Midi, Mac, Eudora, and Quicktime Surprises
By Philippe Rouillé
Paris, May 8th, 1996
On the 96.05.07 mailing list digest, I was complaining that, as a Macintosh user, I could not extract the MIDI files which are sometimes encrypted in the e-mail text of the mailing list. I really wanted for example to see
> The MIDI file of the recreated master roll of Ampico recording
> [ 68283, Paderewski's Minuet, Op. 14, No. 1, played by Sergei
> [ Rachmaninoff.
which was part of the 96.03.05 digest of this mailing list.
It infuriated me so much, not being able to convert these files into good MIDI files, although Robbie Rhodes and Jody Kravitz did their best to patiently explain to me everything about Macintosh files, UU-encoding, Binhex, etc., that I read some literature about it. I am not at all a computer expert, but I thought I could find an explanation to my failure in the manuals of the softwares we use : Eudora (132 pages), StuffIt, Compact Pro, etc. It is in the small User's guide of UULite 2.0, a UUdecoder of files, that I found the truth, which may be useful to those other members of the list using Eudora to get their e-mail and who are not experts in computers. Page 27 says :
> If UULite seems to be having problems with Eudora ...
> Users of Eudora should confirm that they have Guess Paragraphs turned
> off. If Guess Paragraphs is left on, Eudora will change the carriage
> returns at the end of all uuencoded lines to spaces. This will cause most
> decoders, if not all (including UULite) to fail<
So I looked for Guess paragraphs ... which I found in the "Save as" command that we actually use to extract this famous e-mailed MIDI files ! So I disactivated the two options : "Guess paragraph" and "Include Headers", and I could work on these files properly. Computers experts will probably say : but all that is evident. It was not for me ...
When these files are too long to appear in one Eudora message (like Paderewski Menuet, which was divided in 2 parts), I save each as 68283-1 and 68283-2, for example, I merge the 2 in one document using the merge command of UULite, and there the miracle occurs : I see a 68283b.zip file appear, that I can unzip and convert into a MIDI file as I explained in the 96.05.07 mailing list digest.
My cheerful thanks go again to Jody and Robbie, who did their best helping me solving the problem.
Speaking about computers, should I mention a bug with Quicktime, which may blurr the rendition of some MIDI files of Orchestrions. Following the suggestion of Roger Stern (digest 96.04.15), I had a look at the web site of PianoDisc (Music Systems Research)
http://www.pianodisc.com/
and imported a MIDI file of a movement of a Mozart concerto. I opened it with Lecture Quicktime (and Quicktime instruments), and was astonished to listen first to an very very vigorous chamber ensemble of tympani, saucepans and various steel bands, and, after 1 or 2 minutes, to a fairly good grand piano - but the percussions did not disappear. Poor Mozart ... I got the explanation on a small web site, and I cannot resist giving you some extracts of the text accompanying a very small MIDI file demonstrating why you get tympani instead of a beautiful mozartean string ensemble :
> QTMIDI-Test.mid (889 bytes)
>
> Charles Kelly, March 25, 1996
> http://www.aitech.ac.jp/~ckelly/mmuig.html
> The Macintosh MIDI User's Internet Guide
>
> This is a MIDI file for testing the instrument program change accuracy of
> any MIDI software which makes use of QuickTime & the QuickTime Musical
> Instruments Extension.
>
> The general MIDI instruments are often assigned the wrong sound in
> QuickTime 2.0 & 2.1. Even Apple's SimpleText 1.3 assigns the wrong
> instrument sounds!
>
> I imagine eventually Apple will change either QuickTime or change
> SimpleText, MoviePlayer and other similar software. (When they do, the
> usefulness of this file will have expired.)
>
> This file was produced to help programmers be aware of the problem and to
> give them an easy way to check that their software is performing properly.
>
> Non-programmers may find this (small, fast-to-download) file useful in
> checking out the accuracy of software which they are using.
>
> General MIDI numbers are 1 to 128, not 0 to 127.
>
> ***There are 4 program changes in this file. (Tune name: Astley's Hornpipe)
>
> 57 Trumpet (not Orchestra Hit)
> 49 String Ensemble (not Timpani)
> 73 Piccolo substituted by FLUTE in QT (not Clarinet)
> 33 Acoustic Bass (not Guitar Harmonics substituted by DISTORTION GUITAR in QT)
>
etc. I am not reproducing the end of the text which says which softwares got the same mistakes ...
Next time, I shall try to make a contribution more about automatic musical instruments and less about computers.
Philippe Rouille
|
(Message sent Wed 8 May 1996, 19:24:10 GMT, from time zone GMT+0200.) |
|
|
Mechanical
Music Digest™ Archives
