Archive of the former Yahoo!Groups mailing list: MOTM
Subject: Re: Delay Module (somewhat OT)
From: "paulhaneberg" <phaneber@...>
Date: 2002-02-27
On the subject of 24, 20 and 16 bit converters. I run a commercial
studio as well. Everything is so-called 24-bit digital Pro-Tools.
However from my own experience my Apogee 24-bit converters sound
better than the Digidesign 24-bit converters. 24 bit is a huge
improvement over 16 bit. However I cannot hear any improvement
between 44.1kHz and 96kHz sampling rates. Studies by the AES and
theory in general back this up. And for me since all my studio work
either ends up on CDs or in commercials, sampling at 96k or 192k is
counter productive because sample rate conversion is imperfect and
the end results in the case of CDs is at 44.1k.
This brings me back to the idea of a MOTM style digital delay module.
Varying the sample rate between 48k and 192k is only a 4:1 range.
I'd like to see at least 64:1 and maybe as much as 256:1 under
voltage control.
I was thinking about using the 10th bit of an ADC as a clock source
and driving the ADC with a MOTM VCO sawtooth, much like the miniwave
is driven. With the VCO at 20k the clock would be at around 10MHz.
drop the frequency of the VCO 8 octaves and the clock would be
around 40kHz. This would be a 256:1 control range. The problem is
in finding an ADC that will operate over such a range with enough
bits for decent audio. I'm not sure if there is such an animal.
The rest of the design is not that hard. I do realize that there
may be a jitter problem if the sawtooth is not linear enough.
I'd like to have as many bits as possible, but as I said before, I
don't think this thing has to have audiophile quality to be useful,
as long as it's better than the BBDs. I used to have an old Emulator
II. It sounded pretty good and it was only 8 bits (but with some
sort of compression/expansion scheme something like the old dbx
noise reduction units I think.)
The other possibility is some sort of 1-bit scheme. Thanks for the
comments Tony. I am aware of the old DeltaLab stuff and I was
thinking along those lines as well.
A whole series of module possibilities opens up if a high quality
voltage controlled delay can be realized.
I could always make my own flash converter, only 65,536 coomparators
needed for 16 bits. (Ha Ha :/)