Midi Pedaling and Disklavier
By George Bogatko
Larry Kellogg wrote:
> George, I downloaded the file and tried it. The pedaling information > you have included in the file caused some problems. ... So there were > half-pedal messages in the Midi file? If so, then maybe they weren't > on the right channel.
The file was composed directly in Cakewalk's "Piano Roll" view -- not hand played on anything. The purpose was to create a master MIDI file that Bill Jelen could then transfer into an 88-note Piano roll. His program (and QRS's also) translate the continuous controller pedal on/off messages into the sustain pedal hole on the paper roll. So the purpose of the CC message was indeed simply to turn the pedal on and off -- there was no attempt to do any 'half pedaling' at all. So I'm curious just what *did* happen that has caused confusion.
Also, the short notes (the "pahs" in "um-pah") are 40 ticks at 240 to the quarter, which is supposed to match the size of 'pahs' on original rolls -- a newer technique that seems to work well on the QRS numbers, and worked well on the BAM-BAM version as well. Do these play nice and crisp, or are they too short and jerky, or do they get dropped completely? In addition, is there enough space between repeated notes for the hammer to reset sufficiently to get the second strike or are they being blurred?
There is a lot of note-bleeding on sustained chords, both to add some depth for those players with broken sustain pedals, and also to take advantage of the increased control over the sonority that this technique offers. Does the Disklavier falter here or is it able to keep up?
I could, of course, go out and just buy a Disklavier, but at $30,000 a pop for a grand, that's more than I can afford, and I'd be killed if I tried to squeeze an upright in the house. So, I'm left with whatever a synthesizer will produce. A good Roland SSC has excellent piano sounds, but there's nothing like fooling with the real thing, hence the original request.
Thanks for the feedback folks!
George Bogatko
|
(Message sent Fri 14 Mar 1997, 14:53:17 GMT, from time zone GMT-0500.) |
|
|