On 2020-01-28 1:27 pm, Bric wrote:
This might be broadening the scope of the original concern, but as
food for thought:
One would want to control the timing of the HH closing for *artistic*
purposes... the HH closing can make for rhythmic expression (and often
does, in live performance)....
General MIDI is limited in its expressiveness of nearly all instruments,
but in particular percussion. The original design was overly simplistic
given the variety of drums and cymbals that exist plus the myriad ways
to articulate them. That is not to say MIDI drums cannot be useful in
their own right; but achieving anything close to a real drummer on a
live kit requires careful programming with custom-crafted sound banks
which frequently ignore the GM standard's mapping of drum sounds to
notes. After all, the synthesizer really does not care whether note 56
actually sounds like a cowbell. If you do not need the sound, use it
for something else.
I use Reason in my music workflow and have several Refills for drums,
many of which substitute the range of supplemental percussion for
alternate playing styles of a standard kit. Some of them, for instance,
include several degrees of partially open hi-hat. Programming these
variations in place of an otherwise fully open hi-hat sound helps to add
a more natural sound.
That said, however, the automation of the above (in lilypond) may be
outside the scope of what the app is intended for (primarily score
engraving). Correct me if i'm wrong.
It would be a difficult undertaking to support all of the non-standard
drum synths and sample banks out there. The principle advantage of the
GM standard is that it is a singular, generalized thing to target.
I suspect the original intent for MIDI output in LilyPond was to allow
users to spot-check their scores, as it may be easier to hear a mistake
in engraving that we might not otherwise see during proof-reading. But
some folks may be using the MIDI output to generate backing
accompaniment or practice tracks. In support of that, the output should
be a practical and high-quality rendering of the score. In fact, I
would love to see something like articulate.ly that humanizes the
performance more so by adding in natural imperfections. Since few
performers are machines, it could be beneficial to listen to several
"takes" of a complex score to gauge its feasibility.
-- Aaron Hill