naturerecordists
[Top] [All Lists]

[gear] one bit technology for field recording [long msg]

Subject: [gear] one bit technology for field recording [long msg]
From: "oryoki2000" oryoki2000
Date: Tue Apr 3, 2007 6:02 pm ((PDT))
While most discussions assume that a PCM (WAV file) recording is
uncompressed, in fact PCM is highly processed and reduced in size from
the original digital form.

Digital recorders begin the analog to digital conversion by employing
a process of modulation called "sigma delta."  This process samples
the analog voltage from the preamp at a very high rate (usually 2.8
million samples per second).  The output is one bit only: a zero if
the analog voltage is lower than the previous sample, or a one if it
is higher.

In a PCM recorder, the sigma delta data stream is then reduced to a
more manageable size (decimated) and sent through a low pass filter to
reduce the quantization "noise."  The result is the PCM data.

To play back a PCM file, the reverse process is used.  The PCM data is
interpolated to add back an estimate of the original sigma delta data
stream.  The resulting data is sent through a modulator that converts
it to an an analog voltage.  This voltage is amplified and sent to the
speakers for playback.

Some people asked, why not just save the sigma delta data stream,
skipping the decimation, filtering and interpolation used in PCM?
This could improve the accuracy of the recording process.

That's essentially what Sony and Philips do in the recording process
called Direct Stream Digital (DSD).  DSD recordings are distributed on
Super Audio CDs (SACDs, a special kind of DVD).

Until recently, there have been no field recorders that use the
one-bit sigma delta format.  This changed recently when Korg
introduced two machines:  the MR-1 and MR-1000.

The Korg MR-1 ($700) is an iPod-size recorder with a 20GB hard disc.
It records one bit data at 2.8MHz through 3.5mm inputs.  Read more here:
http://www.korg.com/gear/info.asp?a_prod_no=3DMR1

Group participant Gianni Pavan purchased an early MR-1, and posted his
initial impressions here:
http://tech.groups.yahoo.com/group/naturerecordists/message/26191

The Korg MR-1000 ($1400) is more of a professional audio recordist's
tool, with XLR inputs, phantom power, and a bigger battery.  The
MR-1000 records at 5.6Mhz to a 40GB disc.  Given the high data
acquisition rate, 40GB holds about 6 hours of stereo recordings.  Read
about the MR-1000 here:
http://www.korg.com/gear/info.asp?a_prod_no=3DMR1000

Initial reactions to recording with the MR-1000 are cautiously
positive.  The Tapers Section forum has an active discussion among
MR-1000 owners.  See
http://taperssection.com/index.php?topic=3D81877
(registration may be required to read this forum)

Both Korg recorders include PC software that can convert the one-bit
sigma delta file to a variety of PCM forms.  This is one of the
strengths of sigma delta recording.  You can convert to any existing
audio format, from 64 kbit MP3 to 24/192 PCM, with equal success.  In
the future, when new formats emerge, you'll be able to translate your
sigma delta file into that format using new software without loss of
resolution.  That's why Korg calls this format "future proof."

OK, I acknowledge that at this point sigma delta recording is a
"bleeding edge" technology for field use.  But I think it's definitely
worth tracking this new development to see how it might improve the
quality of recordings made in the field.

And it's by no means assured that sigma delta files "sound better"
than PCM.  24/96 and 24/192 PCM files seem to test out as well as
Message: 2.
Subject: 8MHz sigma delta (hence the increase to 5.6MHz data rate in the
MR-1000).

The Korg recorders can also be set to record in PCM mode. They're
worth a look as conventional PCM recorders, too.

--oryoki





<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU