Is there anything else which affects playback timing in Sonar

[norfolkmastering]

I'm getting some inconsistency with the timing of audio being sent to the SEND side of external inserts (I don't use the return so this is not a question about external insert round trip delay compensation).  I use these to feed analogue tape tracks in my hybrid studio.

I'm finding that the timing of the audio being sent to the external insert sends is varying from session to session on the same project.

So if I bounce a track from Sonar to an analogue tape track, then I bounce the same track from the same project on another session, I'm getting different timing (I check using an audio waveform click).

It's varying by about 250ms from session to session.

I have tested it with all FX switched out and so the PDC is showing as zero.

Is there anything else which might be varying on a session by session which would affect the time delay from an audio waveform being replayed from Sonar to the SEND of an external insert?

[Wookiee]

@norfolkmastering You may have done all this but......

1. Check Windows Audio settings are the same as Sonar/Cakewalk 

2. Have you disabled HDMI and the Realtek audio in device manager

3. In Windows Audio remove the check that says allow application to take full control of your audio device.

Some people lock the AUD.INI file this ensures your audio settings are the same everytime you start Sonar/Cakewalk, just remember to unlock it if you intend to make changes to your audio device  or settings. 

[msmcleod]

Is there any reason you're using External Insert's with no return rather than buses?

IIRC, the PDC pretty much relies on the return to calculate the overall roundtrip latency (although it may be that in "send" mode for external inserts, the input and output are essentially are a passthrough... so PDC = 0).

[norfolkmastering]

3 hours ago, Wookiee said:

@norfolkmastering You may have done all this but......

1. Check Windows Audio settings are the same as Sonar/Cakewalk 

2. Have you disabled HDMI and the Realtek audio in device manager

3. In Windows Audio remove the check that says allow application to take full control of your audio device.

Some people lock the AUD.INI file this ensures your audio settings are the same every time you start Sonar/Cakewalk, just remember to unlock it if you intend to make changes to your audio device  or settings. 

1. Yes the Windows Audio settings are the same as Sonar/Cakewalk

2. I have disabled the Realtek audio in device manager.  HDMI audio doesn't appear on the list.

3. Windows 11 doesn't have exactly the same options (that I can find anyway) but all the 'Audio, Allows apps and Windows to use this device for audio' are set to Don't allow'.

I don't think the problem is timing variations in the Sonar external insert sends.  I've tested these extensively today and they are very consistent.

I've done some more testing and think the problem may be slight timing variations in my sync system.   I use Sonar's MTC output as the timing master.  This  gets converted to LTC using the Show Cockpit MTC to LTC convertor app.  Sadly this PC based convertor only supports native Windows audio drivers (so not ASIO).  I've been using my PC's Realtek digital audio output via a D/A convertor to get the LTC in analogue format.  It's been okay up till now and tight enough 'timing wise' for my use.  So I'm not sure why I'm getting this variation now, although I did update to Windows 11 recently.

So what I'm saying is that the path from Sonar MTC output to Show Cockpit's MTC input (happens inside the PC) to LTC conversion (still inside the PC) to LTC emerging from my PC's digital audio output, is varying in the path time delay, a little bit from session to session.  That it's worked okay up till now is a miracle!

I can pull 'the system'  back into sync by putting a frame or two of offset into Sonar's SMPTE/MTC Offset.  This (I think) slightly advances or delays Sonar playback timing compared to Sonar's MTC output.  It's tedious but it works.

I probably need to find a master MTC/LTC generator which I can sync the tapes machines and Sonar.  That would still rely on a constant time delay getting the MTC into Sonar (so via a MIDI interface), or perhaps easier would be to find a high quality MTC/LTC unit which I can sync from Sonar (so bypassing the Show Cockpit app).

All suggestions welcomed!

 

[David Baay]

4 hours ago, norfolkmastering said:

So if I bounce a track from Sonar to an analogue tape track, then I bounce the same track from the same project on another session, I'm getting different timing (I check using an audio waveform click).

I was just going to ask what's controlling the tape sync when you posted. It seemed likely that was the culprit rather than the EI latency being variable.

[Starship Krupa]

14 minutes ago, norfolkmastering said:

I've been using my PC's Realtek digital audio output via a D/A convertor

I'm no expert, but I do know that one of the drawbacks of using onboard Realtek CODEC's is that they clock using the CPU's clock, which isn't designed with the same accuracy as the JetPLL'd clocks usually found in external audio interfaces.

I've only ever thought of it in terms of jitter, and 250mS seems like a HUGE error outside of any jitter that I've heard of, but I think you're correct in looking to your clocking system, and an external box seems like a good way to address it.

Although the Realtek CODEC isn't as terrible at playback as some would have us believe, I wouldn't expect it to perform well at providing a clock signal. As you say, it seems like a miracle that it's worked up to this point.

[Xoo]

What if the CPU clock is dynamically changing (most modern CPUs do)?  Can you disable that in the BIOS as a test at least?

[Starship Krupa]

23 minutes ago, Xoo said:

What if the CPU clock is dynamically changing (most modern CPUs do)?  Can you disable that in the BIOS as a test at least?

The Realtek CODEC shouldn't be using the CPU clock in that way. I should have said that it uses the same clock source as the CPU. Which is more prone to drift and jitter than the clock sources found in interfaces or dedicated hardware clock providers.

[Xoo]

Ah, OK.  Even so, worth trying :-)

[msmcleod]

@norfolkmastering - I always found the JL Cooper PPS-100 rock solid.  That's what I used in the 90's.  I've still got one, and it still works.

They've been out of production for a long time though (I think they were only in production between 1989 and 1995), but you can usually pick one up on eBay for anything between £75 to £150.

For something more modern, HHB do one (based in Norfolk too!) :

https://www.hhb.co.uk/prod/cb-electronics/tc-5b-midi/

 

[Starship Krupa]

7 hours ago, Xoo said:

Ah, OK.  Even so, worth trying 🙂

I'd try an external sync box, they're not so expensive.

The whole business of syncing and master clocks and whatnots kinda makes my head spin. If I ever need to figure it out I'll probably ask norfolk!

[Noel Borthwick]

I think you need a hardware mtc sync generator. I used to have a digi smpte slave driver that I gave away a decade ago.

[norfolkmastering]

I came across this MTC to LTC converter on Amazon, https://www.amazon.co.uk/DOREMIDI-LTC-MIDI-Timecode-MTC-20-synchronize/dp/B0CSFGCYZV

It's reasonably priced so I will probably get one and see helps with the issue.