linuxaudio.org

On Wed, April 14, 2010 13:38, Ralf Mardorf wrote:

quoted text
> Jens M Andreasen wrote:

>> On Tue, 2010-04-13 at 22:04 -0400, Paul Davis wrote:

>>

>>

>>> MIDI-driven synth engines offer various choices about how to handle

>>> the situation, because people realized early on that it wasn't that

>>> difficult to make it happen, and that a useful way of handling it

>>> would be better than pretending it wasn't possible.

>>>

>>

>> But .. In my book it is still a bug, deliberate or not.

>>

>> Accidents can happen and therefore we have airbags. But this doesn't

>> mean that we are endorsing reckless driving as an acceptable method of

>> minimizing the time taken for commuting.

>>

>> As for the problem presented by OriginalPoster, send a note-off

>> immediately after each note-on and be done with it. Then use

>> sustain-pedal for playing several identical notes as a chord.

>>

>

> Full ACK. There are some drum sample players that have problems when the

> duration between note on and note off is to short and there ... there

> are several reasons when not sending a note off might be nice, similar,

> for some equipment it is nice if running status could be disabled, some

> old sequencers can do this etc. pp. ... somebody wrote that one note off

> would finish all note ons of the same number, this must not be true all

> the time. Especially old MIDI equipment can behave very unusual. Anyway,

> I'm using MIDI since it was on the market and even if there might be

> some advantages when not sending a note off, there will be on

> disadvantage, the most negative thing I know about MIDI, the "hanging

> note". Or is it "frozen note" on English? The note that never will end

> ;). In case of doubt for MIDI we shouldn't think about advantages for

> computer equipment, but how it would be for a natural instrument. I

> guess on that list we found out, that every natural instrument will turn

> a note off before the same note could be played again.

that makes three of us who don't know what i'm talking about..
the case of the missing note offs within a string of same-pitch note-ons,

i'm having difficulty classifying as accidental or intentional. these

missing note offs occur because of the design concept - you don't need to

have a working program to prove it. unfortunately it's a design i wish to

implement. the separation of rhythmic data (position + duration) from

pitch + velocity data, and the method of generating the pitch+velocity

data does not mean that note-offs are never sent, it only increases the

likelihood of note-offs being absent when the rhythmic data overlap each

other, and the same pitch is generated in sequence due to the

pitch/velocity generating algorithm. it would be an error if this error

did not arise in these circumstances. i'm beginning to formulate ideas for

solving it, but there is no absolute solution. the end.
_______________________________________________

Linux-audio-dev mailing list

Linux-audio-dev@lists.linuxaudio.org

http://lists.linuxaudio.org/listinfo/linux-audio-dev