Audacity Forum

Hi,

I have 538 audio files that have been converted to MP3 from tapes. They all have an annoying hiss sound (very low but noticeable when listening to the audio with headphones).

I can remove the noise on individual files with the Noise Removal effect on Audacity 1.3.4 and it works great. Now I want to automate the process to the remaining files but I cannot get the new chain NoiseRemoval command to get the noise profile. I have read the wiki but I don't seem to find a way to do it. Help!

Thanks.

We have the same exact problem (help anyone?) However, in some cases the chain does work. What we did was first select one file, get the noise profile, then save it as an .aup project, close audacity (which annoyingly seems to be necessary to be able to apply chains), then apply the chain to all files in the folder (which all contain the same sort of noise). Oddly enough, this procedure first worked perfectly, now it doesn't anymore - no clue why.
What we would like to know is where and how the noiseprofile is stored. When we apply the noiseremoval process in our chain, where does it get its noise profile?

Thanks for any useful advice,

Laura Menenti

I have tried to capture the Noise Profile they way you described and in my installation it does not work. I also tried with a clear copy of windows and running as an administrator, in case it has anything to do with it, and it doesn't.

One thing that I noticed though is that when I apply a chain, the very first file is loaded and a window pops up saying something like "creating noise profile". But that happens only with the first file. Anyways, I exported my noise profile to a new file and renamed some files as "01.mp3", "02,mp3", "03.mp3" and so on, being the noise profile the file "01.mp3" and tried again. No luck!

Any ideas?

Thanks,
Xavier.

I tried this on a hunch:

Keep in mid I'm running Linux (Ubuntu 7.10)

Set up the chain just as I did one audio file 'by hand'
Exported 2-3 seconds of representative noise from an audio file call it something like '00 noise'
(not sure if there is any re-sorting by Audacity when it runs the chain)
Applied the chain to a selection of split out files named 00 noise, 01 xxxx, 02 xxxx, ...
Seams like it took the first file as the pilot for generating the noise profile and remembered it that way for the rest of the selected files.
Seamed to work fine.
Hope this helps.