[gmpi] Re: [OT] Re: 3.15 MIDI
- From: Paul Davis <paul@xxxxxxxxxxxxxxxxxxxxx>
- To: gmpi@xxxxxxxxxxxxx
- Date: Thu, 17 Jun 2004 00:58:17 -0400
>If by reset "immediately" you mean "on the next sample", I agree
>that would probably not be advisable... But I don't think the
>developers of MIDI expected response that that soon.
>
>As you say, in MIDI time, "immediately" really means something
>more like "in the next .6ms"... at minimum. It could also mean
i.e. a timeout :) and its not really the next 0.6ms. there could be a
big sysex on the wire. or even something small like a MIDI clock and
then a MMC locate command. and *then* the LSB might show up. you
basically have no idea how long you should wait.
i believe that the main reason this hasn't come up much so far is (1)
hardly any MIDI h/w sends 14 bit messages (2) most implementations
kludge various aspects of the problem. i haven't had the priviledge of
seeing any implementations, however.
>It also depends on things like how you have implemented level
>changes in your sound engine, what parameter you are controlling,
>and how much of a change it is to drop to zero. For example,
>I believe hardware volume ramps are often used in audio ICs to
>avoid zipper noise. Maybe this is not something software
>developers typically do, but if not, then maybe you are all
>just lucky so far <g>.
no, we can ramp, but the MIDI spec sorta-kinda implies that we're not
supposed to. one can interpret it liberally, and i think most of us
s/w people would tend to do that, but its really quite unclear.
>The other thing that might explain why response time was not
>considered an issue by the authors is that the reset of the
>LSB only applies with 14 bit messages, where the control is
>supporting 16,384 values, so the LSB only has a maximum affect
>of 128/16384 (1/128). It doesn't seem that a change of 1/128
>in volume (or any other CC parameter) over .6 ms could result
>in audible zipper noise, but if I am wrong, I would love to
>see the data.
any abrupt change in many parameters will cause a discontinuity in the
output that leads to an audible click. how much is abrupt? it depends
on the parameter. for volume, 1/128 is about -0.07dB, so its not
incredibly audible to a human. in absolute terms, it would reduce a
0dB 16bit signal by 256, which is definitely in the "click" zone once
run through the D/A. for a scalar, non-logarithmic parameter like
"number of FFT bins", a change of 1 will cause drastic shifts in the
plugin's operation.
--p
----------------------------------------------------------------------
Generalized Music Plugin Interface (GMPI) public discussion list
Participation in this list is contingent upon your abiding by the
following rules: Please stay on topic. You are responsible for your own
words. Please respect your fellow subscribers. Please do not
redistribute anyone else's words without their permission.
Archive: //www.freelists.org/archives/gmpi
Email gmpi-request@xxxxxxxxxxxxx w/ subject "unsubscribe" to unsubscribe
Other related posts:
