MTC receive using MBox 3 with PT 12

[BScout]

Originally Posted by Top Jimmy

Syncing the two audio interfaces via word or S/PDIF is the only thing that will keep the two timelines from drifting as MTC is not sample precise.

Word clock via word clock or carrier along AES specs (SPDIF) has nothing to do with timelines or drift. It is only digital clocking.
MTC is not sample accurate because it take 4 serial messages for complete timecode info to transfer and there is inherent delay in sending/receiving.

Word clock is digital clock accurate (no positional info in word clock).

Two completely different issues used for two different reasons.

Only way to get sample accurate lockup is via two Sync HD with genlock and Satellite but I think the OP realizes these limitations.

[Top Jimmy]

Originally Posted by BScout

Word clock via word clock or carrier along AES specs (SPDIF) has nothing to do with timelines or drift. It is only digital clocking.

Not clock syncing has everything to do with MTC drift. If the two digital clocks are synchronized then there won't be any. Kinda self explanatory don't you think?

[BScout]

Originally Posted by Top Jimmy

Not clock syncing has everything to do with MTC drift. If the two digital clocks are synchronized then there won't be any. Kinda self explanatory don't you think?

No. This is incorrect. MIDI Timecode is a serial communication via quarter frame messages.

Digital clocking is for keeping digital signal clocked via a series of pulses.

MTC can exist without any relation to digital clocking (in fact it preceded digital.) There is no connection between the two. And you can have digital systems clocked by wordclock and still have MTC drift (in fact you will.)

https://en.wikipedia.org/wiki/MIDI_timecode
https://en.wikipedia.org/wiki/Word_clock

[Top Jimmy]

Originally Posted by BScout

No. This is incorrect. MIDI Timecode is a serial communication via quarter frame messages.

Digital clocking is for keeping digital signal clocked via a series of pulses.

MTC can exist without any relation to digital clocking (in fact it preceded digital.) There is no connection between the two. And you can have digital systems clocked by wordclock and still have MTC drift (in fact you will.)

https://en.wikipedia.org/wiki/MIDI_timecode
https://en.wikipedia.org/wiki/Word_clock

In your race to be right you toss the baby out with the bathwater because you fail to consider how Pro Tools operates. We're not talking about using Sync HDs or any hardware based sync. We're talking about how Pro Tools uses MTC and generates its timecode reference.

In Pro Tools, the sample clock is king. It is the only absolute timing reference, all others (min:sec, timecode, feet+frames, bars|beats) are relative to sample time. This means that if you have identical session settings in two different sessions on two different computers and you sync one interface clock to another digitally, then they will run in perfect sync as long as you can figure out a way to start them together at exactly the same place and time. No drift. Why? Because at a sampling rate of 48K, Pro Tools needs exactly 48,000 samples to make 1 second.

This does not vary from one install of Pro Tools to another, nor does it vary from one computer to another. And since timecode is derived relative to sample time, it too will not drift in this scenario. On the occasion when there is MTC drift due to independent clocking, Pro Tools cannot speed up or slow down the sample clock so it instead will "hiccup" or skip small sections of time in an effort to maintain MTC lock, but if the sample clocks are exactly the same, Pro Tools should never have the need to hiccup.

The trick though is starting them together in time. When you are using MTC to sync two Pro Tools systems, there is a sizable margin of slop within which the MTC slave can successfully lock up. This means that the first time you hit play it may lock up with an additional 250 sample margin, while the next time you hit play, it could have an additional 400 sample margin. But again, this will never drift while playing as that margin will always be the same from beginning to end. Hence the variability I mentioned in my first post.

So realistically speaking, you can get one PT system to chase another via MTC without drift as long as they share the same sample clock reference. The only thing that could play havoc with this is something that interrupts or messes with the MTC transmission timing, hence the common practice of sending MTC only on its own port. The one thing I don't have any specific knowledge about is how much margin of error Pro Tools will tolerate before it attempts to fix the time by skipping. Since MTC takes 2 frames to transmit a single time message, this may be rather large. And since this margin of error will always be changing every two frames, it's largely ignorable as long as it never crosses the correction threshold. If this is of concern, one could always use jam sync since one only needs to trigger the transport when clock synced.

