Sign in with
Sign up | Sign in
Your question

Wordclock&SMPTE in Composer/Digital system

Last response: in Home Audio
Share
Anonymous
February 4, 2005 10:38:56 PM

Archived from groups: rec.audio.pro (More info?)

I know this has been asked in different ways over the past, but...

A TV/Film composer friend is in the midst of some equipment upgrades.

He wants to start routing most of his audio digitally. He composes mostly
with standalone samplers and some tracks of audio.

His samplers have digital outs and can be fed with wordclock.
He is using Cubase (current version) for sequencing the samplers, plus
he is tracking his audio in Cubase. Cubase is on a new Mac G5 with
an RME interface. Midi is by MOTU MTP AV (USB).

He will not track individual sampler outputs to Cubase, but instead feed them
to his Mackie D8b. These tracks will be blended with individual audio tracks
coming out of Cubase in the D8b. Remember, everything is happening in the
digital domain.

The output of the D8b will go a couple of places: DA88 and back to the
Cubase/RME audio ins. He provides music to producers on DA88 and Protools
files.

He works from VHS tapes! SMPTE from VHS goes to the MTP, and cubase will
slave to that. No Blackburst. When he is working on parts, will switch
cubase to master and not bother seeing tape as he works.
(Yes, one day he may capture video to his Mac and work from it there...but
he likes to move carefully).

So...how to distribute the clock?

1)Does WC have to be integrated with SMPTE? The MTP does output WC.
Should he use that as his master clock?

2)Or can SMPTE and WC freewheel against each
other? If the MTP is not the master, what should be? D8b? Cubase?
And external WC generator?

3)Lastly---WC distribution. He is running about 8 samplers in addition to
the mixer, audio interface, DA88 etc. Does he need to send the WC master
to some sort of WC distribution box to avoid all the chaining? Or is
there a way around?

Thanks in advance for any help,

Rob R.
Anonymous
February 4, 2005 10:38:57 PM

Archived from groups: rec.audio.pro (More info?)

Rob--

Wow. That's a complicated setup. I've scored from VHS work tapes or
"chase copies" for years. Here's what I've done with my D8B/Logic/VHS
master rig:

I have a video sync/black burst house clock. It sets the speed for my
word clock. I've done it 2 ways, and it's worked both ways:

1. video sync to DIGI video slave driver, which converts video to word,
then word gets distributed via Aardvark word clock DA.

2. MOTU 2408 Mk III accepts word, video, or linear SMPTE and converts
it to word clock, word clock distributed via Aardvark DA.

The reason that video sync is the first clock in the chain, is because
of the video deck. What most people don't know is that almost every
consumer and semi-pro video deck out there locks to the video signal
that comes into the video input. I lock my Mitsubishi S-VHS machine
(and have done this with Sony, Panasonic & JVC also) to video sync by
plugging a video sync signal into the video input, and also sending it
to the video slave driver or MOTU 2408 Mk III. The SMPTE from one of
the audio channels of the VHS machine goes to my midi interface (an
Emagic Unitor) for locator purposes for Logic & the D8B.

In theory, if you could find a word clock device that will derive video
sync from incoming word clock, you could work with a word clock master,
and have the video lock to the video sync output.

Over the last year, I've pretty much taken to just digitizing the
video, opening it in logic, and scoring to that. My VHS machine is
getting pretty dusty, and I post approvals on the internet, instead of
shipping VHS dubs via FedEx or courier. That's eliminated the need for
a video sync master signal; although I still have the setup if I need
it. I now mostly work from the video slave driver as the word clock
master, distributed by the Aardvard DA, and no video sync is involved.

I don't have any of my samplers set up via digital output, so I don't
have any advice about that. I'm moving very swiftly to software synths,
which negate the need for outboard synths more and more all the time.

Now... to answer your questions in order:

1. Word clock has to be integrated with video sync clock. SMPTE is
another issue, having to do with where you are (mostly) as opposed to
how fast the frames go by (mostly). SMPTE will tell the MTP what to
translate to Midi Time Code, and from there, communicate to the D8B and
Cubase where you are. The MTP can be the "God clock," if all the gear
agrees with it. Some clocks are more unstable than others. I don't know
about the stability of the MTP's clock.

