QLab losing time?

[Steve Richardson]

Hello!  Bit of a weird one...

Fairly busy QLab audio show being triggered from Qsys over OSC.

All running fine, except there are a couple of cues where QLab feeds back (via OSC) for certain visual elements in time with audio content, thing like actuators opening or closing etc...

Over time, these cues seems to "drift" slightly and there is a delay on them triggering, up to 5 or 6 seconds.

There's also another element of the show where staff acknowledge an alert with a push button.  Again, this response seems to drift over time.  Weirdly, can see I/O coming in to Qsys instantly, and the OSC string coming in to QLab, but the cue triggering will sometimes have a delay or lag to it.

Machine is a Mac Mini M2 and the network is a Netgear M4250 system.  Like I say, all running well and efficiently, other than these two weird bugs.

Any helpful ideas?  Thanks in advance...

[Chris Ashworth]

Hiya Steve,

To troubleshoot we’ll need a detailed description of what is hooked up to what, which signals are going where in what order, which audio devices are used, etc.

There are a lot of clocks at play in a QLab workspace. For example: OSC is following the computer clock while audio is following the clock of the device it’s playing to. If video is involved, that could be following either the video clock or an audio clock. If timecode is involved, that’s a whole mess on its own.

If cues are running for a long time, different clocks will drift from each other over time which can have noticeable results.

If that’s what’s happening, we could look for ways to avoid the drift if we have a more detailed understanding of how all the parts hook together.

It might be necessary to send a copy of the workspace itself to sup...@figure53.com so we can look in detail.

Best,

Chris

[Sam Kusnetz]

Hi Steve

A few questions:

1. Are you rebooting the Mac daily?
   
2. If you manually trigger the cues on the Mac directly, do you see the same lag?
3. In QLab, open the Workspace Status window, go to the Logs tab, and log incoming OSC. Does the timestamp on the message come before or after the perceived lag? That is to say, does QLab get the OSC message on time but then wait before starting the cue, or does QLab get the OSC message late?
4. Is this happening on specific cues only, or across the whole workspace?

Best

Sam

 

—

Sam Kusnetz (he/him) | Figure 53

s...@figure53.com

--
Contact support anytime: sup...@figure53.com
User Group Code of Conduct: https://qlab.app/code-of-conduct/
 
Instagram: https://www.instagram.com/Figure53
Bluesky: https://bsky.app/profile/qlab.app
---
You received this message because you are subscribed to the Google Groups "QLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qlab+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/qlab/766a0840-a4e5-4fcd-9854-7855fddad910n%40googlegroups.com.

[Steve Richardson]

Hi all,

Apologies for the slow response - We actually implemented a few things (and updated Qlab) and wanted to soak it for a while before coming back.

The good news, is the issue seems to have disappeared, I suspect the Qlab update had some other bug fixes which potentially also fixed a few other unknown ones.

The system appears to have been stable for a full week now, but will report back if it presents itself again.  

[jastan...@proton.me]

To Chris' point about clocks, I'll add that using an earlier version of QLab with Focusrite 2i2 USB Audio Device, things seemed fine for just audio.

  When adding video to the equation, audio and video would drift out of sync. The 2i2 wasn't a robust enough device to clock from. 

As advised by our fabulous QLab team, switching to MOTUs (at the time the Ultralite Mk3) resolved that.

  Since then I'd suspect any inexpensive audio device in a sync problem.

Jim

[Steve Richardson]

Hi all,

Sadly I spoke to soon and the issue presented itself over the weekend.

I will send a copy of the workspace file over but in the meantime thought I would go into the system configuration in case others are having a similar issue:

Qlab machine : MAC MINI 8C CPU 10C GPU M2 16GB 512GB

Qlab version : 5.4.9

Dante interface : 2 x RME Digiface Dante with USB C - USB3B adaptors

Dante primary clock : QSYS Core Nano

Main OSC source : QSYS Core Nano

In terms of cueing, there is a "show start" OSC string for the QSYS Core Nano which starts all background atmos cues, they will run indefinitely until a "show stop" command is fired at the end of day.

All other cues come in on OSC strings relevant to individual triggers and other IO triggers received by the QSYS Core Nano.

There are a few cues where Qlab sends feedback back to the QSYS Core Nano at certain times during those tracks.

The issue we're seeing, is with one or two cues in particular, where the OSC trigger is being received by all controllers (lighting, Brightsign and Qlab) but for whatever reason Qlab is sometimes delaying output by 10-40 seconds.

I have verified that the OSC string is coming in to Qlab correctly and at the right time, the delay seems to be within the cue or Qlab itself, but not every time (before this message, it hadn't presented itself for a number of days).

Closing Qlab and reopening seems to fix the issue.

[Sam Kusnetz]

Hi Steve

Please send a copy of the workspace to sup...@figure53.com and include as much information about the system setup as you can.

Best

Sam

—

Sam Kusnetz (he/him) | Figure 53

s...@figure53.com

--
Contact support anytime: sup...@figure53.com
User Group Code of Conduct: https://qlab.app/code-of-conduct/
 
Instagram: https://www.instagram.com/Figure53
Bluesky: https://bsky.app/profile/qlab.app
---
You received this message because you are subscribed to the Google Groups "QLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qlab+uns...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/qlab/07142baa-59e4-4d20-ba04-218439db2527n%40googlegroups.com.

[Jeremy Lee]

Probably a very long shot- but have you tried making the RME the clock leader? It’s a far better audio device and clock than the core. Depending on how many flows you’ve got going on, and uni or multicast, and IGMP/ QoS, you may be having a clocking/ PTP issue. 

 

Have you looked in Dante Controller for issues when this happens?

mobile +1 347.276.9468
web www.jjlee.com
email  jer...@jjlee.com

To view this discussion visit https://groups.google.com/d/msgid/qlab/CAOOK5AbjFUCeTDcUVt45TECjX_w6aAJFzUx8sZJkqvpGj7%2BOxg%40mail.gmail.com.

[Steve Richardson]

Hi all,

This issue presented itself again today after a few months of happiness!

Have uploaded logs etc through the "Contact Support" in help.

I hope it made its way to you - there was around 35 days worth of logging!