[LAU] Improving a bad recording

Jörn Nettingsmeier nettings at folkwang-hochschule.de
Fri Jun 11 23:58:12 UTC 2010


On 06/02/2010 12:10 AM, Philipp wrote:

> 2) a very loud noise, possibly from some part of a laptop. To my
> knowledge the recording was made in two takes, guitar and voice, on a
> laptop with a cheap mic of some sort. The noise seems to be located
> mainly somewhere around 12kHz, a region which seems to be important for
> the clarity of the voice. Any attempts to just pull down  with an eq
> (4-band parametric for example is what I tried today) had a significant
> impact on the voice.

have not tried on mp3s, but i think it would work on .wavs, assuming the 
voice is dead center mono, no important instruments are dead center 
mono, and the noise is uncorrelated, i.e. different on both channels.

convert the L/R signal into an M/S signal: M = L + R, S = L - R. plus 
here means "mix", minus means "invert the phase of R and mix".

with luck, you will have voice and not much else in M, and the rest in 
S. if the noise crept in after the voice recording, chances are it's 
"stereo noise". if the original voice track was noisy, you lose.

if it indeed is "stereo noise", try to kill it in the side channel only 
with aggressive notch filtering. the clarity of the voice should be 
mostly unaffected, since it resides in M.

now reconvert back to L/R: L = M + S, R = M - S.
while you're at it, you can also play with the base width, by changing 
the M:S ratio before converting back.

how well this trick works on mp3 depends largely on how well the stereo 
information is preserved in the mp3 (joint stereo, discrete, or even 
mono)...

this method is how naive karaoke filters work: eliminate center content.
another good way to accomplish this is to plug a stereo jack (that 
carries unbalanced stereo) into a balanced line input. frequently seen 
in the wild, usually unintended :-D




More information about the Linux-audio-user mailing list