From: Rob Danielson <>
> Mic, mic pre quality (low noise and low harmonic distortion
> particularly) and favorable A-D characteristics ("quality" vs low
> price ) do seem to be larger factors than 16 vs.24 bit recording
> resolution, but I'm not sure my experience bears out Pohlmann's
> assertion that field resolution is effectively maximized close to the
> performance of 16 bits (if I understand this correctly). I've had a
> half dozen experiences with bringing home both 16 and 24 bit files
> from the same location through the same front end. It's easier to
> work with the 24 bit files compared to the 16 bit files. There's more
> "body" providing more subtle tonal/amplitude differentiation making
> equalization changes easier and more effective. The lower the file
> saturations, the more pronounced the difference. I do up-sample 16
> bit recordings to 24 bits in post when I know they will go through
> further changes downstream, but this does not create a file that
> handles as well as a 24 bit orig. All plugs are converting/working
> with 32 bit floating regardless of orig sample rate of course. Maybe
> there are better A-D units/methods out there today than Pohlmann was
> basing his analysis on. It looks like the 4th edition was printed in
> 2000. Rob D.
You have misread what I put up of Pohlmann's info. He states a limit of
20 bit resolution, which is 4 bits better than 16 bit. So, there would
be differences if using a 24 bit, just not 24 bits worth of difference.
I have strong doubts anyone has gotten around this point of Pohlmann's.
It's based only on the voltage difference involved in a 24 bit
resolution compared to the inherent thermal noise voltage variation of
resistors. The damage occurs in the filters on the analog half of the
A/D that are part of the A/D process. That voltage variation will exceed
the step voltage, wiping out the fine step's accuracy.
My read out of what he says is that only the high quality 24bit chips
will get the full 20 bit resolution. So, cheap 24 bit recorders will
probably exhibit lower quality in this regard than good ones. And even
the good ones the resolution limit will be somewhere around 20 bit. And,
of course, any higher bit recorders will be pure hype. Even now, a cheap
24 bit recorder may not be giving a significant gain over a quality 16 bit.
The way around this limit is complex. He discusses other A/D designs
than straight linear, and also discusses their noise problems too. I
believe that, based on what I've read, farther increases in digital
accuracy will have to come from moving away from linear. Go to targeted
systems where the accuracy is distributed more like it's needed. It
will, of course, require getting around those fine voltages to do it.
Walt
________________________________________________________________________
________________________________________________________________________
|