[rec.video] DAT/SCMS experiments

clements@bbn.com (Bob Clements) (11/21/90)

Here are a couple hundred lines about the SCMS copy protection
mechanism in DAT decks.  If uninterested, quit now.
/Rcc

	DAT/SCMS experiment report -- Bob Clements -- 20 Nov 1990
	Copyright 1990 by Robert C Clements, Lexington MA 02173.
	All rights reserved.

	This report may be copied for individual, non-commercial
	use only.  It may not be published or sold in any form
	for which a fee is charged, electronically or otherwise,
	without explicit handwritten permission of the author.
	This copyright notice must be retained with any and all
	copies of this report.

I've been wanting to learn more about the exact details of the SCMS
copy protection scheme in the current DAT decks for some time.  I
finally got together enough stuff to do some experiments and am
reporting my results here.

Since this deals with one of the great moral/legal issues of the
day, I had better be explicit in my disclaimer first.

	DISCLAIMER:  The information in this report is my own opinion
	based on some experiments.  It is intended only to inform
	myself and others about the present state of affairs and to
	help me decide whether I want to buy a DAT deck.  It is not
	intended to encourage violation of copyright laws.  It was not
	supported by my employer.  I will not make any unprotected
	tapes for anyone.  I do not own any of the DAT equipment
	described and I have returned the units to their owners.

I hope that's clear enough.

The equipment used in the tests:

	Two Sony DTC-700 Consumer DAT decks
	Philips CD-680 CD player with digital output
	Sony PCM-601ES PCM-audio-VCR converter (Modified)
	Ancient RCA VHS VCR for use with PCM-601

The Sony DTC-700 DAT deck (henceforth just "DAT deck") is the current
non-ES consumer product.  It is not a "professional" deck and
therefore it has the infamous "SCMS" (Serial Copy Management System).
It operates at 48.0, 44.1 and 32.0 KHz sampling rates, with some
restrictions.  The SCMS system allows digital copying for only one
generation away from a protected source such as a Compact Disc.  CDs
are sampled at 44.1 KHz.

The Sony PCM-601ES converter (henceforth just "601") is the unit which
was commonly used about five years ago to produce digital recordings
on videotape, before the advent of the DAT.  It operates only at one
sampling rate, namely 44.1 KHz.  It has a single bit of copy
protection.  A digital signal is either protected or it is not.
Signals applied to its analog inputs are NOT protected.  Signals
applied to the digital input may or may not be protected, according to
a bit in the datastream.  The protection bit is contained in both the
video format (for recording/playing on a VCR) and in the digital I/O
format.  The 601 will not make a video or digital output from a
protected video or digital source.  This protection mechanism is
simpler than SCMS and predates it.

As has been reported before, a perusal of the schematic diagram for
the 601 shows that a simple modification will bypass the copy
protection mechanism.  Thus a 601 can also be used to make digital
copies of CDs, though not in its original unmodified state.

The purpose of the tests reported here is to examine the interactions
between these systems in order to learn more about the SCMS
implementation.

Some general comments:

I had access to the service manual for the 601, and permission to
modify it.  I did NOT have the service manual for the DAT decks or
permission to open them up, let alone modify them.

This DAT deck records from its analog inputs at either 48 or 32 KHz
sampling rates.  It does NOT record at 44.1 KHz from an analog input.
This greatly restricted the number of combinations to be tested, since
the 601 operates ONLY at 44.1 KHz.  In particular, there was no way to
test the effect of the modified 601 on a 48 or 32 KHz data stream.
One could reasonably assume the mechanism in the DAT deck is the same
as at 44.1 KHz, but I couldn't test it.

This also answered the top-level question of whether I wanted to buy
one of these decks.  Since I can't do analog recordings on the DTC-700
and make copies onto the 601 video system that I have used for years,
the DTC-700 isn't fully useful to me.  I'll have to wait for one that
can record analog inputs at 44.1 KHz.

The DAT deck does not read track numbering (or even track boundary)
information from the digital input.  It appears to, but it is actually
just noticing the silences between tracks.  This was verified by using
the Beatles "Abbey Road" CD which has a long medley with many
separately numbered tracks.

Some people have suggested that the copy protection is applied to any
digital signal that is sampled at 44.1 KHz.  This turns out to be
false.  See below.

