Re: [ARSCLIST] New CD Report NIST

[Jerome Hartke <jhartke@xxxxxxxxxxxx>]

Link is http://nvl.nist.gov/pub/nistpubs/jres/109/5/j95sla.pdf

Lots of effort, but the paper has scientific weaknesses.

1. Recordable information layers can be irreversibly modified either
chemically (cyanine) or physically (phthalocyanine).

2. Disc end-of-life occurs when test results exceed any specification of the
relevant standard, even mechanical requirements such as tilt or deviation.
The occurrence of uncorrectable errors is unpredictable, since some drives
are sensitive to certain parameters while others are quite tolerant.

3. BLER is a poor quality indicator for CD discs as discussed in detail at
http://www.mscience.com/longev.html  (NIST has been advised of this paper,
but has chosen to ignore it). Error rates such as E22, or even E12, are
better indicators. Likewise, DVD PI Errors (PIE) are poor DVD indicators,
while PI Fails (PIF) are much better (these are a specific DVD+R quality
requirement).

4. Standards specify a non-condensing humidity environment, while the
profile used by NIST has a significant risk of condensation. Media Sciences
has developed a proprietary profile that avoids this risk.

5. NIST incorrectly states that "Samples containing phthalocyanine performed
better than other dye types." In fact, only one cyanine brand was tested by
NIST, while four phthalocyanine and two azo types were tested. Some
phthalocyanine discs tested better than the cyanine sample while others were
worse.

Lots of data, but avoid taking it all at face value.

Jerry Hartke
Media Sciences, Inc.

> -----Original Message-----
> From: Association for Recorded Sound Discussion List
> [mailto:ARSCLIST@xxxxxxx] On Behalf Of Steven Smolian
> Sent: Friday, December 10, 2004 3:58 PM
> To: ARSCLIST@xxxxxxxxxxxx
> Subject: [ARSCLIST] New CD Report NIST
>
> There is a new NIST report which is a product of an NIST CD and DVD
> recorable disc test in collaboration with the Library of Congress.  It
> discusses dyes, reflective surfaces and variations in storage conditions.
>
> Conclusion: The gold/silver CD discs using phthalocyanine dyes were
> consistantly more stable than the others.  Though not mentioned by name,
> the Mistui Gold was the only such disc on the US market during this study,
> though Tayo-Yuden was phasing out its own such product. "It is
> demonstrated here that CD-R and DVD-R media can be bery stable... these
> media types will ensure data is availble for several tends of years and
> therefore may be suitable for archival use.
>
> The 8 page study appears in the Journal of Research of the National
> Institute of Standards and Technologt, Vol. 109 no. 5, Sept-Oct, 2004.
>
> I found this on the web this AM but was unable to locate it (and its URL)
> this afternoon.  Prehaps it's been reposted by now.
>
> One weakness in the study is that it ignores the issues of disc capacity
> and, particulary, writing speed.   Better we have this guidance now,
> however, than wait for a more comprehensive report.
>
> Steven Smolian