The D8B must have the Apogee Clock I/O card to lock to anything outside
itself, and most people have said that it improves the quality of a
free-standing D8B too.

2. SMPTE and WC cannot freewheel if you expect sync. A lot of people
get a master word clock generator to be the "God clock." Look around,
there are a lot of recommendations around. One will be right for you.

3. WC distribution with a WC DA is the best way to go. No messing with
termination, induced delay, grounding, etc. etc. etc. Yes, you should
send the WC master to a Distribution Amplifier, or get a model that has
the WC master and several outputs in the same box. I haven't seen one
with 8 outputs, though.

As an alternative on a box with 5 or 6 outputs (not your required 8),
you can experiment (if you have the time) with putting BNC "T's" to
distribute *a few* of the word clock DA's outputs, but you're messing
with loads and impedances and some devices don't like that, and won't
lock reliably.

Happy scoring.
Anonymous
February 4, 2005 10:38:57 PM

Archived from groups: rec.audio.pro (More info?)

Rob--

Wow. That's a complicated setup. I've scored from VHS work tapes or
"chase copies" for years. Here's what I've done with my D8B/Logic/VHS
master rig:

I have a video sync/black burst house clock. It sets the speed for my
word clock. I've done it 2 ways, and it's worked both ways:

1. video sync to DIGI video slave driver, which converts video to word,
then word gets distributed via Aardvark word clock DA.

2. MOTU 2408 Mk III accepts word, video, or linear SMPTE and converts
it to word clock, word clock distributed via Aardvark DA.

The reason that video sync is the first clock in the chain, is because
of the video deck. What most people don't know is that almost every
consumer and semi-pro video deck out there locks to the video signal
that comes into the video input. I lock my Mitsubishi S-VHS machine
(and have done this with Sony, Panasonic & JVC also) to video sync by
plugging a video sync signal into the video input, and also sending it
to the video slave driver or MOTU 2408 Mk III. The SMPTE from one of
the audio channels of the VHS machine goes to my midi interface (an
Emagic Unitor) for locator purposes for Logic & the D8B.

In theory, if you could find a word clock device that will derive video
sync from incoming word clock, you could work with a word clock master,
and have the video lock to the video sync output.

Over the last year, I've pretty much taken to just digitizing the
video, opening it in logic, and scoring to that. My VHS machine is
getting pretty dusty, and I post approvals on the internet, instead of
shipping VHS dubs via FedEx or courier. That's eliminated the need for
a video sync master signal; although I still have the setup if I need
it. I now mostly work from the video slave driver as the word clock
master, distributed by the Aardvard DA, and no video sync is involved.

I don't have any of my samplers set up via digital output, so I don't
have any advice about that. I'm moving very swiftly to software synths,
which negate the need for outboard synths more and more all the time.

Now... to answer your questions in order:

1. Word clock has to be integrated with video sync clock. SMPTE is
another issue, having to do with where you are (mostly) as opposed to
how fast the frames go by (mostly). SMPTE will tell the MTP what to
translate to Midi Time Code, and from there, communicate to the D8B and
Cubase where you are. The MTP can be the "God clock," if all the gear
agrees with it. Some clocks are more unstable than others. I don't know
about the stability of the MTP's clock.

The D8B must have the Apogee Clock I/O card to lock to anything outside
itself, and most people have said that it improves the quality of a
free-standing D8B too.

2. SMPTE and WC cannot freewheel if you expect sync. A lot of people
get a master word clock generator to be the "God clock." Look around,
there are a lot of recommendations around. One will be right for you.

3. WC distribution with a WC DA is the best way to go. No messing with
termination, induced delay, grounding, etc. etc. etc. Yes, you should
send the WC master to a Distribution Amplifier, or get a model that has
the WC master and several outputs in the same box. I haven't seen one
with 8 outputs, though.

