Thanks for your input Eric. I agree that 2D IRENE's most promising
application (in theory) is with delaminating (or broken) lacquer discs. The
groove profile problem is definitely an issue, though my tests showed that
it was a sharply V-shaped, not rough or round, groove bottom that caused it.
In these cases (about half, in my experience), audio must be derived from
the groove top, which is much noisier.
Other current issues include detecting the surface of the disc properly to
keep the camera in focus (due to reflectivity or translucence of the
lacquer), edge-detection algorithms picking up on wrong features, and the
fundamental principle of discarding the groove wall information.
I think it probably sounds like I'm griping. That's not my intention. There
are several unrelated problems preventing IRENE from reaching its potential,
but I don't think any of them are unsolvable. I think it's important that
people are aware of these issues so that they can make informed decisions
about when to use the system.
IRENE's 3D capability has been a hot topic of debate since this year's ARSC
conference, but the vast majority of my work at NEDCC was with lacquer
discs. There are huge collections of lacquer discs at risk, and there isn't
a clear public consensus about how to approach preserving them. To repeat,
there have been no third-party tests comparing IRENE's sound quality for
discs against turntable and stylus. Maybe people are assuming that it
doesn't sound as good and that's ok, but I think people are assuming the
opposite in the absence of good information. NEDCC are still accepting large
projects of intact materials, and I think it's an irresponsible waste of
time and money.
IRENE has so much potential for broken/damaged/obsolete media, but using it
when it's not the best option doesn't do anyone any good. If the audio
archives & preservation community wants to see it reach its potential for
the materials it's actually needed for, then people need to make their needs
and expectations known.