[Audio Empire]

On Tue, 4 Sep 2012 10:40:00 -0700, Arny Krueger wrote
(in article ):

"Audio Empire" wrote in message
...
On Mon, 3 Sep 2012 07:54:49 -0700, Arny Krueger wrote
(in article ):

"Audio Empire" wrote in message
...

You've brought-up a good point. When recording digitally, you just
don't
want to come too close to that MSB.

This is an audiophile myth, just as surely as the idea that having
extensive
unused power reserves makes power amps sound better.

I'll tell you what you do, overmodulate PCM so that you're trying to use
more bits than you have and watch what happens!

You seem to labor under the false belief that I don't already know. In fact
the true statement is that not being there, you have no idea about what
happens.

Having heard the effects of overmodulating digital into clipping many years
ago, I "don't go there". Why should I?

Craftsmanship suggests that you don't want to clip, but actual listening
reveals that a few short overages will escape even the most critical ear.

In analog that's true.

Also true with digital. Brick wall filters in the amplitude domain are just
like brick wall filters in the frequency domain - they can remove
significant amounts of program material and as long as it is below the
thresholds of audibility, it doesn't matter.

Thing is, there is no such thing that is as well known for the amplitude
domain as the Fletcher Munson curves are for the frequency domain. However,
people who observe such things know that the human tolerance for clipping is
greater than zero and very dependent on the circumstance.

A certain amount of lore about digital clipping was based on converters that
didn't clip cleanly. As a rule, modern ones do clip cleanly. So much for
thinking rooted in the 1980s...

Well since I never allow clipping, I wouldn't necessarily know that, now
would I?

Especially on a good pro tape recorder at 15 ips.
I've had recorders (like my old Otari MX2020) where you could bang the
needles against their pins momentarily with no APPARENT audible effect
(although I'm sure you could measure it at greater than 3%, you just can't
hear it) In digital recording, that kind of laissez-faire attitude toward
100% modulation is a distinct no-no.

Apparently, only in the minds of people who lack experience with good modern
equipment.

I suspect that I have at least as much experience with "good modern
equipment" as you do.

While a pro analog tape machine can go
over the 0 Vu mark occasionally with little or no consequences, you
never
want to do so in digital.

That's primarily because the position of the 0 dB mark on analog tape
recorders is almost always a judgement call.

It shouldn't be a judgement call. There is an ANSI spec for reel-ro-reel
tape. It's supposed to be: 0 VU = +4dBu = 1.23 V AC RMS or a fluxivity of
320
nWb/m.

That would be a contemporary standard whose conformance is as it is in the
judgement of the people who do the work. Legacy recordings vary for
systematic reasons, and contemporary recordings vary because people are
people.

That "contemporary standard" of 0 Vu = 1.23 VRMS as you call it was first
agreed upon by Bell Labs, CBS and NBC in 1939 originally for network use of
phone lines to carry network programming and for STLs. The coercivity spec
was agreed upon by the NAB (then the NARTB) and SMPTE in the early 1950's. As
far as the use of the coercivity spec is concerned, you can't calibrate a
tape recorder and guarantee it's round-trip frequency response (and
distortion) specs without using a standard calibration tape and every studio
in the country had them. They all conformed to the 320 nWb/m = 0 Vu because
that's the standard. If you don't calibrate your playback side first, you
can't calibrate your record side. Now, it's true that the DIN spec was a
little different, and the British used a different spec as well. But sometime
in the 1960's they converged as I recall (possibly because Willi Studer and
Nagra sold so many of their pro machines here in the USA).