the other tools that you are referring to, so I can't quickly create a pipeline. About the placement of Stereo Tool: It doesn't really matter where you put the stereo to mono conversion. The command line version of Stereo Tool assumes the input to be "somewhere near" 44.1 kHz (32 should be close enough), if it's really far off (say 16 kHz or 192 kHz) you might get some artifacts in the sound.

Does this help?

I "simply" need this pipeline:

FLAC decode > Stereo Tool downmix > Vorbis encoding

About parameters: should i set "fix AZIMUTH" and "Center Bass" to on ?

Thanks in advice, again...

_________________
>>Forward Agency
In progress we (always) trust.

About the Azimuth and center bass:

If you have the Phase and Width sliders set to 0 then "Center bass" has no effect (is basically overrides both sliders with '0' at very low frequencies - but they are set to '0' anyway).

Azimuth correction is only useful if you're using tape or vinyl recordings, where there's a phase offset ERROR between the two channels. If that is never the case, it's better to leave it turned off. (It will only consume extra CPU power). Even if you have such tape or vinyl recordings, the effect of the Azimuth correction filter is nearly nonexistent when the Phase and Width sliders are set to 0. So I would just leave it turned off.

The pipeline that you describe now sounds much easier than what I understood from your previous post. I cannot look into it right now, but maybe the following will help: A pipeline for MP3 decoding and encoding (using LAME):

lame --decode in.mp3 - | stereo_tool_cmd - - C:\st\settings.sts | lame - out.mp3

Note: '-' indicates standard input or output. So by using the pipe symbols, the output of lame (sent to standard output) is piped into the standard input of Stereo Tool, and Stereo Tools output is piped into the standard input of Lame. Other command line tools might not support this - in that case you'll have to use intermediate files instead. In that case this would look like:

lame --decode in.mp3 lameout.wav
stereo_tool_cmd lameout.wav stout.wav C:\st\settings.sts
del lameout.wav
lame stout.wav out.mp3
del stout.wav