Delay compensation using multiple interfaces and sends/buses on PT 2022 - help!

[casiotone]

Hi All!
I’ve got a complicated parallel set-up for paralell processing and i’m having some problems with delay compensation, wondering if someone could help me figure this out.

So i have an unprocessed track (1) that has a send to bus 1-2 and this bus feeds three other tracks for parallel processing and all four tracks feed a second bus (3-4) to be summed together.
Track 2 receives bus 1-2, monitoring on, i/o 1-2(interface A), perfectly in sync with unprocessed track
Track 3 receives bus 1-2, monitoring on, i/o 3-4(interface A), also perfectly in sync with unprocessed track
Now, track 4 receives bus 1-2, monitoring on, i/o 1-2(interface B) and it completely messes the delay compensation

Interface A is a prism titan that reports latency perfectly to PT, everything in sync
Interface B is an old Apogee AD-16X/DA-16X with X-HD card that requires 1.22ms delay compensation (at 48K) and that’s how’s set up.

Now the delay compensation presented on the mixer for interface A is 2195 for track 2 and 3
But the delay compensation for interface B for track 4 is 2189

I’ve tried everything, to compensate on the i/o until the delays of both interfaces were 2195 on all tracks, doesn’t work, still same problem.
I tried adjusting the +/- 6ms both to positve and negative, both on the track 4 with the 2189 delay and on the other two tracks for 2195 and nothing appears to change.

If do a copy of track one for track 2, 3 and 4 and bypass the aux send and just use the i/o individually, it compensates perfectly for both interfaces even with different latency values on the mixer, but this system doesn’t really work well for me.

So, my question is what am i doing wrong here? How can i adjust the slight difference of delay compensation on the track with interface B to make it work in tandem with interface A ?
(pretty please, avoid suggestions of only using interface A or the solution i already presented, i just need to know what am i doing wrong with the delay compensation)

I hope that was clear?!
Thanks in advance!!!

(Pro Tools 2022.12 running on NHD on OSX)

[casiotone]

So i just found out if i put the timeadjust plugin on the track of interface B with 6 samples delay it compensates perfectly, so that's a solution, but it's burning a plugin slot.

Is there any way to add these 6 samples delay to the plugin compensation?

[originalscottyg]

Go to I/O Setup and then the HW Insert Delay tab. You can enter hardware insert offsets per channel in msecs. You will have to calculate what 6 samples at 48kHz is in msecs.

[Darryl Ramm]

Do you have the MDIO-HDX card in the Titan? If so the Titan is not reporting anything to Pro Tools about latency. Pro Tools assumes it has the latency of whatever DigiLink box it is emulating, aka lying to Pro Tools it is. Now wether it actually has is up to the implementation, some vendors are sample accurate (adding additional delay to their hardware to match the old Avid boxes) some are not. Seems it is.

I am struggling to follow what exactly you are doing. In plain simple English. Is this running as hardware inserts or you just running this as track physical inputs and outputs? Or you running Aux output to the physical IO? You are talking about 6ms and 6 samples, so what is it?

You might make more progress with a simple as possible test session, no software plugins, if you are not using HW inserts you should be able to turn off delay compensation and see no change. Please share that session.

Either way I would turn off ignore errors and low latency monitoring to avoid possible known stupid latency bugs.

[Darryl Ramm]

I am also lost as to where you are doing stuff, where exactly are you setting 1.2ms latency… if that is in the HW insert setup panel, that only affects things if using HW inserts, if you are using HW insets maybe you just need to micro adjust that setting.

Screenshots of a very simplified tests showing the HW Insert settings and track delay compensation values might help.

[casiotone]

Originally Posted by Darryl Ramm

I am also lost as to where you are doing stuff, where exactly are you setting 1.2ms latency… if that is in the HW insert setup panel, that only affects things if using HW inserts, if you are using HW insets maybe you just need to micro adjust that setting.

Screenshots of a very simplified tests showing the HW Insert settings and track delay compensation values might help.

Hi Darryl
Thanks for your input, it was actually helpful, i started again and traced back the problem to the interfaces.
I was suspecting my explanation wasn't clear, but upon further testing it doesn't really matter, because of the underlying conflict with the interfaces.
I'm going to leave what i found out here because it can be helpful for someone_

What's happening is that the Prism Titans with the MDIO card do emulate perfectly the AVID HD interfaces and report the latency correctly.
And the old Apoogee AD-16X/DA-16X with the X-HD card also emulate the AVID HD interfaces but don't report latency correctly and need an offset on the HW inserts ( that's the 1.22ms at 48k)

Now, the real problem is that when only the Titans hooked up to the system all the inserts latency is solved (almost full phase cancel with inserts).
But the moment i hook up the Apogee devices, the reported latency from the titan is off by 1 sample (it's 1 sample back to be exact), which means i can't offset it on the hw insert.

So i added 1 sample delay to the track without any inserts and everything canceled again on the Titans.
But after doing this, the Apogees are now 1 sample off, but this was easy to offset by 0.01 ms on the HW inserts.

So now it's working, kinda, but for other users, be warned, the Apogee AD-16X / DA-16X with the X-HD card will mess up the latency reporting of your other interfaces, even if connected to a different port of the HDN card.

Thanks for your help again!