Antelope 32 HD delay compensation issues [SOLVED]

[casiotone]

Right...

So i checked all the manuals and re-did all the loop sync connections:

Antelope HD loop sync out -> titan loop sync in -> titan loop sync out -> avid hd loop sync in -> avid hd loop sync out -> Antelope HD loop sync in.

Tried to sync the hardware via the WC in of the first interface (titan) and to the internal from antelope hd (the only two options that actually work) and...

...no change. exact same behaviour. i would say it's even a bit worse now, because all three are off (the titans and antelope by 1 sample and the avid by some magical mysterious number that sounds like it's less than one sample, but it's not.

*sigh*
any ideas?

[JFreak]

Contact support of all said vendors

[casiotone]

Already done that waaay before posting this... been around this for a week now. Originally i thought it was just the titans problems, but since the it keeps changing it looks systemic.

A strange thing i noticed is that when i change the clock source in one of the interfaces, all the interfaces automatically change to that clock source.
It appears to be impossible to set up a single clock source for each interface in my system, even if all the manuals (and even the first answer from avid support) say it's possible and that's how its configured.

Tried and search the forum about that but didn't found any clear info / other case...
If someone knows something about this problem i would be greatly appreciated.

[JFreak]

Loop Syncs works just that way: there is one clock master and all other boxes lock to it. Just like normal word clock chain does, except -- every Loop Sync box somehow adjust the clocking so that the whole loop however large it is keeps in sync and don't start drifting if you have +10 boxes together. The clock master that gets the loop signal back is somehow responsible for it. Way too technical to explain here, but there is a reason for it being a loop.

If you can isolate the problem by using one brand interface only, then you have a winner. First thing you should do is connect Avid interface *only* and verify you have a problem with it. Then you have a valid case.

[Top Jimmy]

In hardware setup, the antelope clock source should be internal and the Titan and Avid clock sources should be on loop sync.

Loop sync is nothing more than a word clock distribution system. The loop master provides the clock on output and termination on input. The rest of the devices in the loop are unterminated and the in/out jacks are paralleled.

[casiotone]

Managed to figure out the problem after some 10+ hours of tests so i'll leave it here just in case someone needs it:

First, disconnecting the Antelope from the port solved all issues immediatly.

Second, sync loop needs to be Orion 32 - Titan - Avid - Orion32.

Third, clock needs to be set on Internal, interface 1 (Orion 32)

Fourth, Antelope Orion 32 doesn't like to be in port 2 of HDN.
Had to change it from port 2 to port 1 for the clocking issues to stabilise with the other interfaces

Fifth, Antelope Orion 32 doesn't auto-compensate correctly in the control panel, "HDX compensation" needs to be set to "manually" in Orion 32 HD control panel and then you need to adjust all the trimmers, either holding command(on mac) for fine tunning or introducing the numerical values.
They were all off by exactly one sample:
Auto compensation put 28 in the control panel, correct value is 29.

Gonna contact them to see if theres a solution for the issue but their technical support has not been great with other issues.

JFreak and everyone, thank you for your contributions it really helped figure it out.

Maybe it would be better to change the title to Antelope 32 HD issues with PT or something, since the issue is clearly the Antelope and it could help someone with the similar problem. Dunno if that's possible, just an idea.

Cheers

[JFreak]

Fixed the title for you. Glad you got it sorted out

[ZEUSS]

Originally Posted by Top Jimmy

In hardware setup, the antelope clock source should be internal and the Titan and Avid clock sources should be on loop sync.

Loop sync is nothing more than a word clock distribution system. The loop master provides the clock on output and termination on input. The rest of the devices in the loop are unterminated and the in/out jacks are paralleled.

So with an interface with loop sync and interfaces without, what would be the best scenario to lock and clock with an HD native thunderbolt box?

In my scenario, I am running a Red Line 16( has loop sync plus WC ports) 1 Rosetta 800 and 1 DA16X with digilink cards (no loop sync) I also have a Hedd connected via Spdif and have a Big Ben available.

I have tried a lot of scenarios but have issues when having to change sessions with different sample rates. Protools gets “stuck” and Swods and becomes unresponsive. I do have a workaround but it’s extra steps that I would love to stop having to do. Any advise would be appreciated.

It seems protools is looking for some sort of signal via loop sync to help it change. One observation, when I quit protools, the Red 16 defaults to 192 sample rate.

When everything is working,something happens and the system seems to get confused when opening another session with a different sample rate.

My workaround. Shut down. Turn off apogee units. Restart Mac, start protools with only the Red 16. Open at new sample rate. Then shutdown and turn apogee on. Reboot and start protools. Open session and it is fine.
Current and best way to get it to work is have Red Line 16 in port 1, Rosetta and DA16X in port 2.

Red 16 set to internal in protools as first interface with clock output via wordclock to Big Ben. Big Ben distribute to Rosetta 800 with T connector and 75 omg termination cap, and DA16X and that has the termination switch on as its last in chain. Clock from BB also goes to the Hedd which also has a proper termination with T and cap.

Apogee are at current firmware and show up as HD I|O’s

Tried having BB clock entire system via WC “Star” method but have huge issues with protools swodding and locking up when changing sample rates.

First scenario works with the work around.

I used to do the Star clock method with 192’s and the apogee with my old TDM system with no problems. Something changed along the way.


Sent from my iPhone using Tapatalk

[BScout]

Originally Posted by ZEUSS

when I quit protools, the Red 16 defaults to 192 sample rate.

This is because in macOS Audio MIDI setup, your AVID CoreAudio driver is set to 192. Change it there and this will stop happening when you quit Pro Tools.

[BScout]

All our interfaces are externally clocked. For all our systems. And we do not have problems at all.

We don't use loop sync at all (and if you use an external clock, you should not have the loop sync bnc connectors also connected; that confuses Pro Tools.)

You have to set external clock per device from the Session setup window. (Yes, stupid, but true. Flip through each device and then go back to whatever you want to signify as the "loop master" as the last device to set clocking even if you aren't using it as a "loop master.")

Interesting thing, on Avid interfaces' word clock and loop sync, it is unbuffered from in to out. With loop sync always working at Fs (vs 2Fs or 4Fs even if that is your session sample rate is a 2Fs or 4Fs rate.) Word clock input/output does stay at the actual Fs. However, the MTRX does not do this as it is not an actual true Avid interface. It is buffered. It also always operates on its internal clock -- only "referencing" external clock input if set to use word clock input. For that reason, all our systems with MTRX units use the MTRX as the master clock (referencing Pro Tools so it will follow the session sample rate) and one MTRX in a system will send its external WC to a WC distributor(s) that then sends to the other Avid or non-Avid hardware (including the SyncHD) in a star configuration.

Also make sure 75ohm termination is set up correctly depending on what the end points of your word clock chain are. For instance, the Sync HD is not terminated and a terminator should be placed on the WC out if you are sending a WC in from an external clock. You kind of have to go through all the hardware connected to find out what they do unless they are very clear about it (like RME is with the blue button/yellow led for termination.) And at worse, you put a T connector on a WC in and terminate the other side of the T.