[AV Media Matters] The Functions and Science of 'CD Polishing'.

[e.ha.zwaneveld@nfb.ca]

Greetings,

1. The problem: Compact Disc media in all their recorded and replicated
forms, are subject to reading surface damage, particularly scratches, gouges
and pits. Disc reading surface damage is a serious nuisance to owners of CD
or DVD  recordings. It is especially a problem when such discs are rented or
loaned as with libraries and in 'dynamic manual archives' when the original
recordings become degraded. The focused laser beam used to read back the
pit-shaped signals below the polycarbonate disc surface is deflected as a
result of scratches, gouges and pits and may be unable to read back the
information. Although it is possible that cleaning of such abrasions could
help remove any accumulated particulate deposits, such cleaning does not
resolve the essential problem which is to cause the reason for light
scattering to be taken away. Another point about the harm done by water
cleaning of CD media is not that it harms the polycarbonate support. But the
unintended penetration of water through the porous polycarbonate and
protective lacquer layer is then able to attack and cause the recording/
reflective layer to be oxidized and lose its ability to hold the signal or
to reflect it to enable playback. The standards of successful treatment to
recover music from a CD is not the same as that of a disc used to preserve
valuable heritage content. That is why the error correction circuitry in the
playback systems has wider tolerances for Audio CD then the for other
applications of optical disc media. The issue is not whether the repair
resolves the problem sufficiently for the eye or ear to consider it
subjectively acceptable. The issue is whether a focused laser beam can do
its job without impairments, dictated and specified for that application.
Potential users of any process may not wish to test some 'solutions'. The
information they have may indicate these 'solutions' to be inadequate.
Although references are made in passing to chemical know-how invested in
these products, we have not yet read anything that explains which scientific
principles are used, not an encouragement to explore such products! The
vendor may also demonstrate disregard for the problems the potential
customer is trying to solve. Just sending samples is no longer good enough!
A little while back I received a film cleaning solvent 'solution' from a
supplier, but without chemical specifications that would enable us to
consider the science behind it or to know whether its use would damage film.
We took it to a chemical waste recycling plant, without even applying it to
any film. A vendor who belittles the use of objective measurements because
the user's objectives are more scientific than most of his customers, has
not understood that he is in the problem-solving business and not in the
business of selling snake-oil. Perhaps we can raise the level of our
discussion by looking at the science of the problem:

2. The scattering of light: J.W. Rayleigh and Lord Strutt, in Phil. Mag. (4)
41 in 1871 on pp. 107, 224 and 447, first demonstrated that the scattering
of light is caused by optical inhomogeneities. The molecules at a damaged
surface act as secondary sources of light, capable of preventing all of the
focused light source to get to the reading surface and by missing its
targets, fail to read the recorded or embedded signal.

3. Light refraction or light-ray-bending power of any light transmitting
medium: The first basic law of optical refraction was formulated by Snellius
(1618) and also independently by Descartes (1637). It required another two
centuries before a correlation was established between the refractive index
and the chemical structure. Gladstone and Dale (1858) found that for organic
liquids, the ratio (n-1)/p, if measured for a standard wavelength, is,
(nearly independent of the temperature), a characteristic "constant" of the
substance considered. This was named "specific refraction". Therefore, the
ratio between the sine of the angle at which the light approaches the medium
(the incident angle) and that which passes through the medium (the angle of
refraction), both angles measured with respect to the normal or optical axis
at the point of entry. It may be written thus: refractive index = sine i :
sine r.

4. The refractive index for polycarbonate (poly[1,1,-ethane bis (4-phenyl)]
carbonate) is 1.585. If we wish to neutralize the undesirable light
scattering behavior of the damaged areas (particularly the 'crater or cliff
edges'), we could then apply a liquid, and if it has a matching refractive
index, the scratch, gouge or pit will become invisible. It now remains to
identify a liquid that can be applied to a damaged CD reading surface, that
has in its dried state the matching refractive index. From the reports we
have, it would appear that  the existing solutions that have been formulated
have their limits.

5. Examples of known similar uses in our industry: Film base scratches are
unacceptable because they are enlarged and annoying when viewed in
projection. When they occur in negative film base they turn white in
projection and are far more annoying then when black as will occur with
print film base scratches. So the traditional approach was to rejuvenate or
'polish' the film base. Using highly polished glass rolls and chemical film
base softeners, the film was wrapped in its soft state around such glass
wheels. If the proper degree of softening was created, any scratches would
be flattened and would disappear again. If they were very bad, the surface
was first roughened by wrapping the film in its softened state around a
glass wheel with a matte surface, thereafter around a highly polished glass
wheel. Another common approach was to regenerate the scratched film surface,
not by grinding, but by coating it with a proprietary 'film coating',
consisting of some form of lacquer. The disadvantage with such lacquer
applications was that sooner or later the film needed to be cleaned, usually
in a solvent, and it degraded the lacquered surface, requiring ultimately
its removal. That was a very laborious and seldom satisfactory process. I
have managed motion picture laboratories where we were not only equipped to
polish film base and to rejuvenate the surfaces. If a customer had been
persuaded to have some outfit apply lacquer to their original negative,
there was little we could do to save their valuable film content. This might
explain my impatience with unverifiable formulas and results that are not
measured! The refractive index for triacetate film base is 1.478, but the
refractive index of gelatine in the emulsion is 1.520. Therefore it is not
possible to exactly match both requirements in a single solution. But there
are several solutions or mixtures thereof that offer a reasonable
compromise. Percholoroethylene has a refractive index of 1.504 and it does
not contain water. Thus it has been applied in immersion and application wet
printing and even projection, to render the scratches invisible when the
(base) scratched film is being printed or projected, as long as the surface
remains wet during the reading. This wet state content transfer in printing
or telecine transfer has become the most desirable solution for dealing with
scratched film base surfaces.

