[LAU] maximum input level, or normalization and dc offset correction?

Fons Adriaensen fons at linuxaudio.org
Sat Jul 13 11:44:10 UTC 2013


On Fri, Jul 12, 2013 at 02:59:03PM -0400, Bill Gribble wrote:
 
> AFAICT "best practice" with modern equipment is to basically work as if
> you were on an analog console; that is, clipping at no less than +20 dB
> with respect to your "0 dB" working level.  That means you record with
> signals working at RMS around -20dBFS and then you just don't worry
> about clipping unless your signal is truly hugely dynamic, and in that
> case you make allowances as necessary with gain and/or limiting. 

Agreed 100%. With 24 bits you have a lot of headroom, and any
'digital gain' used post-recording won't affect the real S/N
ratio at all. 

A working level of -20 dB RMS FS will preserve the full dynamic
range even if you have to boost the signal later.

A high end preamp/AD such as the RME Micstasy will have a S/N
ratio of around 120 dB maximum when configured for a high level
line input. With typical gain settings for e.g. recording an 
acoustic instrument it will be around 100 dB. Most mic preamps
will give you much less, and we're ignoring mic self noise and
ambient noise. In many cases the latter will determine the real
S/N ratio you can have. If you're recording electronic instruments
they and the DI-box will determine the noise floor. It won't be
near anything that 20 bits can't handle.

Anyway, for music with a restricted dynamic range (that includes
almost everything presented/discussed here) this is completely
academic. You could record the tracks at -40 dB FS and nobody
would notice.

Ciao,

-- 
FA

A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)



More information about the Linux-audio-user mailing list