As an alternative on a box with 5 or 6 outputs (not your required 8),
you can experiment (if you have the time) with putting BNC "T's" to
distribute *a few* of the word clock DA's outputs, but you're messing
with loads and impedances and some devices don't like that, and won't
lock reliably.

Happy scoring.
Anonymous
February 4, 2005 10:38:57 PM

Archived from groups: rec.audio.pro (More info?)

Well, I can only tell you the ideal situations which you probably can't
do....but the best setup would be to have a master word clock generator
that also outputs video sync and then feed word clock through a
distribution amp and video sync to the video sources.

Now the fact that the VHS can't slave to video sync is problematic. Of
course, you can just make the VHS deck the master, but almost all VHS
decks (especially those that can't slave to external video sync) have
horrific time bases. I'm guessing the guy doesn't have a TBC either to
refresh the video and make the sync nice and pretty again, huh?

To answer your questions directly:

1) WC does have to be integrated with SMPTE in the sense that most NLE
will lock to a SMPTE location and then run on word clock which (after
awhile) will cause it to drift apart from the SMPTE output if the WC is
not matched to the SMPTE. Since the VHS is probably going to play back
at some random speed every time, this really isn't an option. The only
option is to have your NLE clock off the SMPTE, which is not ideal for
analog audio quality. But you really don't have another choice.

2) Your master is the VHS deck, unfortunately, since it can't slave to
anything. Your MTP should receive the SMPTE and then create WC based on
the SMPTE input. I'm not an expert on the MTP, but with the Digi Sync
I/O it's just a matter of setting the WC to be generated from the SMPTE
input.

3) Ideally, WC is distributed through a DA. Whether or not daisy
chaining will work depends on the devices and your specific situation
which I can't really guess about.

As long as you are sending audio back and forth, the stability of your
WC against SMPTE is not as big an issue because as long as the samples
get sent over correctly, they will be reclocked correctly once you get
off the SMPTE clock and get back to the MTP's internal clock (or
anything else decent).

So really, digital transfers make the most sense in this situation.
Just hope that the SMPTE output isn't so bad from the VHS that the MTP
just gives up trying to generate WC from it. The Beta decks around here
almost always give up when trying to sync to a VHS without sending the
VHS through a TBC.

Cheers,
Trevor de Clercq

Rob Reedijk wrote:
> I know this has been asked in different ways over the past, but...
>
> A TV/Film composer friend is in the midst of some equipment upgrades.
>
> He wants to start routing most of his audio digitally. He composes mostly
> with standalone samplers and some tracks of audio.
>
> His samplers have digital outs and can be fed with wordclock.
> He is using Cubase (current version) for sequencing the samplers, plus
> he is tracking his audio in Cubase. Cubase is on a new Mac G5 with
> an RME interface. Midi is by MOTU MTP AV (USB).
>
> He will not track individual sampler outputs to Cubase, but instead feed them
> to his Mackie D8b. These tracks will be blended with individual audio tracks
> coming out of Cubase in the D8b. Remember, everything is happening in the
> digital domain.
>
> The output of the D8b will go a couple of places: DA88 and back to the
> Cubase/RME audio ins. He provides music to producers on DA88 and Protools
> files.
>
> He works from VHS tapes! SMPTE from VHS goes to the MTP, and cubase will
> slave to that. No Blackburst. When he is working on parts, will switch
> cubase to master and not bother seeing tape as he works.
> (Yes, one day he may capture video to his Mac and work from it there...but
> he likes to move carefully).
>
> So...how to distribute the clock?
>
> 1)Does WC have to be integrated with SMPTE? The MTP does output WC.
> Should he use that as his master clock?
>
> 2)Or can SMPTE and WC freewheel against each
> other? If the MTP is not the master, what should be? D8b? Cubase?
> And external WC generator?
>
> 3)Lastly---WC distribution. He is running about 8 samplers in addition to
> the mixer, audio interface, DA88 etc. Does he need to send the WC master
> to some sort of WC distribution box to avoid all the chaining? Or is
> there a way around?
>
> Thanks in advance for any help,
>
> Rob R.
Anonymous
February 4, 2005 10:38:57 PM

Archived from groups: rec.audio.pro (More info?)

