Re: [ARSCLIST] Is recording to Reel-to-reel still the preferred preservation method?

[Claus Trelby <ctrelby@xxxxxxxxxxxxxxx>]

This is a paper I had not read before. I am not one to get into technical
details... I fundamentally believe that the ears have to do most of the
work. Ultimately they have to decide.

That being said... Dan makes all the right points. PCM technology has to (by
definition) be filtered above (half) the frequency range that the sampling
rate allows to reproduce. In his paper you see this along all the "mirror"
images across frequencies. I don't believe that is the conclusion to this
subject.

I have a very good relationship with the guys at Sony, and I am not sure if
the people reading this knows that DSD originally was conceived as an
archival format. It was realized that the archival business was not "ripe
for the picking" (financially) and the DSD format turned into SACD. The
theory of 1 bit and Gig's of sampling makes sense due to the simple fact you
don't have "ultra" frequency and other PCM artifacts that have to be
filtered...

In my opinion a logical short analysis would be... The "stuff" filtered at
192HHz PCM is beyond the human hearing range (filtered at 96K - our hearing
at 20K). From experience I know that early equipment and software "forgot"
to filter what had to be gone... This let to data files (and tests) too
large for the allowed space allotted for DVD-A essentially ending up to be
too large to fit on a disk. DSD does not have this problem. It moves several
Gig samples in time and moves one bit (dynamic range) up or down. This will
by definition not create artifacts. Technically and by definition DSD should
be the format of choice.

I support it, but if the dialog turns into what hardware will be available
in the future, Sony has a lot of work to do. I know they are in the process
of generating hardware that will suit production beyond direct recording,
but do we believe that the better "BETA" format will be around? Sony and
Phillips it trying to make it so... I hope they succeed, but in the mean
time I get mostly requests for PCM digital formats. In professional video
"BETA" is king so don't rule anything out...

To justify PCM I have to say this. . . (99% of all production hardware and
software is based on this technology) Early pioneers in CD had to fight for
the "cleanliness" of the sound, lack of "ambience" etc, etc... No matter
what you filter out at 96Khz (audible) at 192Khz (digitally) you'll never
hear it unless your hardware or software is not functioning correctly... I
might have "a complete lack of signal fundamentals", but double the sampling
rate makes for double the amount of "snapshots" in-time in the audible
hearing range of sound, no matter what is heard and available in the source
material. I know speakers have limitations, and so forth... but I think we
are in the business of NOT being able to predict the future.

Going back to my bit depth (PCM) issue I truly believe that unless you use
laboratory microphones directly to digital (DSD or PCM)... any frequency
stored on tape will delude (due to mechanical, tape restrictions and
electrical noise-floor) above 32 bit but always be noticeablly better if
captured at maximum level (0 TO -1 dB digital peaks).

Again... When is the limit reached for the human ear to hear the difference?
AtoA will inherently produce two levels of analog generation loss. AtoD will
deliver one level of analog loss and one level of digital "artifacts"... no
matter how we play it back and listen to it. That will always be the case
and ultimately our ears have to do the judging. Bottom line... We will
always be slaves to our hardware and software, no matter if we test for
artifacts or for human hearing.

Sleep tight,

Claus.

Claus Trelby
Managing Engineer/Partner

XEPA Digital
1137 Branchton Road, 19-N-3
Boyers, PA 16020-0137
www.xepadigital.com
ctrelby@xxxxxxxxxxxxxxx
P:724-794-3686
F:724-794-3292
C:805-490-1730





-----Original Message-----
From: Association for Recorded Sound Discussion List
[mailto:ARSCLIST@xxxxxxx]On Behalf Of Parker Dinkins
Sent: Wednesday, September 29, 2004 10:07 PM
To: ARSCLIST@xxxxxxxxxxxx
Subject: Re: [ARSCLIST] Is recording to Reel-to-reel still the preferred
preservation method?


At least one study has shown that using a sample rate above 96kHz provides
no benefit when used with well designed filtering, and in fact using 192 kHz
results in reduced accuracy as well as unnecessary additional expense, both
in storage and equipment requirements. See Dan Lavry's white paper on this
topic at

http://lavryengineering.com/documents/Sampling_Theory.pdf

---
Parker Dinkins
MasterDigital Corporation
CD Mastering + Audio Restoration
http://masterdigital.com



on 9/29/04 8:06 PM US/Central, David Lewiston at david.lewiston@xxxxxxxxxxx
wrote:

> Upon the advice of a highly-regarded "golden ear." In his opinion, if the
> very best A/D converters are used, 176.4kHz and 192kHz are audibly
superior
> to the lower sampling rates. I don't have access to a broad enough
selection
> of equipment to make my own judgements, so I rely on him.
>
> He has had a chance to listen to nearly all of the high-end A/D
converters,
> and recommends the Pacific Microsonics box. This may not be available for
my
> project, so instead I've specified the new Weiss box.

---
Parker Dinkins
MasterDigital Corporation
CD Mastering + Audio Restoration
http://masterdigital.com

---
Incoming mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.769 / Virus Database: 516 - Release Date: 9/24/2004

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.769 / Virus Database: 516 - Release Date: 9/24/2004