Automate these tasks and apply them to a set of files

Good evening,

I’m a member of a free association. I use Audacity 1.3.5 and would like to improve the quality of tens of vocal files, each 40 min long, by the same actions with the same parameters, the recording being made in the same conditions :

1/ Noise removal based on a noise sample located in a file,
2/ high-pass filter
3/ low-pass filter
4/ Compressor with normalization
5/ Amplify each and every sentence separately - there is always some silence between the sentences
6/ Silence removal
7/ MP3 export

Each time, processing a file manually takes 30 min, would like that my PC performs all these actions alone during the night.

Tried to move forward:
A/ How to remove the noise with a sample file as a parameter (1st step)?
B/ Several above tasks are not available in batch mode ; how to make them available?
C/ How to call a compression function from the Nyquist prompt?
D/ How to call a Nyquist plugin from the Nyquist prompt (any doc ?)

Any solution?
Any pointers much welcome.


Noise removal does not work well in chains as the noise profile should match exactly the noise present in the specific sample.

The high pass and low pass filter can be done reasonable well with the Equalization effect if you create custom presets.

Nyquist effects can not be called in chains - only (some) of the built in effects can be used in this way.

You do not have to “call” Nyquist effects from the Nyquist prompt - just enter the Nyquist command. For example - to apply a low pass filter at 1000 Hz to a selected region of a track - open the Nyquist prompt and type:

(lp s 1000)

then press OK.

In this example, “lp” is the command for a first order lowpass filter. “s” tells Audacity to pass the selected sound to Nyquist. and 1000 is the frequency in Hz.

For more information on Nyquist - see these sections in the Audacity wiki:

Thanks SteveTheFiddle. In between have downloaded the source code, plus the required MS tools. Will see whether I’d implement these things or not… Would mean to get back to my C++ developer old reflexes… Is there any effort ongoing on implementing those functions in the batch commands?


The built in effects in Audacity seem to be gradually appearing in the Chains options. I don’t think anyone is working on making chains available for 3rd party plug-ins.

Do you believe the previous set of commands is interesting in comparison to the CleanSpeech Chain?
(I mean … from a sound processing point of view)

Have just discovered the clean speech mode of Audacity.
Interesting, to take the noise once for all and applying the noise reduction on a full set of files.