Trevor -

Yes, I've been able to sync with even 2-head VHS decks to video input
on playback. Video syncing to input on playback was undocumented in all
but my Panasonic deck which also had a "sync" input, but actually
synced more reliably to the video input.

This was told to me in the late 80's by a video engineer, and has
worked reliably on all the machines I've owned. I've beat them up
pretty bad, and I always needed ones with a nice frame-by-frame wheel
remote, so I've been through a number of them. I agree though that you
should always check and be sure it will do it first. I would sync with
a 1 hour program as a test. You should be able to tell within about 5
to 10 minutes though if it isn't syncing.

The biggest problem that I've had was that some video houses have no
idea what the audio levels are going to their video machines (even on
masters sometimes!), and I would sometimes get a linear time code track
that was either way hammered, or barely there. Usually a visit with the
engineer would smooth things over and get me a clean copy. I always
request burn-in too, just for confidence.
Anonymous
February 4, 2005 10:38:58 PM

Archived from groups: rec.audio.pro (More info?)

This is interesting. I know consumer video decks will sync off the
video input when recording, but you're saying that some junky old VHS
deck will sync to the video input when playing back? I'm guessing that
wouldn't be universally true and would depend on a deck to deck basis.
I would be leery about that unless I tested the actual machine I was
working with.

Anyway, if the deck does sync to the video input on playback, what
you've said seems like good advice. It sure would make things a lot neater.

Cheers,
Trevor de Clercq

singproinc wrote:
> Rob--
>
> Wow. That's a complicated setup. I've scored from VHS work tapes or
> "chase copies" for years. Here's what I've done with my D8B/Logic/VHS
> master rig:
>
> I have a video sync/black burst house clock. It sets the speed for my
> word clock. I've done it 2 ways, and it's worked both ways:
>
> 1. video sync to DIGI video slave driver, which converts video to word,
> then word gets distributed via Aardvark word clock DA.
>
> 2. MOTU 2408 Mk III accepts word, video, or linear SMPTE and converts
> it to word clock, word clock distributed via Aardvark DA.
>
> The reason that video sync is the first clock in the chain, is because
> of the video deck. What most people don't know is that almost every
> consumer and semi-pro video deck out there locks to the video signal
> that comes into the video input. I lock my Mitsubishi S-VHS machine
> (and have done this with Sony, Panasonic & JVC also) to video sync by
> plugging a video sync signal into the video input, and also sending it
> to the video slave driver or MOTU 2408 Mk III. The SMPTE from one of
> the audio channels of the VHS machine goes to my midi interface (an
> Emagic Unitor) for locator purposes for Logic & the D8B.
>
> In theory, if you could find a word clock device that will derive video
> sync from incoming word clock, you could work with a word clock master,
> and have the video lock to the video sync output.
>
> Over the last year, I've pretty much taken to just digitizing the
> video, opening it in logic, and scoring to that. My VHS machine is
> getting pretty dusty, and I post approvals on the internet, instead of
> shipping VHS dubs via FedEx or courier. That's eliminated the need for
> a video sync master signal; although I still have the setup if I need
> it. I now mostly work from the video slave driver as the word clock
> master, distributed by the Aardvard DA, and no video sync is involved.
>
> I don't have any of my samplers set up via digital output, so I don't
> have any advice about that. I'm moving very swiftly to software synths,
> which negate the need for outboard synths more and more all the time.
>
> Now... to answer your questions in order:
>
> 1. Word clock has to be integrated with video sync clock. SMPTE is
> another issue, having to do with where you are (mostly) as opposed to
> how fast the frames go by (mostly). SMPTE will tell the MTP what to
> translate to Midi Time Code, and from there, communicate to the D8B and
> Cubase where you are. The MTP can be the "God clock," if all the gear
> agrees with it. Some clocks are more unstable than others. I don't know
> about the stability of the MTP's clock.
>
> The D8B must have the Apogee Clock I/O card to lock to anything outside
> itself, and most people have said that it improves the quality of a
> free-standing D8B too.
>
> 2. SMPTE and WC cannot freewheel if you expect sync. A lot of people
> get a master word clock generator to be the "God clock." Look around,
> there are a lot of recommendations around. One will be right for you.
>
> 3. WC distribution with a WC DA is the best way to go. No messing with
> termination, induced delay, grounding, etc. etc. etc. Yes, you should
> send the WC master to a Distribution Amplifier, or get a model that has
> the WC master and several outputs in the same box. I haven't seen one
> with 8 outputs, though.
>
> As an alternative on a box with 5 or 6 outputs (not your required 8),
> you can experiment (if you have the time) with putting BNC "T's" to
> distribute *a few* of the word clock DA's outputs, but you're messing
> with loads and impedances and some devices don't like that, and won't
> lock reliably.
>
> Happy scoring.
>
Anonymous
February 4, 2005 10:38:58 PM

