First note of loop missing

[DrApostropheX]

I'll try with Surge tonight, which is free, WIN/MAC/Linux: https://surge-synthesizer.github.io/

I'll also try Sitala, which is a sampler, but also WIN/MAC/Linux and free: https://decomposer.de/sitala.html

I can only test the Windows versions, though.

[DrApostropheX]

I did some quick testing and with Surge and Sitala, it plays the first note perfectly, every single time through the loop (they're fun instruments and very easy on CPU too, so I'm going to have to play with them more often).

However, when I put the free AAS Player instrument (https://www.applied-acoustics.com/swatches/overview/) on its own track with a note right at 1:0:0, some patches it would play most or all of the time, and other patches (just scrolling through the list of freebie patches) would not play at all... the "Blue Rhythms -> Swagger Wah (key)" patch wouldn't play a note at 1:0:0 at all in loop mode. I know the AAS Player has a variety of synth engines in it, so it seems like the behavior is engine dependent.

I made a quick video, if it would help.

[fde101]

This hardly qualifies as light-weight, and I believe it is sample-based, but it is free and it does demonstrate the problem - IK Multimedia (who makes SampleTank) offers "Syntronik Free" if that is still helpful. I tried a bunch of other free synths and this is the one so far that chokes on this issue.

https://www.ikmultimedia.com/products/i ... eesoftware

[fde101]

the "Blue Rhythms -> Swagger Wah (key)" patch wouldn't play a note at 1:0:0 at all in loop mode.

Curiously, that one does play for me.

[DrApostropheX]

the "Blue Rhythms -> Swagger Wah (key)" patch wouldn't play a note at 1:0:0 at all in loop mode.

Curiously, that one does play for me.

Yeah. It's really bizarre. My project was at 94bpm so it's not like I was doing 280bpm doom core or anything.

[GoRhino]

Not sure if this is relevant to this interesting thread, but I often experience MIDI loops where the first note sometimes is louder than it should be. As I recall the behaviour is not consistent (ie first pass OK, second is loud, third is OK) and occurs other than 1.0.0 (I usually leave a couple of bars empty at the start of a project).

Most of the time I am using kontakt - so sample-based.

Is this a known issue?

[Stuttaton]

Not sure if this is relevant to this interesting thread, but I often experience MIDI loops where the first note sometimes is louder than it should be. As I recall the behaviour is not consistent (ie first pass OK, second is loud, third is OK) and occurs other than 1.0.0 (I usually leave a couple of bars empty at the start of a project).

Most of the time I am using kontakt - so sample-based.

Is this a known issue?

This happens here too. At least with W9. I just switched to W10, so I'll see if it's still there.
I'm using Collective and hardware synths.

[fromwithin]

Are you sure that's just not a problem with the MIDI event viewer you're using?

That's a very good point, so I have checked and it looks like that MIDI Monitor might indeed have a problem. That's that deleted from my machine then. I can't find any other 64-bit MIDI Monitor VSTs, but Reaktor 6 has one in its library. I've just tried it and it shows the commands in the expected order.

However, presuming Reaktor's MIDI implementation works correctly, it has shown up some other bugs. And it would be a huge surprise if Reaktor wasn't working correctly.

Bug 1:
On first play, Reaktor is receiving a Note On before a Start and Reset command. The Reset command only happens if play is started when the Play Position is at zero. The Start command is benign, but the Reset?

From the MIDI specification:

Reset. Reset all receivers in the system to power-up status. This should be used sparingly, preferably under manual control. In particular, it should not be sent on power-up.

Could this be the source of the occasional missing notes at the start of tracks?

Waveform erroneous Reset command.png

I've also tried sending MIDI data out to MIDI-OX via a MIDI loopback. It too shows the events in the correct order, but doesn't show a Reset command. Is the Reset command being filtered out before it goes to a system MIDI Output? I don't think that it should be there in either case.

Bug 2:
The Reset All Controllers (CC#121) and All Notes Off (CC#123) commands are only sent in loop mode if the Out Marker position is within a clip.

Waveform loop inconsistency.png

[fromwithin]

Not sure if this is relevant to this interesting thread, but I often experience MIDI loops where the first note sometimes is louder than it should be. As I recall the behaviour is not consistent (ie first pass OK, second is loud, third is OK) and occurs other than 1.0.0 (I usually leave a couple of bars empty at the start of a project).

Most of the time I am using kontakt - so sample-based.

Is this a known issue?

Sometimes a duplicate note gets sent, so you're hearing two of the same note stacked on top of each other.

[fde101]

Yep, that "Reset" certainly should not be there, though it is hard to point to that as being the sole culprit as it appears to have happened only for the first iteration of the loop - the repeated iterations don't have that.

Also, I have observed that in some cases the note will play the first time and not the other times - so it would be playing when the reset is there and not after that.

I am curious about the "SongPos" column, however. It looks like after the first iteration the note is being turned "on", then earlier in the song being turned "off"?

[fromwithin]

Yep, that "Reset" certainly should not be there, though it is hard to point to that as being the sole culprit as it appears to have happened only for the first iteration of the loop - the repeated iterations don't have that.

It's only sent when the Transport is first started and Play Position is at 1,1,000.

I am curious about the "SongPos" column, however. It looks like after the first iteration the note is being turned "on", then earlier in the song being turned "off"?

I didn't even notice that, but it looks like the transmitted song position info is being updated after the Note On and CCs have been sent. It could be worth looking into.

[mtelesha]

I have this issue with loops. I noticed that when the note started right at the start of the clip it would play every other time (Using Zebra 2 and SynthMaster but I think this always happens)

FIX: I would give space between the start of the midi clip and the start of the note. Then it plays fine but if the clip is trimmed the the start of the note every other time it wouldn't play.

[mtelesha]

Are you sure that's just not a problem with the MIDI event viewer you're using?

I've looked at the code for a bit (https://github.com/Tracktion/tracktion_ ... .cpp#L1047)
and we add the allNotesOff message to midiBuffer on line 1074 (which is what gets sent to the plugin) before adding the MIDI notes on line 1092.

It seems to me that the all-notes-off should always happen first, unless the synth is ignoring the order of the buffer and going purely on sample position and getting a random order because they're all at time 0?

I'LL BE A CUSTOMER FOR LIFE! When a company has code on github and shows you what they are doing in regards to a bug that is just amazing. That these pieces are open sourced is even better. Thank you!

[chico.co.uk]

I'LL BE A CUSTOMER FOR LIFE! When a company has code on github and shows you what they are doing in regards to a bug that is just amazing. That these pieces are open sourced is even better. Thank you!

This might be interesting - https://www.youtube.com/watch?v=G8YofMK6eY0
And - https://www.tracktion.com/develop/tracktion-engine

[ogquality808]

I found a fix, at least it worked for me.

When you press D and it copies it, that's when the bug occurs.

If you click and drag the small icon to copy it manually basically, the missing note is fixed.

Here's what worked for me

https://youtu.be/3x2zzsSE1rI


**EDIT** Ok so it's back again, whether you copy it with "D" or press the icon and drag, it's back on missing so it's not that I guess.

***EDIT 2*** Again, sometimes it works, sometimes it doesn't.
This time I changed the synth and it did the same
https://youtu.be/RY2ooAjAcGU