24 Aug
2009
24 Aug
'09
9:43 a.m.
Fons Adriaensen wrote:
I guess this is the leading point. Even professional IO amps at times
need calibration.
We should imagine a setup like 'sound card --> tube pre-amps -->
mastering recorder'. For this scenario a VU meter needs to be an
external one.
Ralf
On Sun, Aug 23, 2009 at 07:27:18PM +0100, Dan Mills
wrote:
'Peak level' vs 'intrinsic level':
* 'dbFS' refers to 'peak level', 'full-scale square wave'
* 'dbFS RMS' refers to 'intrinsic level, 'full-scale sine wave'.
This is how is in Germany distinguished and because of the words on
English, it might be an international way to distinguish it.
Lets say your card is aligned so that 0dbFS =
+18dbu (EBU standard),
then 0Vu = +4dbu = - 14dbFS, so a software VU calibrated for 0Vu =
-14dbFs should read the same as an external Vu calibrated for +4dbu =
0Vu. If it does not then either a calibration setting is off somewhere
or one of the meters is faulty.
True.
But even a definition such as 'dB FS' is ambiguous, and
it's easy to make mistakes as a result of that.
Consider a sine wave that is just below digital clipping.
This would be called '0 dB FS', but the actual RMS level
is 3 db lower. I've seen at least one context where this
same signal would be called -3dB FS. Which somehow makes
sense as well.
AFAIK the first interpretation is the more common one.
actually
fairly common with professional cards.
And it avoids a lot of problems. Semi-pro cards will not
have the correct levels, but some can be quite consistent
between channels. For example my Terratec EWS88MT has less
than +/- 0.1 dB variation between its 8 channels, both
for input and output. But the actual level is just +3.5dBu
for a FS sine wave.
The real misery starts when (as reported in a recent post
on LAU or LAD), a user finds four volume controls between
his audio file and the physical output (plus of course
a fifth one on his amplifier), and then of course gets
confused on how to set all of them.