A word clock challenge for you. Please help.

[Eric Lambert]

I'll try to stick to the basics to keep it simple.

MacStudio Ultra is connected to an HD Native Thunderbolt box.

The HD Native Thunderbolt box has separate Digilink connecttions to a Prism Titan and a Lynx Aurora (n), both units having HDX modules. The Titan is connected to Port 1 of the HDNTB box and the Aurora to Port 2.

Word Clock cables are connected from the output of the Titan to the input of the Aurora, and from the output of the Aurora to the input of the Titan. A loop.

In ProTools Hardware Setup I have set the master clock to the Titan's Internal clock.

For a week this worked perfectly at every sample rate. I could switch to 96K and both interfaces showed Locked word clock. Then, I unplugged the word clock output of the Aurora briefly, then plugged it back in. Since then the Aurora interface refuses to show its word clock as "Locked" at sample rates of 88.2K and above. 44.1/48, on the other hand, are fine.

Prism states that they only use the base rate for loop sync, so at SRs above 48K the Titan is outputting either 44.1 or 48. Lynx states that their box is expecting to see 88.2 when the session is at 88.2, hence the unlocked word clock, because the incoming clock rate is a mismatch (44.1 in that case). Lynx states that because the HDX box is being told to operate at 88.1 via HDX, and because its incoming word clock is out of range, it has switched to Internal clock mode.

It's a no-win scenario for me at this point and each company is saying "the other company's unit needs to handle clock differently. Prism says that Lynx needs to use Loop Sync and Lynx is saying that the Titan needs to send the SR of the session, not base clock."

That would be an understandable explanation, EXCEPT THAT IT WORKED PROPERLY FOR A WEEK!?

Something about unplugging and reconnecting that word clock cable seems to have changed how the Titan operates, but Prism refuses to accept this, stating that they only ever send base rate loop sync.

Does anyone have any advice here? At the moment there's no acceptable scenario with both interfaces in the chain, and I seem to have reached a dead end with the advice of the two companies' tech support.

[Darryl Ramm]

Loop sync is a word clock signal, but by definition uses single speed aka base rate clocking. So Prism would be 100% correct here... "Prism states that they only use the base rate for loop sync". Except what do they mean by Loop Sync? The clock input and output they put on their Titan is labeled Word Clock not Loop Sync, so I open the Prism Titan manual...

Wordclock output

As well as outputting a clock at the selected sample rate, Titan's Wordclock output can be configured to produce a '256x clock' (a clock at 256x the selected sample rate, e.g. a 'superclock') or a 'Baseclock' (44.1kHz if the sample rate is a 44.1kHz multiple, or 48kHz if the sample rate is a 48kHz multiple).

So seems you can do everything you need if you clock from the Titan WC. I don't know why you need a loop (Pro Tools breaks the loop ahead of the master anyhow) Am I missing something, does something not work if you just connect WC source to destination (being PC about naming)?

Or you can sell that beautiful Titan to me, I'll offer you a great deal at $500

If you have different needs, you can get a Mutec clock box like a MC3+ and it can apply any multiplication to a WC signal, distribute multiple different signals, etc. (but IIRC only does multiply not divide).

[Top Jimmy]

The Titan manual says it can output actual rate, super clock, or base clock via word. Did this somehow get switched? It looks like you have to run the software to toggle this setting.

Also, the word inputs on both these units are terminated meaning neither of these units are built to do loop sync as Avid designed it so it’s useless to configure it this way. Just use the one cable and send word clock one direction.


Sent from my iPhone using Tapatalk

[JFreak]

Avid design is this:

- Loop Sync is always 1x Word Clock, regardless of the session sampling rate.
- Ext.Clock is representing session sampling rate, unless:
- In the olden days of Legacy Port for MIX hardware, Ext.Clock would turn into 256x SuperClock when Legacy Port is activated.

3rd party DigiLink boxes still don't always work like this and require at least one Avid box or a proper clock master to spread the sync.

[Eric Lambert]

Thanks guys! I have peaked through the manual and seen those sections, and because the Prism appeared to send session rate during the first week of use I'm optimistic that it can again. It's Prism's insistence that, while running in HDX mode for PT, the Titan sticks to 44.1 or 48K, and the Titan's apparent performance at the moment which seems to echo that statement, that has me scratching my head.

Yes, when running the Titan outside of HDX, via the Prism app, I'm able to select the session rate and everything is great. But I don't use PT in that way, I use it via HDX, and that's where the odd limitations begin.

I have played with word clock cables -- removing an input here and an output there -- but it's possible that I've not completely removed the cable leading from the Aurora's out to the Titan's in. I'll play some more.

[Eric Lambert]

Originally Posted by Darryl Ramm

Or you can sell that beautiful Titan to me, I'll offer you a great deal at $500

As a down payment?

[Eric Lambert]

Originally Posted by Top Jimmy

Also, the word inputs on both these units are terminated

I wonder if this has something to do with the weird anomaly where all sample rates functioned normally for a week until I disconnected a cable. This Aurora replaced an Avid HD IO, which was also slaved to the Titan, so I'm wondering if the Titan had configured itself to send the session rate (explaining why the Aurora could lock to WC at 88.2K and above), then unplugging the word clock cable changed its mode.

This was the theory I'd presented to Prism but they didn't bite. Lynx actually thought it was precisely the culprit and repeatedly asked why the Titan couldn't send the SR of the session.

[JFreak]

Sometimes you need to count seconds how long it takes for a box to provide stable clock. This was a huge problem in MIX era, literally you needed a stopwatch after power outage

[Top Jimmy]

Originally Posted by Eric Lambert

I wonder if this has something to do with the weird anomaly where all sample rates functioned normally for a week until I disconnected a cable. This Aurora replaced an Avid HD IO, which was also slaved to the Titan, so I'm wondering if the Titan had configured itself to send the session rate (explaining why the Aurora could lock to WC at 88.2K and above), then unplugging the word clock cable changed its mode.

This was the theory I'd presented to Prism but they didn't bite. Lynx actually thought it was precisely the culprit and repeatedly asked why the Titan couldn't send the SR of the session.

It puzzles me that Prism would assume that they should only support base clock when running HDX. It'd make sense if they supported true loop sync I/O, but they don't. I have a RED 8Pre that I use via HD Native. It has both word I/O and loop sync I/O. Loop sync jacks are on a relay. When a device is the loop master, the relay position sources the output from the internal clock and the input is terminated. When a device is a loop slave, the relay position parallels the input and output and sends the input to the PLL circuit unterminated. If the Titan doesn't have this functionality, then Prism should not lock it to base clock when used via the HD card.

I can still use REDNet Control for most functions when using the RED 8Pre with Pro Tools HD Native. Maybe the Titan software does the same and you can change the clock rate setting.

[Eric Lambert]

It's definitely puzzling.

After rereading all of Prism's and Lynx's responses the summary is this: the Titan only puts out loop sync. The Aurora is expecting only word clock.

Each company thinks that the others' design choices are odd, and both are beginning to recommend 3rd party clocking like a Big Ben. Geez, I sold my BB 15 years ago -- seems odd not to use these great modern clocks instead. Maybe I need clock distribution, like that Mutec, but spending another $1k when I shouldn't need to would bug me.

Or maybe I should just sell the Titan which seems to be the stubborn component here. Beautiful sounding conversion but man oh man I have their tech support dept on speed dial.