Audio output quiet for first bar after starting sequencer?

Elektron Gear Octatrack
#1

to set the stage a bit the main way I use the octatrack is for live input and live re-sampling. I have a project set up I have been using for a while (different variations of it over the past year or two) where there are 3 tracks+buffers set up with flex machines that continuously record live input (1 stereo input, 2 mono inputs) and I have the master track buffer set to continuously record the main outputs. These 4 buffers record 4 bars maximum. I run the octatrack mk1 at 24bit. The other 4 tracks are set to flex machines and buffers that will only record when activated manually.

I have 2 octatrack mk1s, and they both exhibit the same behavior where when I first start the sequencer when using this project, the first bar or two is very quiet before the octatrack kicks into full volume on the main outs. A few times here and there one of the channels (seems to be mainly the left channel) will be completely silent, and take a bar or two to ‘kick in’.

it’s not really an issue per se, since I basically start the sequencer when I turn my units on and as long as the sequencer keeps running it’s all good - it only happens when stopping and re-starting the sequencer, but it’s pretty annoying. For a while I thought I was having cable issues or mixer issues or something and it took a while to track down

mostly I’m curious if anyone is able to replicate this issue or has experienced this issue - I would assume someone should be able to, since it happens on both of my mk1s. I’m also curious if the mk2 does this. I have experienced this phenomenon with the most recent OS as well as 1.25H

my suspicion is that my heavy-duty live-input and continuous recording approach is choking a data bus or processor

few things I need to investigate

  1. how much additional output latency my live-input approach is creating vs. thru machines vs. direct inputs (with delay compensation turned off)
  2. if the headphone and cue outputs show the same behavior
  3. if 16bit exhibits this issue
  4. set all the flex machines to continuously record and output live feeds to see if that makes things worse

probably a few other things I could look into. any and all thoughts appreciated! I can and already do work around this issue, but I would like to understand things more deeply so I can make better and more informed decisions about how I want to configure and use the octatrack

#2

I remember someone reporting similar symptoms last year: it turned out that the input signals or samples were at low levels and the user was compressing them in order to increase the effective level, and this caused initial playback to be quiet. Are you using the compression effect on your audio tracks?

2 Likes
#3

well, very interesting! thank you for the response! it appears to be specifically related the master track compressor, and does not occur with individual channel compressors in my testing so far. The more I dig into this the more it seems there is a bug in the master track implementation of the compressor…

my input signals aren’t quiet - they’re coming from a modular setup so I actually have to avoid clipping the $@&! out of the OT inputs. What’s also interesting is that I have absolutely no makeup gain on my master track output compressor - I’m using it mainly as a soft limiter, but the master track is noticeably quiet for the 1st bar if I have the mix at 127, and then kicks up in volume, whereas if I leave the mix at 0 there will be no effect/gain applied and I won’t experience any level change after the first bar. So apparently the compressor effect adds gain even if you have makeup gain set to 0, but for some reason on the master track it takes a full bar after starting the sequencer for the compressor to actually ‘kick in’? Changing the mix values on individual track compressors does not duplicate this behavior, it only appears to happen on the master track. I think since I was using the compressor as a soft limiter with the mix set to fully wet this problem was much more obvious to me, whereas before if I had a parallel compression kinda thing going on for the master bus I would be much less likely to notice this behavior, since the mix wouldn’t be fully wet with the compressor effect

edit to add: OMFG ignore everything I said - there was a plocked trigger hiding on the second sequence page of my gd master track…

2 Likes