The one thing I can say from personal experience though is that the few times I have used MTC and clock sync between two Pro Tools systems, I haven't had any drifting issues.

[BScout]

I think we are getting nowhere on this. You have some things right (timecode is derived from samples as per an embedded number calculated from 0 and then derived by sample rate) but you get things wrong about how ProTools receives/transmits positional time. Sample pulses (word clock) has no bearing for positional time. ProTools does not “count” samples from transport start. It uses positional protocols. This is why we use genlock(Black burst/trisync), ltc, biphase, etc. And MTC. MTC transmission has its inherent delays (interface, drivers, etc).

So I’m going to appeal to reason: why did Avid introduce Satellite if all it took was MTC and wordclock for sample lock accuracy? And Avid are very clear that is the only way to get sample rate accuracy (2x Sync, ltc, genlock)

I’ll also appeal to your sense of experiment: two ProTools system sending only analogue signal between them (no digital) and linked by MTC and wordclock. Drop into record 4-5 times sending a serious of pulses cut sample accurately on the frame. Then do it again with the wordclock disconnected. They should differ greatly per your hypothesis: wordclock perfectly in sync at the sample level and without, random drifting.

Or talk to anyone in the Avid post-production client interfacing side.

One note for the OP (who probably exited out the side door awhile back) MTC cannot send 23.976 FPS (Or as Apple refers to it as 23.98). That came after the standards were set for MTC. Sending 23.976 via ProTools is actually 24 MTC transmission but that leaves you with frame drift depending on where your pkayhead starts. The easiest solution is to set both sides to 29.97 for the main counter/session and ignore any burn-in. 29.97 is in the MTC standard and due to the 5:4 situation between the two, you can flip it back to 23.976 after and be accurate (with no frame drift; not speaking samples accurate which requires Satellite). Just make sure your session and picture start on a second (not between seconds on a frame) to make flipping framerates seamless between those two. 23.976 is so common that this could be an issue for a post house suddenly using MTC.

[Top Jimmy]

Originally Posted by BScout

So I’m going to appeal to reason: why did Avid introduce Satellite if all it took was MTC and wordclock for sample lock accuracy? And Avid are very clear that is the only way to get sample rate accuracy (2x Sync, ltc, genlock)



I’ll also appeal to your sense of experiment: two ProTools system sending only analogue signal between them (no digital) and linked by MTC and wordclock. Drop into record 4-5 times sending a serious of pulses cut sample accurately on the frame. Then do it again with the wordclock disconnected. They should differ greatly per your hypothesis: wordclock perfectly in sync at the sample level and without, random drifting.

Where did I make the claim that it would or even could be either frame or sample accurate? I only said it wouldn't drift. Which is all the OP seems to need since he's just wanting two PT systems on an Mbox3 Pro and a Tascam 2x2 to run together.



Sent from my iPad using Tapatalk

[its2loud]

You’re running HD on both machines. You have a Sync HD. Put them both on the same network. Drop all this midi TC and LTC nonsense and enable satellite link on both machines. It became free and included in PT 11. Make one the master(administrator) and one the slave.

Read about it here

http://akmedia.digidesign.com/suppor..._v10_73571.pdf

[KingTor]

About to test S/PDIF. Will test word clock when I get both rigs on to the mix stage.

[KingTor]

…almost an hour into the S/PDIF test and that **** is locked. Sync source FTW!

[Top Jimmy]

Originally Posted by KingTor

…almost an hour into the S/PDIF test and that **** is locked. Sync source FTW!

Just so there isn't any misunderstanding here, this method isn't really actively synchronizing. One machine is not really chasing the other, other than being triggered to start and stop. They stay in sync because each system is playing the exact same number of samples in the exact same amount of time when the digital clocks are synced, and Pro Tools on the receiving machine is ignoring whatever margin of error there is between the received MTC and its own timeline. So as far as synchronization goes, this isn't by design but by happenstance of the way Pro Tools works.


Sent from my iPad using Tapatalk