At 09:13 PM 9/9/05 -0400, David W. Fenton wrote: >Well, the problem is that you've got 180 degrees and 128 MIDI values,
Now I see what you're getting at. Let me take you through why my preference is that scaling be abstracted and generalized. In most cases, percentage works for me because I'm a child of knobs. I built my first synth modules in 1970, and bought my first synth in the way-pre-Midi dark ages of 1973. Knobs did not (and still do not) all have the same travel. The "o'clock" method doesn't work if your knob's travel can't reach 7:00 or 5:00 ... plus knobs with "off" positions didn't actually start controlling until 8:30. Kinda like my mornings. :) And then, many knobs are sliders, and sliders vary in length even more than knobs vary in travel. None are replicable. So controls with a zero point at the start were 0%-100%. Controls with a center zero point ran -100% to +100%. Filters had a variable Q point depending on temperature. For me, it was always a percentage relative to the turn of the knob or reach of the slider. Another example of abstraction is the volume control, which is logrithmic in output, but "looks" linear (in physical knobs and in Midi). And that doesn't include so-called loudness taps, or even off-the-shelf multi-band (hardware or software) equalizers, which work in unequal divisions approximately related to logrithmic ones. Your VU meter in the physical world and in software displays in dB. It's all what "feels" right to the eye and the hand, and that does not necessarily map to the actual voltages or bits being employed. Even resistor and capacitor values are approximate to the voltage equations -- that's why they're sold in the apparently strange multiples of 1.0, 1.2, 1.5, 1.8, 2.2, 2.7, 3.3, 3.9, 4.7, 5.6, 6.8, 7.5, 8.2, and 9.1 For those who actually design software synths, there are lots of ways of treating the values for pitch, volume, positioning, etc. Download a copy of SynthEdit and notice that all the work is done in *virtual voltages*, with Midi information translated in and out of a 'container' before it hits the interior modules! Now back to Midi panning. 16-bit values can be used for many parameters. Whereas panning within Finale might still be set to only 128 discrete values, in other software an assigned controller may replace the older pan controller, using a more refined scale of 32,768 values. Such higher resolution is very important if you're mixing the music to a 3D space. I'm suggesting that locking an abstract behavior to a fixed numerical identification may be appropriate for certain ways of working. And, of course, it's only software -- so the choice of scale should be configurable to the user's preference. There will be times when the Midi value is going to be critical, and I'd want to know it. But usually not. It seems to me that percentages abstract position of knobs and sliders pretty conveniently, because most actual values don't have importance -- only their relative position does. Isn't the reason for abstracting the numerical values to knobs and sliders in the first place meant to exploit the convenience of the physical position, however you prefer to name it? Dennis _______________________________________________ Finale mailing list [email protected] http://lists.shsu.edu/mailman/listinfo/finale
