Subject:How To Normalize L & R Channels

Posted by: kooki

Date:6/1/2008 7:52:31 AM

I use SF9s batch converter to process hundreds of wav files.

Part of the processing is to ensure that the files are normalized correctly.
My problem is ... sometimes the L & R channels of a stereo file are not balanced.

Is it possible to normalize the L channel and then the R channel independently? I know it can be done manually, but I need to automate the process.

Any help appreciated.

Subject:RE: How To Normalize L & R Channels

Reply by: Kennymusicman

Date:6/1/2008 9:38:17 AM

A script would be the only way forward that I can think of at this time

Subject:RE: How To Normalize L & R Channels

Reply by: jumbuk

Date:6/1/2008 10:23:19 PM

Would you really want to do this?

EG: LH channel has one REALLY loud sound at time t. Normalising will adjust the entire channel level so that the peak hits zero dB. However, suppose RH channel is uniform throughout, with no big peaks. Entire channel will be lifted above LH average level. Result - RH swamps LH.

If the channels are really unbalanced, the thing to do would be to balance them first. If it's too tedious to do this manually, use a compressor.

Subject:RE: How To Normalize L & R Channels

Reply by: ATP

Date:6/2/2008 11:08:17 AM

agreed with jumbuk. by simply normalizing left and right individually chances are the track will still be unbalanced, perhaps even towards to other side after processing. so your best option is to balance the tracks manually first ... or, if you're lazy, convert all your tracks to mono ;)

Subject:RE: How To Normalize L & R Channels

Reply by: Geoff_Wood

Date:6/2/2008 5:51:25 PM

They are possibly 'unbalanced' because that is the way the source is and should be. To normalise each channel separately would totally stuff up stereo imaging. It is proper that different sized peaks should normally appear at different points in a stereo file, unless all instruments are situated dead centre !

There is no requirement that both channels hit the same peak amplitude, although crass asymetry may (or may not) indicate a problem elsewhere i the original recording. That would be just plain kooki ! (Sorry, couldn't resist ).


geoff