Word Clock

[HapHazzard]

Have not tried this yet but will be soon.

There will be a second machine (not mine) attaching to my system strictly for recording. In the past I would fire up SAW or drag out my splitter and have it but now I'm thinking if their machine can take it I could provide the ADAT feeds coming from the per-amps. My Aphex 188 has redundant ADAT outs.

Q: Do I have to send my WC to the second machine.

Hap

[RBIngraham]

Depends. There has to be one master clock in all digital systems. So what is your wordclock master right now?

[Bob L]

If you do a digital signal split, you have to have one master clock... you can send it from your machine with WordClock if you want.

If you do an analog signal split, the machines can run clock independent with no issue.

Bob L

[HapHazzard]

It will be a digital split.

My SAC host will be the master coming from the RME card.

I will have to insure that the 2nd machine can except a WC input.

Thanks

[Sean McCoy]

The second machine can derive its clock from the incoming ADAT signal, but it's generally accepted that ADAT clock is less reliable than Word Clock, so I'd suggest the WC option if it's available.

[PhaseShifter]

The second machine can derive its clock from the incoming ADAT signal, but it's generally accepted that ADAT clock is less reliable than Word Clock, so I'd suggest the WC option if it's available.

I've seen this posted a lot, but have yet to see anyone prove that clocking via ADAT is less reliable.

[Grekim]

I'm not sure what the downstream computer is supposed to be doing? But, if it's just recording, then jitter is not at all a factor unless it was so severe as to be off by a half sample or something. The only time jitter comes into play is if you are capturing analog and converting to digital (or the reverse, going digital to analog) and relying on a clock to do that. In the OP's case it has already been converted.

Yes, there were rumors that W/C is lower on jitter than ADAT. And even that you shouldn't try to clock and send audio signal through the same pipe. I don't know if any of it is true. The downstream interface receiving the clock would be the big factor I believe, not so much whether it gets there by W/C or ADAT.

[RBIngraham]

I've seen this posted a lot, but have yet to see anyone prove that clocking via ADAT is less reliable.

What kind of proof would you like?

I find it depends a lot on the manufacturer of the gear. For example I can easilly get many Yamaha digital consoles to act up when you try to have them lock to an incoming ADAT optical. But they send out the ADAT clock just fine.

It's weird **** like that, for me, which makes it worth the couple seconds of extra effort to just run a BNC and then I never have problems.

[RBIngraham]

It will be a digital split.

My SAC host will be the master coming from the RME card.

I will have to insure that the 2nd machine can except a WC input.

Thanks

When you say the split will be digital, are you doing it at the preamps? Those preamps have two digital outputs, right? Or am I mistaken? If that is how you are doing the split, I might just make one of the Aphex the master clock. Use wordclock to lock the other Aphex units to themaster unit. Then I would just try locking both the mixing computer and recording computer to the preamps via the ADAT opticals.

That is what I would try first anyway.

[HapHazzard]

When you say the split will be digital, are you doing it at the preamps? Those preamps have two digital outputs, right? Or am I mistaken? If that is how you are doing the split, I might just make one of the Aphex the master clock. Use wordclock to lock the other Aphex units to themaster unit. Then I would just try locking both the mixing computer and recording computer to the preamps via the ADAT opticals.

That is what I would try first anyway.Will give a try.

[DominicPerry]

Try this if you want to know more...

It's a paper describing the accuracy of a selection of digital transmission formats.

http://www.tcelectronic.com/media/1018473/frandsen_lave_2004_plug_play_aes116.pdf

Dominic

[Carl G.]

Try this if you want to know more...

It's a paper describing the accuracy of a selection of digital transmission formats.

http://www.tcelectronic.com/media/1018473/frandsen_lave_2004_plug_play_aes116.pdf

Dominic
After reading that abstract... it seems John Flemming had it right the first time. :)

[cgrafx]

Try this if you want to know more...

It's a paper describing the accuracy of a selection of digital transmission formats.

http://www.tcelectronic.com/media/1018473/frandsen_lave_2004_plug_play_aes116.pdf

Dominic

Its a paper from 2004 and the conclusion of the paper basically says if its digital, it might work or it might not. We don't really know why and we don't have any suggestions about how to make it better.

As for me, I've got two rock solid systems connected over ADAT and have had zero clocking issues.

I'd always start with the simpler solution. If ADAT clock works, great use it. If for some reason its not working with the equipment you have, than try a separate word clock, just know that its adding additional complexity to the system.

[RBIngraham]

Its a paper from 2004 and the conclusion of the paper basically says if its digital, it might work or it might not. We don't really know why and we don't have any suggestions about how to make it better.

As for me, I've got two rock solid systems connected over ADAT and have had zero clocking issues.

I'd always start with the simpler solution. If ADAT clock works, great use it. If for some reason its not working with the equipment you have, than try a separate word clock, just know that its adding additional complexity to the system.

I kind of agree, but when you get to larger systems where you have 2 or 3 or more computers and consoles and god knows what else all talking to each other.... trust me, running a dedicacted word clock generator and making everthing lock to it, actually ends up being the simple route rather than some complex spider web of adat clocks or the like.

[Naturally Digital]

I've seen this posted a lot, but have yet to see anyone prove that clocking via ADAT is less reliable.I've had problems here at the studio trying to get a LynxTwo card (with ADAT exp) to sync to incoming ADAT from an RME card. That proved it for me. :)

I will add this... 9 times out of 10, when I have problems with ADAT connections, it turns out to be a problematic cable.

[cgrafx]

I kind of agree, but when you get to larger systems where you have 2 or 3 or more computers and consoles and god knows what else all talking to each other.... trust me, running a dedicacted word clock generator and making everthing lock to it, actually ends up being the simple route rather than some complex spider web of adat clocks or the like.

In multi-connection configurations the simpler setup is going to be a dedicated word clock. Actually in some of those configurations a dedicated word clock is the only sane option.

Just saying that there isn't anything inherently wrong/poor about ADAT clocking (even though there are instances where it doesn't work).

[RBIngraham]

In multi-connection configurations the simpler setup is going to be a dedicated word clock. Actually in some of those configurations a dedicated word clock is the only sane option.

Just saying that there isn't anything inherently wrong/poor about ADAT clocking (even though there are instances where it doesn't work).

Agreed. There just seems to be more instances where ADAT clocking is poorly implemented than most other digital interconnects. No idea why really, just my experience.

[DominicPerry]

I agree that the paper is out of date, but some of the interfaces they tested were from 1997 - seven years before the paper was written. The standards have been around for many years, but that doesn't mean devices are adhering to them.

I think it's worth bearing in mind that there are several things in play. ADAT, as a digital standard, should work just fine as a sync method. In the same way, unbalanced audio at -10dB over a short run, connected with 3.5mm mini jacks, can and often does work as well as balanced audio at +4dB connected with locking XLRs. But the combination of interference, lossy cables and unreliable connectors make the latter preferable to the former.

An ADAT connection should be fine, but sometimes it isn't. The connectors are fiddly, the cables can be low grade. A good locking BNC, dedicated to the job of wordclock is a more reassuring approach. It won't sound 'better' than ADAT if the ADAT is working properly, but it's more likely to work.

Dominic

[Carlos Mills]

I kind of agree, but when you get to larger systems where you have 2 or 3 or more computers and consoles and god knows what else all talking to each other.... trust me, running a dedicacted word clock generator and making everthing lock to it, actually ends up being the simple route rather than some complex spider web of adat clocks or the like.

(y)