Archived from groups: rec.audio.pro (More info?)

Very good to know, thanks!

Yeah, you gotta love those VHS decks that don't even have reliable audio
meters, much less audio level adjustment controls.

Cheers,
Trevor de Clercq

singproinc wrote:
> Trevor -
>
> Yes, I've been able to sync with even 2-head VHS decks to video input
> on playback. Video syncing to input on playback was undocumented in all
> but my Panasonic deck which also had a "sync" input, but actually
> synced more reliably to the video input.
>
> This was told to me in the late 80's by a video engineer, and has
> worked reliably on all the machines I've owned. I've beat them up
> pretty bad, and I always needed ones with a nice frame-by-frame wheel
> remote, so I've been through a number of them. I agree though that you
> should always check and be sure it will do it first. I would sync with
> a 1 hour program as a test. You should be able to tell within about 5
> to 10 minutes though if it isn't syncing.
>
> The biggest problem that I've had was that some video houses have no
> idea what the audio levels are going to their video machines (even on
> masters sometimes!), and I would sometimes get a linear time code track
> that was either way hammered, or barely there. Usually a visit with the
> engineer would smooth things over and get me a clean copy. I always
> request burn-in too, just for confidence.
>
Anonymous
February 10, 2005 6:12:29 PM

Archived from groups: rec.audio.pro (More info?)

Bob and Trevor,

Thanks for your highly informative replies.

Rob R.

Rob Reedijk <reedijk@hera.med.utoronto.ca> wrote:

> I know this has been asked in different ways over the past, but...

> A TV/Film composer friend is in the midst of some equipment upgrades.

> He wants to start routing most of his audio digitally. He composes mostly
> with standalone samplers and some tracks of audio.

> His samplers have digital outs and can be fed with wordclock.
> He is using Cubase (current version) for sequencing the samplers, plus
> he is tracking his audio in Cubase. Cubase is on a new Mac G5 with
> an RME interface. Midi is by MOTU MTP AV (USB).

> He will not track individual sampler outputs to Cubase, but instead feed them
> to his Mackie D8b. These tracks will be blended with individual audio tracks
> coming out of Cubase in the D8b. Remember, everything is happening in the
> digital domain.

> The output of the D8b will go a couple of places: DA88 and back to the
> Cubase/RME audio ins. He provides music to producers on DA88 and Protools
> files.

> He works from VHS tapes! SMPTE from VHS goes to the MTP, and cubase will
> slave to that. No Blackburst. When he is working on parts, will switch
> cubase to master and not bother seeing tape as he works.
> (Yes, one day he may capture video to his Mac and work from it there...but
> he likes to move carefully).

> So...how to distribute the clock?

> 1)Does WC have to be integrated with SMPTE? The MTP does output WC.
> Should he use that as his master clock?

> 2)Or can SMPTE and WC freewheel against each
> other? If the MTP is not the master, what should be? D8b? Cubase?
> And external WC generator?

> 3)Lastly---WC distribution. He is running about 8 samplers in addition to
> the mixer, audio interface, DA88 etc. Does he need to send the WC master
> to some sort of WC distribution box to avoid all the chaining? Or is
> there a way around?

> Thanks in advance for any help,

> Rob R.
!