Scanning Music Box Disks
By Wayne Stahnke
| [ Forwarded by Robbie Rhodes:|
Robbie Rhodes sent me your recent notes regarding scanning music box disks, and it piqued my interest. I have a few suggestions that might be helpful.
First, however, I hope you can clear up a point for me about nomenclature. I know that historically we have used both the spelling "disc" and "disk" in English. I believe that both spellings have been accepted as correct for some hundress of years, but that in the very recent past (since about 1980) there has been a division along functional lines. The spelling "disc" seems to be used generally, where no specificity is wanted or needed, but "disk" seems to be preferred when referring to items in a a computer. Thus,. we have the "Compact Disc" digital audio system and "disc" brakes, but a home computer user will say that his "hard disk" has a capacity of so-and-so many megabytes. Floppy disks are really "diskettes" and therefore have only one spelling. Do music boxes use "discs" or "disks" and why?
Your suggestion to scan music box disks strikes me as being timely and valuable. The technology to do such a thing is with us; this was not the case as recently as ten years ago. There is not the urgency connected with this effort that there is with music rolls, which are becoming increasingly fragile with each passing year. Nevertheless, it seems to me that scanning music box disks is worth while.
Robbie Rhodes has suggested an experiment to explore the problem. His idea is to photocopy a small, relatively coarse music box disks and scan it, and then do the translation as required. This is an excellent approach, but I think too small a step. It would be better, it seems to me, to take a larger first bite out of the problem.
I suggest that you select a disk of particular interest. This should be a disk for which there is or could be a widespread interest because of its scarcity or musical value. I suggest that you go to the trouble of making a high-quality photograph of the disk using a flat-field lens (intended for copying photographs) fitted to a 35 mm camera. I imagine that the disk would be mounted horizontally with the tangs down, and uniformly illuminated from below, to provide high contrast. A white bed sheet lit by 3 or 4 desk lamps would provide a suitably diffuse surface. The camera would be mounted on a tripod, looking straight down at the disk. After developing the film, the 35 mm negative would be scanned with very high precision by an optical shop that provides such a service. Robbie could operate on the resulting disk file and prepare it for other use. I do not see any fundamental reason why you could not use this technique to scan any number of disks at relatively low cost, and without the effort required to develop a special scanner.
The only problem that I can imagine is resolution. We would like about 10 to 20 pixels per channel. If a disk contains 200 channels (100 on either side of center), we would like 2000 to 4000 optical lines of resolution. High-quality lenses and film are good for about 100 optical lines per millimeter; this may not be sufficient when using 35 mm film. A larger format camera may be required.
I hope this note is of value. I wish you the best of luck in your efforts.
[ Editor's note:
[ It would be interesting to know if the Kodak "Photo-CD" format
[ preserves enough resolution from 35mm film for this purpose.
[ Its relatively easy to get an entire roll of film converted.
[ The CD would probably have to be shipped to the interested
[ party, as it quickly becomes a lot of data.
(Message sent Mon 5 Aug 1996, 04:06:17 GMT, from time zone GMT-0400.)