LP cv/gate latency?

[CTRLSHFT]

has anyone else experienced problems with sequencing the LP via cv/gate? I've had mostly decent success with midi (although midi is constantly hanging on this little guy), but the cv/gate issue is frustrating.

i'm using a future retro mobius to sequence the LP, and for some reason the LP is giving me some pesky latency. it responds late by about 100ms or so, and on top of that the pitch detunes about a half-step when you plug it in.

i don't experience these problems on other cv/gate analog synths. otherwise i'd blame the mobius!

has anyone else had this problem? could it be that the LP needs an OS update to address this kind of latency? my moogerfoogers respond perfectly in time with the mobius, so the OS is my guess (love it all the same! ).

any ideas? i'll definitely contact moog on the issue, but if any of you guys out there have experienced this problem with sequencing, please share your thoughts/stories, and if you found a solution, please help!!!

[Amos]

Hi there!

I had a look at this situation on the oscilloscope yesterday, and this is what I found:

In normal operation, there was a consistent 9 millisecond latency between the incoming gate trigger and the firing of the LP keyboard gat. This was the same even if notes were also played on the keyboard, and knobs/pitch/modwheel were tweaked.

There is a momentary lag when the panel function buttons are pressed, which adds no more than 25 milliseconds of latency depending on the timing of the events; that is to say that if you press a panel button at the instant the LP receives a trigger pulse, there could be an additional 25ms of latency just for that pulse; if the events don't coincide exactly then there is less delay.

Now, when the LCD display has to update, that causes the most latency. If you are scrolling through the master menu or moving the data encoder while you are playing a sequence, you will definitely get some wonky delayed notes as the processor writes all the new information to the LCD display. Precision Mode is constantly updating the LCD with high-resolution parameter info, and so if you are in Precision Mode then you will definitely have latency problems. Precision Mode is intended for scientific patch tweaking and is not optimized for live performance.

Check to be sure you are not in precision mode and see if perhaps that is the problem.

In other news, we have fixed a number of note-handling issues that have been causing the stuck notes; look for that in the next OS release which should be coming within a few weeks.

Cheers,

Amos

[CTRLSHFT]

Hi there!

I had a look at this situation on the oscilloscope yesterday, and this is what I found:

In normal operation, there was a consistent 9 millisecond latency between the incoming gate trigger and the firing of the LP keyboard gat. This was the same even if notes were also played on the keyboard, and knobs/pitch/modwheel were tweaked.

There is a momentary lag when the panel function buttons are pressed, which adds no more than 25 milliseconds of latency depending on the timing of the events; that is to say that if you press a panel button at the instant the LP receives a trigger pulse, there could be an additional 25ms of latency just for that pulse; if the events don't coincide exactly then there is less delay.

Now, when the LCD display has to update, that causes the most latency. If you are scrolling through the master menu or moving the data encoder while you are playing a sequence, you will definitely get some wonky delayed notes as the processor writes all the new information to the LCD display. Precision Mode is constantly updating the LCD with high-resolution parameter info, and so if you are in Precision Mode then you will definitely have latency problems. Precision Mode is intended for scientific patch tweaking and is not optimized for live performance.

Check to be sure you are not in precision mode and see if perhaps that is the problem.

In other news, we have fixed a number of note-handling issues that have been causing the stuck notes; look for that in the next OS release which should be coming within a few weeks.

Cheers,

Amos

WOW, thank you for such a comprehensive clarification on this issue! The latency I've been experiencing seems to be greater than than 30ms or so, but i'll give it another shot with those items in mind and see if i can't get it to lock a bit better tonight.

another thing i didn't mention: i have a +5 accent pulse attenuated down a bit through the cp-251 that is plugged into the filter cv input, for good ole moog acid, while the sequences are running.

also: the gate output on the mobius is actually +-12v.

don't know if those things would have any detrimental effects on the response time of the gate itself, but theres another thing to consider. i'll try dropping the filter modulation, and attenuating the gate signal to 5v and see what happens.

at any rate, thanks again for your research, i can't express how happy i am to have a studio full of products supported by brilliant people such as yourself. i'll post another reply tonight with my findings on getting this little guy locked in sync.

[CTRLSHFT]

here are the findings from my set-up:

*test one*: With my Mobius sequencer synced to external midi timecode via my DAW, and the LP recieving the sequence via CV/GATE, the latency was consistently 70-74ms.

*test two*: With MIDI notes running from my DAW directly into the LP, I consistently found latency of 25-30ms.

*test three*: With my Mobius sequencer synced to external midi timecode via my DAW, and the LP recieving the sequence via a MIDI cable instead of the cv/gate cables, I found a completely consistant (and AMAZING) latency time of only 6ms!!!

here's the clincher.

*test four*: I used a cv-midi converter box and ran a midi sequence from my daw out it, and connected the cv/gate inputs into the LP. latency: EXACTLY 9ms.

now I know for a fact that it's the mobius' gate output causing the problem here. (now it's time to go bug them i guess )

Thanks Amos, you really helped save the day here and eased my mind about the LP's capabilities!

Cheers,
Josev

[rachel]

It might be that the mobius is sending not a square pulse of a fixed
voltage, but instead a waveform that looks more like a slope. This would
show as a sort of latency like you experience.

My System 104 sequencer (which was recapped early this year) sends
a nice tight square wave. There's a slight latency when triggering this
from my Frostwave Quad MIDI/CV box but very little when just letting the
104 run free. I triggered the phatty from it and got consistently
low latency from both these, but I didn't try to check absolute times or anything,it just sounded better.


rachel

[CTRLSHFT]

i managed to get it locked!

for anyone out there using a mobius w/ a LP:

output gate +12v on the mobius. attenuate to +5v via a cp251. plug output to LP gate in.

tight as a drum.

dunno why i didn't think of this before.. guess it was becuase the manual for the mobius suggests -12v for moog units. this works on the moog, but the sync doesn't happen and the ADSR is really wonky.

whatever, there it is. yay!

[Amos]

Hey, sweet! Congratulations, and thanks for posting your results here for others to learn also...