Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site oakhill.UUCP
Path: utzoo!watmath!clyde!bonnie!akgua!sdcsvax!dcdwest!ittvax!decvax!genrad!mit-eddie!godot!harvard!seismo!ut-sally!oakhill!don
From: don@oakhill.UUCP (Don Weiss)
Newsgroups: net.audio
Subject: More on PCM tape dubbing
Message-ID: <287@oakhill.UUCP>
Date: Fri, 28-Dec-84 17:01:01 EST
Article-I.D.: oakhill.287
Posted: Fri Dec 28 17:01:01 1984
Date-Received: Mon, 31-Dec-84 01:57:12 EST
References:  <260@oakhill.UUCP> 
Reply-To: don@oakhill.UUCP (Don Weiss)
Organization: Motorola Inc. Austin, Tx
Lines: 64
Keywords: PCM CD digital dbx
Summary: 

>> The best way I have found to dub CDs is actually onto open reel, 1/4 inch,
>> half-track tape at 15 ips with DBX type 1.
>	...
>> Said method actually provides better noise performance than dubbing to
>> digital VCR tape, using either a Technics SV-100 (EIAJ 14-bit) or Sony
>> PCM-701ES (using 16-bit format).  
        ...
>> in fact if not for the cost, I'd probably do all my dubbing onto analog in
>> the manner mentioned above.

>Hmmm... I don't hear any noise when I dup CD's with my Nak DMP-100 (16-bit).
>Maybe I am not listening well enough or the tweaking that Nak requires is worth
>it (the Nak is a Sony that meets Nak's tolerance specs).  Any idea where the
>noise comes from?  Were those CD's digitally mastered?
>	--Cliff [Matthews]


Let me give some detail on the experiments that led to the conclusion about
the relative noise performance of the two methods described above.

Frst, I digitally dubbed the new digitally-mastered Columbia recording by
pianist Cecile Licad of Rachmaninov's second Piano Concerto, with Claudio
Abbado conducting.  I used a Technics SLP-7 CD player, a Technics SV-100
PCM box, and a Quasar W5310 VCR (I guess that last is irrelevant!).

I decided to critically A-B the digital dub with the original CD.  They were
quite indistinguishable exept in the intervals of digital zero between cuts
and also in some of the very quietest recorded passages...for instance,
at the end of the slow movement, with some very pianissimo piano keystrokes,
the digital dub displayed audibly greater background hiss than the straight
CD playback.  Being startled at this phenomenon, I repeated it with a very
careful adjustment of recording level (to the absolute maximum possible on
crescendos) and balancing of playback levels between the tape and CD; the
result was the same.

I then decided to fire up my trusty Technics RS1500US open reel tape machine
with its outboard dbx model 150 (type I) noise reducer.  I ran the 1/4"
half-track machine at 15 ips just for grins.  The result of the same "live-vs.
memorex" test was that there was absolutely NO noise buildup audible under any
signal condition.

Later, I borrowed a friend's Sony PCM-701ES PCM box and reran the experiment
with it, in 14 and 16 bit modes.  The result was essentially the same as with
my Technics box.

The source of the noise was suggested by the experiment of turning the record
volume on the PCM box all the way down, beginning a (blank) recording with
record muted, and then recording some more with record unmuted.  There were
substantial increases in the playback noise level between (a) playback muted,
(b) playback unmuted and record muted, and (c) playback and record unmuted
with record level set at zero.  It would seem, therefore, that the analog
circuits preceding the A->D conversion are in need of improvement; I wouldn't
be surprised if your (Sony --->)Nakamichi tweaks include this area.

I ran one final experiment:  I set up a 1 KHz oscillator with a wide range
step attenuator on its output so that it produced maximum record level (as
verified on a video monitor).  I then ran a test recording, starting with
the maximum level and reducing it in 10 dB steps to -80 dB of its original
value.  On playback, the tone seemed to sink below the broadband noise by the
-80 dB step.  Thus, I'm not convinced that the PCM boxes are delivering all
the S/N (i.e. 86 dB) that they're supposed to.  I'll have to find a really
sensitive AC voltmeter to find out.

-----Vive l'numerique!       Don Weiss