Specific experiments:

First I verified that an SCMS system actually exists, i.e., that
it isn't all just a psychological ploy.

The equipment was first connected as follows:

     ------         ------         ------
    |      |       |      |       |      |
    |  CD  | ----> | DAT1 | ----> | DAT2 |
    |      |       |      |       |      |
     ------         ------         ------
 	     coax	    coax
	    44.1 KHz	   44.1 KHz


With both DATs in record, the signal from the CD player is passed
directly through the first DAT and both units make a first generation
digital recording.

Playing back such a first generation recording on DAT1 and trying to
record it on DAT2 causes DAT2 to display "PROH" and to refuse to
record.  DAT2 goes into "Rec Pause" mode.


Next I checked out copying via the analog inputs.

The equipment was connected as follows:


     ------         ------         ------
    |      |       |      |       |      |
    |  CD  | ----> | DAT1 | ----> | DAT2 |
    |      | ----> |      |       |      |
     ------         ------         ------
 	    analog	    coax
	    stereo  	   48.0 KHz



Note that there are only line level inputs, not microphone inputs.
Also, of course, the DAT has no way of knowing whether the analog
source is copyrighted or not.  As we feared, it assumes that all
analog sources are copyrighted.  The results were:

A first generation digital recording was made on DAT1.  That recording
was then played on DAT1 and a second generation recording was made on
DAT2.  This was allowed.  The second generation recording was played
on DAT1 and a third generation recording was attempted on DAT2.  This
was NOT allowed, resulting in the "PROH" display as described above.
So you get one more generation of copying, but only one, when
recording from analog sources than when recording digitally from a CD.

All of the above pretty much agrees with the early speculations.

Experiments with a modified 601:

A modified 601 was used to create a digital videotape with the copy
protection bit off.  This datastream was then fed to the DAT decks.
The hookup was:

     ------         ------         ------         ------
    |      |       |      |       |      |       |      |
    |  CD  | ----> | 601  | ----> | DAT1 | ----> | DAT2 |
    |      |       |      |       |      |       |      |
     ------         ------         ------         ------
 	     coax	    coax           coax
	    44.1 KHz	   44.1 KHz       44.1 KHz



Counting the videotape as a first generation tape, the second
generation digital tape was made on DAT1.  This was played back and a
third generation tape was made on DAT2.  The third generation tape was
played on DAT1 and a fourth generation tape was made on DAT2.  And a
fifth generation was made.  All these were allowed.  Note that the 601
was used only once and that thereafter many generations were made
using the DAT decks alone, all at 44.1 KHz.

So it is NOT true that "any 44.1 KHz digital stream gets protected by
the DAT deck".


Next, I checked to see whether a first generation DAT tape was still
copyable by the modified 601.  This verified that the SCMS copy
protection format wasn't totally strange.  The setup was:

   ------         ------         ------         ------         ------ 
  |      |       |      |       |      |       |      |       |      |
  |  CD  | ----> | DAT1 | ----> | 601  | ----> | DAT1 | ----> | DAT2 |
  |      |       |      |       |      |       |      |       |      |
   ------         ------         ------         ------         ------ 
 	     coax	    coax           coax         coax
	    44.1 KHz	   44.1 KHz       44.1 KHz     44.1 KHz


This test also worked, allowing multiple generations after a single
pass through the modified 601.

Experiments with an unmodified 601:

I then removed the modification from the 601, restoring it to its
normal single-bit protection scheme.  I tested tapes from the previous
experiments to see which could be copied.  The results were:

	CD player			Protected
	First generation DAT		Protected
	n-th order unprotected DAT	Copied OK


Further conclusions:

There was no way with this equipment to create a completely
unprotected DAT at 48 or 32 KHz.  In fact, without a modified 601 or
some other protection stripper, a consumer cannot create an
unprotected DAT tape at all with the DTC-700, even if the audio source
material is his own performance.  However, once an unprotected tape
has been created by some other means then it can be copied to any
depth by a pair of DTC-700s.

I will be interested to hear from anyone who has examined the DTC-700
service manual and/or come up with more information about the
possibility of getting it to make unprotected tapes from public domain
audio sources.

Bob Clements, K1BC, clements@bbn.com
Affiliation for mailing address only.  My employer was not
involved in these experiments in any way.