Possible MIDI tuning problems

[Joseph Pehrson, TD 475.19:]
>I'm beginning to question, however, whether I am experiencing the "full
>effect." I first tried the experiment on a sound card -- a
>Soundblaster Awe 64 -- but I fear that the bending was not being fully
>implemented. The just intonation thirds sounded, to me, awfully large
>-- and not too different from 12-tET. Is this bending information not
>always properly interpreted by sound cards (???) The three different
>versions 12-tET, 5-limit just and 7-limit just really didn't sound all
>that different...

It can happen! General MIDI specifies that the DEFAULT pitch bend range
is +/- 2 semitones (400 cents for 16384 bend units, or 40.96 bu/cent),
and that is the range I'm assuming in my retunings. But the value can
be modified in various ways on any given sound module.

Most of the way down on my web page are two tiny files you can download
to assess and try to correct the problem (the correction is a "General
MIDI reset/enable" command followed by explicit instructions to make
each channel the proper bend range).

You can download the files directly from:

http://www.idcomm.com/personal/jadl/bendRangeCk.mid
http://www.idcomm.com/personal/jadl/bendRangeSet.mid

Playing bendRangeCk.mid, you should hear E, F#, and D. If not, or on
general principle (it plays only channel 0, and other channels may have
different ranges set!), play bendRangeSet.mid (no notes sound...), then
the Ck file again. E-mail me on or off list if the problem persists.

I'm assuming you're not playing the "half retuned" versions of the first
movement of the Schubert D894 sonata.

The contents of bendRangeSet.mid should probably be at the top of all my
retuned files; I'll make a mental note to add that.

Thanks for listening!

JdL