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

New Message Reply About this list Date view Thread view Subject view Author view Other groups

Subject: Re: [linux-audio-user] Fwd: [Jamin] Re: soft clip: Achieving Gain, inconsequential overloads
From: Mark Knecht (mknecht_AT_controlnet.com)
Date: Tue May 04 2004 - 22:08:23 EEST


Paul Winkler wrote:
> On Tue, May 04, 2004 at 10:57:21AM -0700, Mark Knecht wrote:
>
>>>Suppose someone turns up the volume for more harmonics.
>>
>>But what harmonics? Where do these come from, and why?
>
>
> The kind of brick-wall limiting and gain-stage-hammering that
> people do to get "loud" mixes introduces significant harmonic
> distortion. That's "where". I can't tell you "what", I don't
> know the characteristics of the distortion technically.
> As for "why"...

OK then I'm going to 'check out' at this point. None of this thread
makes any sense in my experience. It is completely inconsistent with my
experience which is mostly mixing pop/rock stuff. It seems to me this
has drifted very far into the Twilight Zone.

Certainly brick wall limiters change the harmonic mixtures. Whether they
'add significant harmonic distortion' should be measureable, but whether
that harmonic distortion has anything to do with being 'favorable' I'll
wait for someone with more time and interest to discover.

Far more likely, in my mind, is that the root cause of this is just
basic Fletcher-Munson stuff taking over at a low level and then
inexperienced listeners and musicians letting it run wild. Inexperienced
listeners often report that they like one mix over another when in fact
the two mixes are identical and the preferred is 1 or 2db louder, at
least over certain listening ranges. There is absolutely no significant
harmonic anything going on when that happens. There are tricks the ear
is playing on them when they listen at low levels vs. louder levels. I'm
always surprised (well, not always anymore, but often) when I
recalibrate my environment with my SPL meter and remember how loud 85dbA
is. I tend to listen at much lower levels but I think that most people
do not, and most engineers mix at higher levels than I do also. With
this in mind it's no suprixe that a song would sound better played at
the level it was mixed.

>
>
>>If I record a
>>violin correctly, why do people think they need more harmonics?
>
>
> Wrong choice of instrument ;-) Classical engineers tend to
> lean toward neutrality and accuracy in their recordings.

Why does a violin equate to classical recording? I saw Laurie Anderson
just last evening, but of course that wasn't really a violin. ;-)

>>Just a
>>strange hearing thing that tends to favor having them? Possible, but
>>also possibly black magic. I guess I'm too 'old school'! ;-)
>
>
> I suppose you use only measurement microphones too ;-)

OK, insult taken, but that's OK. I wasn't getting personal. Why are you?

>
> Really it's nothing new. Plenty of "old school" pop/rock/blues/etc
> engineers pushed their analog tape and analog consoles hard to get some
> soft clipping, which both introduces harmonics and acts as a sort
> of peak limiter.
> Now we're still doing much the same thing, only we're using
> high-tech brick-wall limiters.

Or I run my digital out through a D/A and immediate back in through an
A/D to do similar things at times. It's another way to dither. It can
help. It can hurt. It changes the harmonics. It does not (again, in my
experience only) have any effect that drives up the volume, but that may
just be me.

>
> I suspect those old-school pop engineers discovered tape saturation
> and related techniques somewhat by accident. Nobody thought it out
> ahead of time, they just noticed it made some records stand out from
> the crowd, maybe carry a bit better on AM radio, and the next thing you
> know your kids are watching "duck-and-cover" films in school. Oh wait,
> that was a different arms race ;-)
>


New Message Reply About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : Tue May 04 2004 - 22:04:06 EEST