Song position reset when a MIDI Start message is received?

Elektron Gear Analog Keys/Four
#1

I’m using an Analog RYTM as my MIDI clock master with both an Analog Four as well as an MPC-1000 as the slaves. I’m also using a MIDI monitor to keep track of what’s going on for debugging purposes.

All of these devices are in “Song mode” and currently stopped.

When I now change the song position on the RYTM (by moving the cursor and pressing “Yes/Save”), the RYTM sends out a Song Position Pointer MIDI message which results in both the Analog Four as well as the MPC 1000 updating their current song position to match the song position on the RYTM. This can be seen on their displays.

When I now press the “Play” button on the RYTM, a Start MIDI message is sent, and the RYTM as well as the MPC start playing from the correct position. However, the Analog Four first jumps to the beginning of the song before it starts playing instead of starting playback from the current song position.

Is there a way to make the Analog Four not jump back to the beginning of the song when a MIDI Start message is received, but instead have it start playback from its current position?

MIDI Message to start Pattern?
#2

Sounds like you’ve been thorough. Not in front of my elektron boxes right now. But I see that one of the bug fixes enhancements from an earlier os update addressed a related issue to do with the play button properly acting like a pause. It may just b a bit buggy in that config. I’d suggest trying a different master firstly the A4 and then MPC just to see if the two elektron boxes share the same trait. If the issue is only with the A4 and not the RYTm it sounds like it could be a bug assuming transport and clock settings are similar. I never really work that way, but I’d try narrowing things down with that test, I’d also limit it to one master and one slave to keep it simple in testing

1 Like
#3

@avantronica Yeah, I still need to test with the RYTM as a slave, just to see if it also jumps back to the beginning.

Please note that the MPC-1000 did not jump back to the beginning when it received a MIDI Start message.

It’s actually a bit of a complex issue; for example, according to the errata on http://www.blitter.com/~russtopia/MIDI/~jglatt/tech/midispec/start.htm, a device should begin playback at the beginning of a song when it receives a MIDI Start message, and a MIDI Continue message should be used to make a device continue playback from the current song position that was set using a Song Position Pointer message.

Regardless of this however, it’s a bit weird that you can make the A4 follow along when you change the song position on the RYTM, but that you cannot make the A4 start playback at the position you just set it to when you start playback on the RYTM.

1 Like
#4

(Please copy the link above instead of clicking it; it seems that this forum doesn’t understand that a URL can contain a tilde.)

#5

I got it that your MPC played ball, but I was guessing that if you had extended midi plumbing it might be a consequence of thru latency or such like. Just thought that a few simple A/B tests might help you point the finger. I understand that the timing schedule of the transport commands is fairly important for it to be robust, relative to beat clock to ensure things work right. I’m sure elektron are well on top of this, but often bugs creep in. I’m not sure if stop or stop-stop at the master will make a difference or whether you could make the pause by pressing play as opposed to stop. Will look at the link outta interest soon

#6

Might try this in Max with the OT and A4 to see what happens
The commands are here in binary

#7

The MIDI setup is simple and should not introduce any latency or other trouble. I’m using a simple hardware MIDI Thru box (http://midisizer.com/midithru/) to distribute the signal coming out of the MIDI Out port on the RYTM directly to the A4, MPC, and to my mac running Midi Monitor (http://www.snoize.com/MIDIMonitor/).

#8

Pressing [STOP] once vs. pressing [STOP] + [STOP] makes no difference; a Start message is always sent thereafter.

#9

Just had a play with the raw commands in Max, the OT responds to start, stop and continue
It also transmits them all, in order to transmit a continue you have to stop by pressing play(pause) as I thought
Might not help you, but at least elektron are on top of transport support, can’t test w A4 at the mo fwiw

#10

Ok, looks like the solution is simply to NEVER use the STOP button on the RYTM, but to always toggle playback by hitting its PLAY button. This makes the RYTM go into “pause mode” which makes it send out a MIDI Continue message instead of a MIDI Start message.

So, to summarize; both the A4 and the RYTM behave correctly according to the MIDI specifications. You just gave to remember to not use the STOP button if you want to be able to set a song position and start playback from that position remotely.

I’m going to send a feature request to Electron to see if they’re willing to make the sequencer always switch to “pause mode” after you’ve set a song position. That seems like slightly more intuitive behavior in this specific case.

1 Like
#11

Also if you have a separate midi controller you can send a specific continue message that is not the same as start or stop, well at least the OT plays ball with that

#12

Yeah, all the Elektron boxes respond correctly to MIDI Start, Stop, and Continue messages as far as I’ve been able to confirm.

It’s just that in song mode, the behavior of the Start button after you’ve changed the song position is not entirely intuitive. It would—IMHO—make more sense if it would always send Continue MIDI message once you’ve updated the song position, no matter if you’re in “stop” or “pause” mode.

1 Like
#13

same issue here. RYTM as master A4 as slave. Fixed by pressing PLAY to pause the song. As I put the cursor on the RYTM pattern from which I want to start the A4 cursor shifts on the same position but when I press play in RYTM the A4 starts with delay (around 64 steps). Any suggestions?