At 9:33 AM 10/2/96, Gerard Arkenbout wrote:

> Robbie,
>
> Your email to Jack (John C. Kane) mentions an experiment in the
> future with 2 dimension imagery, to get a precision better than
> 0.003-inch or < 0.1 mm near the edge of the disc. This is a great
> challenge, for how is that to be done for a 20-inch diameter
> metal-disc for the CHORDEPHON-60 music-box. We prefer the DIAZO
> sunlight-amonium hydroxide-vapor method, for the paper stays dry, so
> the exposure is the right image of the disc. With a metal puncher a
> copy of the disc, with the aid of this print, is a very good one. Also
> the exchange of music titles will be an easy one. Because the DIAZO
> copy can be send by letter (not electronically) to the party
> interested. While a flatbed-scan may be transferred to paper, the
> printer will change the dimensions of the image by heating or wetting
> the paper with the exposure. Therefore another electronic transfer
> method must be applicated. (may be a mathematical correction
> during printing, for the dimensions to be achieved are known)
>
> With regards   Gerard Arkenbout from the Netherlands

Dear Gerard,

The end-result of my experiment is a computer file which contains all
of the information necessary to punch an exact duplicate.  Therefore
the computer file may be considered the "Master File" for that
song-arrangement.  Furthermore, the music may easily be modified for
smaller, or larger instruments.  For example, the computer file for
punching a large music box arrangement could become -- with a bit of
modification -- the master file for a new Chordephon Zither disc.

I admit that it will not be possible to attain 0.1 mm resolution with a
20-inch disc -- it will be difficult to achieve with the 9-inch Stella
disc!

While a small disc image may be made with photographic paper, I agree
that the best method for larger discs is the Diazo/ammonia method.  It
will be necessary to cut the paper image into narrow segments, in order
to fit in an 8.5-inch flatbed scanner.  Then the computer files must be
"stitched together"; fortunately standard software programs are
availble for this task.

You are correct: mathematical corrections may need be applied in the
case of elliptical distortion.

Best regards,

Robbie Rhodes

• ----------------------------------
|         Robbie Rhodes            |
| Return-Path: rrhodes@foxtail.com |
 ----------------------------------