samplomatic 5000 short note MIDI bug

[olkash i psih]

hello
Reaper 6.37 Win 10 64bit, Win 7 64bit

in samplomatics 5000 when you turn on "obey note offs" and when playing very short notes (approximately shorter than 0.005), the actual duration of the sound becomes unstable and random- this can be seen on the oscilloscope

[olkash i psih]

Well, the bug is still here. Reaper 7.01

[olkash i psih]

bug is still here. Reaper 7.07

[axel_ef]

set Release to a higher value

[olkash i psih]

7.33 64 bit, win 10 bug still there

[axel_ef]

With obey note offs a release of 1 is too small.

[Sid]

Originally Posted by olkash i psih

7.33 64 bit, win 10 bug still there

I was wondering how many ticks per quarter note you have your MIDI preference set to.

[olkash i psih]

Originally Posted by axel_ef

With obey note offs a release of 1 is too small.

The problem is not that the sample is choppy, but that it becomes unstable in duration with short notes. I've tested other VST-s (TAL Drum, TAL Sampler) that can handle note-offs, and everything works fine there. I tried to increase the release up to 100, but the sound still floats in length

[olkash i psih]

Originally Posted by Sid

I was wondering how many ticks per quarter note you have your MIDI preference set to.

Default 960, but I'm sure that the problem is in the Samplomatic 5000, everything is stable in the event list of the midi editor, and the other VSTs that I tested are working correctly

[Sid]

You say the note is approximately shorter than 0.005. That is 0.005 what? Milliseconds? Seconds? beats? What is the BPM you are using?

Something I found online:
"The theoretical minimum length of a MIDI note is considered to be two-thirds of a millisecond based on the MIDI standard, which is the time it takes to send a single MIDI message with "running status" (a way to optimize data transmission); however, the actual playable minimum length can vary depending on the instrument and sequencer capabilities due to processing delays. "

Anyhow, I would suspect that if it is written anything like how the JSFX engine is written, information is delivered in blocks, and so depending on where the note is, you could have the note-on and note-off in the same block, or in separate blocks of data, and thus have the unaviodable delay of reading two blocks for some notes and one block for others.

[olkash i psih]

Originally Posted by Sid

You say the note is approximately shorter than 0.005. That is 0.005 what? Milliseconds? Seconds? beats? What is the BPM you are using?

0.005 is the length of the note, 125 BPM


le français ru a1

[olkash i psih]

Originally Posted by Sid

Anyhow, I would suspect that if it is written anything like how the JSFX engine is written, information is delivered in blocks, and so depending on where the note is, you could have the note-on and note-off in the same block, or in separate blocks of data, and thus have the unaviodable delay of reading two blocks for some notes and one block for others.

yes, this is probably a feature of the Samplomatics engine, when the note duration gets longer than about 0.006, then everything plays fine. This is not to say that it is a terrible problem and manifests itself in rare scenarios, but it should not be.

[Justin]

Thanks, fixing!