6. Other 'solutions'` applied to film which also must maintain an optically
pure light transmission surface have included various lacquers, not unlike
the so-called 'protective lacquer layer' that is routinely applied to the
label side of optical CD media. In view of the above, it should be obvious
that the lacquer approach for film scratch concealment is hardly used today.
But we do not have the luxury with optical discs to play them back in a wet
state, we cannot diffuse the light source and polycarbonate is not the same
as cellulose triacetate film base material.

7. Coming back to the question of scratched, gouged or pitted polycarbonate
CD or DVD disc media, we note from the previous discussion that there are
several approaches. The most radical one is to use a more or less abrasive,
over a variable period of time. This method essentially grinds away the
majority of the perfectly good and smooth optical disc surface surrounding
the scratches, gouges and pits, until its surface has reached the thickness
that equals the depth of the damage. Whether this disc surface
'rejuvenation' happens while the polycarbonate is in a softened state is
doubtful, nor is it advisable because softening would disturb the sandwich
structure of the bonded disc surfaces. It should be kept in mind though that
the thickness of the polycarbonate substrate should not be made less than
the specification tolerances provided. Therefore the sharp cliff or crater
edges, if it is done well, become smooth again and perhaps arriving at a
disc surface that will be optically true enough not to deflect the laser
beam anymore. The abrasive used for the tests at the Canadian Conservation
Institute was apparently a 'light abrasive'. Remarks made earlier about the
experience factor are probably also quite relevant. Another approach, not
applicable in the case of optical disc media, is to increase light source
diffusion, it has the effect of making the hard edges of scratches softer
and therefore less visible. Another approach that has also been used with
film base materials, was to apply clear shoe polish or wax to the scratches
and hopefully also smoothen the sharp edges! The other solution apparently
used by the gentleman who cannot stand insistent and admittedly rather
impatient questions, is based on known techniques for the elimination of
scratches on optically critical surfaces in which if I understand his
argument correctly the 'cleaning' of the scratches is emphasised.

8. Several times the argument has been used that to measure the improvement
of the treatments is too expensive, assuming that the user would have to buy
such testing equipment. Understandably, many users would find the cost and
technical complexity of such testing methods prohibitively expensive. That
is not the only solution though.

9. In all cases, users who have damaged discs that are causing drop-outs,
are advised to use the services of commercial testing services to benchmark
and do a quality evaluation of their disc before choosing the 'polishing' or
any other 'rejuvenation' method of their choice. Thus they will have a
passport of the conditions before treatment. Then, after treatment, send it
back for another test report. One example of such a professional service is
Media Sciences, Inc. (Dr. Jerome Hartke) in Marlborough, MA, USA. (See
http://www.mscience.com or jhartke@mscience.com) This company is equipped
with a wide variety of test equipment and measured many relevant criteria.
If I am not mistaken, VidiPax (Jim Lindner)in NY, NY also provides some CD
testing services (See: http://www.panix.com or vidipax@panix.com/~vidipax).

10. And if you want to be really nice, share the results with the
manufacturers of these disc rejuvenation and cleaning products. Because of
their market focus, the present discussion partners assume that all
customers (and potential customers if they show some sensitivity to their
applications) are supposed to evaluate their work subjectively. Many of us
who have extended-term life expectancy of content in mind, do not subscribe
to that approach, because neither the eye nor the ear is a reliable
measuring instrument. In addition, please do not forget to share your
results with us on this listserv. Most archivists are not housed in ivory
towers but in basements. Their collections suffer neglect not because they
do not care, but because the agenda of many media and recording system
manufacturers is focused on the short-term playback of the content. Those
priorities dictate serious investments in copying content, usually when the
storage medium is in its worst condition. Sharing measured and repeatable
quality measurements will facilitate our ability to save our audio-visual
works. Many discs suffer abuse. Because the people whose services or
products are focused on consumers, have not provided us with reliable
information to go by that will enable us to predict whether a solution will
solve a particular problem or not. Why we set our standards so high, is that
we somehow will have to establish when damage approaches a level where we
cannot without loss of content integrity copy the recording anymore. Because
optical disc playback technology does not enable error rate indication, we
cannot determine without measurement when the critical point approaches. If
there is any conflict at all, it is between a short-term solution focus vs a
long-term content preservation focus.

Thank you for an enlightening discussion,

Ed H. Zwaneveld,
Technological Research and Development
National Film Board of Canada,
and Chair AMIA Preservation Committee
August 14, 2000