Weird hapax bug with USB midi tracks cutting each other out?

So forgive me as I’m still new to the hapax, but the strangest thing is occurring and I’m scratching my head.
I’m new to hapax, but definitely not new to midi troubleshooting, sequencers, etc in general. I have been dealing with a complex setup for years including doing things like midi loopbacks, using a midihub filter/routing box, etc. That said, it could 100% be something that I’ve just overlooked, or am just not getting about the hapax.

I’m sending midi over USB to my MPC one, then piping one channel out to tr8s over the MPC din midi port.

Setup:
T01 - Drum track - USBch1 to mpc track 1 (midi track) then on to tr8s - no issues seen here
T02 - Poly track - USBch2 to mpc track 2 (synth program, pluck/bloop)
T03 - Poly track - USBch3 to mpc track 3 (separate synth program, string/pad)

I will leave out further T01 as it’s not relevant.

T02 - 64 step pattern, but a repeating 32 step phrase (just doubled) -the 32 steps are a 2 note “chord” played on 1st beat and 6th beat, and a handful of single notes in between, fairly sparse. The ‘chords’ are 2 16ths long and the single notes are a 16th long. Then repeating for steps 33-64.
T03 - 32 step pattern - 4 note chord, played a few times shorter (3-4 steps long), then one time longer starting on the 5th beat and held until the end of the 7th beat.

Now for the weird part -
Every 2nd time T03’s sequence plays, a single note from the T03 held chord on beat 5 cuts out when the T02 “chord” plays on beat 6. The first time it plays through T03 normally (so the first 32 steps). The other parts of the pattern where notes overlap between the two tracks there are no issues at all.

If I mute T02, T03 chord plays correctly every time. Unmute T02 and the problem with T03 returns.

I assumed it was something going on with my routing, but the mpc midi monitor confirms that the notes are being received to correct alternate channels (2 & 3). the tracks are only set to receive on a single channel each, so there shouldn’t be any interference (confirmed with midi keyboard and setting it’s channel).

The midi monitor actually shows a note off message being generated for the T03 note that is cutting out, when according to the step sequence on the hapax, the note should still be held/not stopped yet. So where on earth is the note off message coming from?

Finally, one thing that is kind of good, because it gave me a temporary solution, but kind of confuses things further - this is only occurring on the usb midi out. If I change to Midi A, the problem disappears.

Other thoughts to mention:

  • I do not have midi output enabled on the MPC for the hapax USB port - ie no data is being sent back down the USB line from the MPC (common problem with usb midi troubleshooting, especially when you start considering usb thru settings).
  • There are no other tracks on the mpc playing or even setup that might be eating polyphony/doing weird things.

Could be a bug. Maybe you can post your project here, so we can load it ourself to see if we can reproduce it?

Also, what is your setup? Try to disconnect everything except the midi output to TS03 and see if it still happens?

This made me think of a MATH condition put by mistake


This invalidated the above theory

One possible culprit could be a large amount of MIDI Time Code / MIDI Machine Control messages sent by the MPC.
On your MPC, would you mind trying :

  • Menu > Preferences > Sync > Send : OFF
  • Menu > Preferences > Sync > Send MMC: OFF

We added a lot of safeguards for DIN ports, to mitigate bandwidth saturation, but less so on USB, this could explain the counter-intuitive behaviour.

1 Like

MMC and MTC are disabled, I send midi clock but as mentioned there are no ports outputting to the hapax, the only output port open is the MPC din midi out going to tr8s (this port on the mpc has sync out enabled, but not relevant as there’s no path out of the tr8s, din midi out and USB are not plugged in, fyi there’s no thru port).

FWIW hapax is clock master, sending out the USB to MPC. MPC receives clock only, no MMC, no mtc

My mistake.

Would you be so kind as to send us your project files and your settings file, so we can try and have a look ?
It might be a bug on our part wrt USB.

Hi,
Unfortunately I didn’t save the original project.
I had chalked it up as just some weird thing, but then I think I found it again -
In this case, unmuting track 10 causes track 14 to suffer dead notes as the notes from track 10 play.

At least that was the process I followed - I had been mucking around with track 10, muted it, started writing on track 14 (and others), then came back to unmute track 10 and all of a sudden notes are being lost like last time.

I have tried with synths that aren’t the MPC now, and am still getting the same issue. I can’t see any way for a midi loop or similar to occur as it’s strictly one-way flow from the hapax (USB device port out - the other end of that (MPC) doesn’t have the hapax return port enabled for global, track, or control) - the only controller I have connected to Hapax is a keyboard via din midi in port and that isn’t connected to anything else.

If I re-mute track 10, 14 picks up playing the notes again.

Very keen to hear if the issue is reproducible on another device. Thanks for your time.

there is a long-identified bug on MPC where overlapping MIDI note values across multiple tracks will choke each other off. it was claimed to have been fixed a few firmware updates ago but i still encounter it when sequencing my MPC from Hapax or Deluge. basically you have to map the MIDI note values of each separate track to not overlap each other. doable for me since you can map all the way down to C-2, which is where i put all my drum program pads. plugins and keygroups i generally map above C2 and then transpose up/down as needed to avoid the overlapping note values (it’s totally an MPC thing and nothing to do with the way Hapax transmits over USB or 5-pin MIDI)

1 Like

Definitely sounds like it then. That sucks. I still like using the MPC as a midi routing tool - eg layering beyond 16ch, midi programs, program change messages stored for songs/sequences.

it’s annoying, moreso that Akai is well aware of the issue but still can’t seem to implement a permanent fix