[linux-audio-user] Fwd: [Jamin] Re: soft clip: Achieving Gain, inconsequential overloads

[tim hall]

On Tuesday 04 May 2004 19:43, Paul Winkler wrote:
> I have a (completely untested) theory that part of the reason is
> due to the distortions that happen in the ear at high listening
> levels (mentioned earlier in this thread).  

This bit interests me. I don't have chapter and verse on it, but I do know 
that this is more than just our personal theory. It possibly has been tested 
in work with hyperactive kids. If I find the reference, I'll post it. I don't 
remember where this idea came from, so don't hold yer breath (or quote me on 
it :-).

As for the rest of this thread it seems like we would need to ground our terms 
a bit better before we can carry on this debate. If you've had enough of this 
thread, you'll be stopping reading about here:-) 'Loudness', 'relative 
volume' and saturated bandwidth and increased harmonics seem to me to require 
looking at separately with an appreciation of the difference between analog & 
digital clipping, brickwall limiting. Again CMIIW - I'm just learning in this 
field, so I'm really just mirroring back what I've understood of this 
conversation.

The lesson I'm getting out of this is that the more 'classical' approach as 
put forward by Mark, Joern & others is that basically when you make 
adjustments, you want to cut rather than boost in order to maintain the 
integrity of the sound, that's if at all. Ideally you have a good room + mic 
sound. If you are recording digitally, then you might use a brickwall limiter 
to protect you from digital clipping. (I think you all know what I mean and I 
think we'd all agree, it's horrible:-) Basically you're aiming to make the 
recording sound like 'Real Life(TM)'.

The 'pop' approach is basically to compress the life out of everything. Again 
I'm not expecting too much disagreement if I say that the nicest possible 
sound here is achieved using some valve compression / distortion and possibly 
saturated analog tape. I recently did a session where the drums and bass got 
compressed to fsck, you can hear the compressor breathe so hard it sounds 
like it has asthma. It's a great recording. It doesn't have to be 'Loud' to 
achieve the desired triggering of the inner ear (see above), say 80dB above 
'silence' (whatever THAT is). It doesn't have to be ear blistering. The same 
sort of effect can be gained from 'singing bowls', which produce a consistent 
and harmonically 'rich' sound. I'm also promted to think of the rather atonal 
chanting of male tibetan monks. I think we're looking at beneficient 
harmonics rather than 'loudness' or the unpleasant forms of distortion / 
noise that are _so_ easy to create, given injudicious boosting of signal 
levels.

That much I think I understand, the rest of this discussion has lost me.
How's my driving?

cheers

tim hall