John Chambers wrote:
>The goal that I was thinking about is: Suppose I've laboriously typed
>in the ABC for Mozart's clarinet quintet in A. It has header lines
>defining the voices and keys:
>
> V:1 name="Clarinet in A" clef=treble
> V:2 name="Violin I" clef=treble
> ...
> V:1
> K:C
> ...
> V:2
> K:A
>
>This comes out on screen and paper just right, with the clarinet part
>in C and the string parts in A. So I feed it to an ABC player, and I
>get cacophony. I need to explain to it that voice 1 sounds a minor
>third higher than the default A=440. Saying C=440 is so obvious that
>it seems hardly worth mentioning. But suppose the notation uses
>something like "transpose=+3". If the past is any guide, what I'd
>expect to happen is that this would make it sound fine, but the
>notation I see on screen or paper would also be transposed up a minor
>third, for an Eb clarinet. Not quite what I had in mind!
That's why I used "transposePlay=" rather than just "transpose=". Actually,
I wanted to make the word "playTransposed=", but I also wanted to make it
possible to abbreviate that to "T=" to stay compatible with MUSE.
>What is needed is an unambiguous way of saying "Show this exactly as
>I wrote it, but play V:1 a minor third higher than written." The
>C=440 notation seems to say this, for the reason you said: It looks
>like a tuning rather than a transposition. This is exactly what I
>want. Programmers working on formatter code would be more likely to
>ignore this, because frequency isn't relevant to printed music. But
>they aren't likely to ignore something called "transpose", because
>this is a constant headache for people working on printed music, and
>they'd expect to have to implement it.
>
>One problem is that "transpose" means very different things to
>different people. It would be best if we could avoid it in most
>circumstances, and use notation that unambiguously says what we want.
As far as I'm concerned "transposeplay= +3" means quite obviously that the
music is to be played three semitones higher than written. "C = 440"
is much less obvious. It could even be a tempo setting. If it's a pitch
setting I have to think seriously whether it means three semitones up
or nine semitones down. When I'm playing music I don't think in terms
of absolute numerical pitches.
>| ... Not that a tuning command
>| isn't useful under some circumstances. (BarFly already has one:
>| you can put tune=n in the V: header field where n is in units of
>| 1/256 semitones. It's useful to slightly de-tune intruments which
>| are playing in unison so you can hear them as separate instruments.)
>
>But it's not useful to me with current technology. I have several
>electronic tuners. If I pick up that A clarinet and play a C, I know
>that it will tell me I'm playing and A. It may also tell me that I'm
>playing at 443. Very few musicians would have any clue about how to
>convert this to units of 1/256 semitones. It's true that the tuners
>have dials that are obviously calibrated in some fraction of a tone,
>but they never have a scale, so you don't know what to type in. All
>you can do is guess, listen, and guess again until it's close enough.
>
>But if the player program implements C=443, I can play one note to my
>tuner, and immediately type in the tuning for the A clarinet part. I
>can add A=443 to the other parts, and everything will sound fine.
>
>(Well, OK; I'd probably type A=435 for the viola part. ;-)
>
Well, tune = n was a quick kludge to fix the problem caused by Quicktime
playing so precisely in tune that instruments playing in unison merge
together. I'm not proposing it as any kind of standard. In general,
I think a tuning command should be a local option, rather than written
into the abc, and probably best implemented via a slider control so you
can tune the program just as you would tune any other instrument. A
numerical readout would be fine, but really I want to use my ears.
I don't possess an electronic tuner, although I did once use a digital
frequency meter to tune a concertina down to concert pitch.
Phil Taylor
To subscribe/unsubscribe, point your browser to: http://www.tullochgorm.com/lists.html
