Hi everyone!
I'm in the process of transferring a bunch of songs from 2" 24
track tape to ProTools. I'm doing each song in two 12-channel passes
with ProTools chasing SMPTE on the unresolved 2" tape using a Universal
Slave Driver. In transferring the first song I was surprised to see
that the second 12 channel pass had drifted 49 samples off from the
first pass.
There were actually less than 24 total tracks, so I recorded the
kick drum track with both passes to ProTools in order to verify sync.
At the top of the song the two kick tracks in PT were 3 samples off
each other, but by the end they were 49 samples off.
I have the USD set to Internal/VSO for Clock Reference, LTC for
Positional Reference, with the correct Frame Rate and the USD solidly
locked to the SMPTE. Although musically 49 samples is close enough,
this still strikes me as strange. Could it be because I'm unresolved
to any type of House Black? I'd still expect sample accuracy since PT
is locked to the SMPTE on the source. Or should I perhaps set the
USD's Clock Reference to "LTC"? It didn't seem to want to lock that
way, probably because the 2" machine is unresolved.
Any insights? Am I doing something wrong or is this as close as
it's going to get, unresolved? I can't get my head around why Lack Of
Black would make a difference, although I suspect that's what the
problems is.
Thanks, Rick.

In this case I would want to set the USD's Clock Reference to "LTC" as
that's the sync master, i.e., the tape, the drifting tape, USD should
follow the drift and keep it locked.

its not the perfect way, but it should work.

The perfect way is as you said, both 2" and USD get a master house sync. for
this you need a synchroniser that is controlling the capstan of the 2",
like a adams smith or zeta 3 and so on.

the synchroniser looks at the smpte, compares it to the house sync and
speeds up or slows the 2" to be more correct.

in your case the 2" is free to do as it pleases, and the USD must follow it
and vary the WC or superclock or whatever it is in PT to follow along, this
has a negative side effect on the AD's quality. but all the same it can and
does work.

I use the MOTU HD192's ability to set any audio input to resolve to smpte.
and it works great as it uses ASIO positioning protocol to sample accurate
sync Nuendo, but its still one step down from a full sync situation where
everyone is locked to one clock.

Did I make any sence, its been a long day.


Adam Calaitzis
Toyland Recording Studio
www.toyland.com.au







"rickymix" wrote in message
ups.com...
Hi everyone!
I'm in the process of transferring a bunch of songs from 2" 24
track tape to ProTools. I'm doing each song in two 12-channel passes
with ProTools chasing SMPTE on the unresolved 2" tape using a Universal
Slave Driver. In transferring the first song I was surprised to see
that the second 12 channel pass had drifted 49 samples off from the
first pass.
There were actually less than 24 total tracks, so I recorded the
kick drum track with both passes to ProTools in order to verify sync.
At the top of the song the two kick tracks in PT were 3 samples off
each other, but by the end they were 49 samples off.
I have the USD set to Internal/VSO for Clock Reference, LTC for
Positional Reference, with the correct Frame Rate and the USD solidly
locked to the SMPTE. Although musically 49 samples is close enough,
this still strikes me as strange. Could it be because I'm unresolved
to any type of House Black? I'd still expect sample accuracy since PT
is locked to the SMPTE on the source. Or should I perhaps set the
USD's Clock Reference to "LTC"? It didn't seem to want to lock that
way, probably because the 2" machine is unresolved.
Any insights? Am I doing something wrong or is this as close as
it's going to get, unresolved? I can't get my head around why Lack Of
Black would make a difference, although I suspect that's what the
problems is.
Thanks, Rick.

rickymix wrote:
Hi everyone!
I'm in the process of transferring a bunch of songs from 2" 24
track tape to ProTools. I'm doing each song in two 12-channel passes
with ProTools chasing SMPTE on the unresolved 2" tape using a Universal
Slave Driver. In transferring the first song I was surprised to see
that the second 12 channel pass had drifted 49 samples off from the
first pass.

That's pretty good. Better than my Ampex can do, anyway.

There were actually less than 24 total tracks, so I recorded the
kick drum track with both passes to ProTools in order to verify sync.
At the top of the song the two kick tracks in PT were 3 samples off
each other, but by the end they were 49 samples off.

The thing is, the recording and playback speeds may differ. So you
can't really trust that. Remember, the machine is never going to run
at exactly the same speed twice in a row. It'll always very a little
bit.

I have the USD set to Internal/VSO for Clock Reference, LTC for
Positional Reference, with the correct Frame Rate and the USD solidly
locked to the SMPTE. Although musically 49 samples is close enough,
this still strikes me as strange. Could it be because I'm unresolved
to any type of House Black? I'd still expect sample accuracy since PT
is locked to the SMPTE on the source. Or should I perhaps set the
USD's Clock Reference to "LTC"? It didn't seem to want to lock that
way, probably because the 2" machine is unresolved.

You worry too much. Just make sure never to split stereo pairs between
the two passes.

Any insights? Am I doing something wrong or is this as close as
it's going to get, unresolved? I can't get my head around why Lack Of
Black would make a difference, although I suspect that's what the
problems is.

This is pretty good.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Adam wrote:
In this case I would want to set the USD's Clock Reference to "LTC" as
that's the sync master, i.e., the tape, the drifting tape, USD should

follow the drift and keep it locked.
.... the 2" is free to do as it pleases, and the USD must follow it
and vary the WC or superclock or whatever it is in PT to follow along,
this
has a negative side effect on the AD's quality. but all the same it
can and
does work.

Thanks Adam.
I think what you are referring to is what the USD calls
"Positional Reference", which I do have set to "LTC. The USD chases
the fluctuating SMPTE very well, it's a quick, solid lock. That's not
the problem.
The other separate selector which I'm confused about is called
"Clock Reference", with choices of Video, LTC, Digital, Pilot, BiPhase,
and Internal/VSO. Video would be House Black which I don't have here,
Digital would be from a digital source, Pilot and BiPhase also don't
apply, so the choices are LTC and Internal/VSO. It's not happy in LTC
mode, probably because SMPTE was not originally striped with any Black
reference, plus it's playing back unreferenced.
Which leaves Internal/VSO mode, which as Scott says is working
acceptably. I can't find my manual on the USD though, and I was hoping
someone who uses a USD more frequently than me could confirm that this
is indeed the correct setting.

Scott wrote:
the machine is never going to run at exactly
the same speed twice in a row. It'll always vary a little bit.

Hi Scott,
The machine varies a lot with each pass, way more than 49 samples.
But the SMPTE I'm locked to stays identical with each pass. That's
why I'm puzzled. As you say, it's working and I'm over-thinking this,
but still I'm curious as to why the drift? If it was a steady 49
samples from beginning to end it would be slightly more understandable
..
Anyways, thanks guys!
Cheers, Rick.

rickymix wrote:
Adam wrote:
In this case I would want to set the USD's Clock Reference to "LTC" as
that's the sync master, i.e., the tape, the drifting tape, USD should

follow the drift and keep it locked.
... the 2" is free to do as it pleases, and the USD must follow it
and vary the WC or superclock or whatever it is in PT to follow along,
this
has a negative side effect on the AD's quality. but all the same it
can and
does work.

Thanks Adam.
I think what you are referring to is what the USD calls
"Positional Reference", which I do have set to "LTC. The USD chases
the fluctuating SMPTE very well, it's a quick, solid lock. That's not
the problem.
The other separate selector which I'm confused about is called
"Clock Reference", with choices of Video, LTC, Digital, Pilot, BiPhase,
and Internal/VSO. Video would be House Black which I don't have here,
Digital would be from a digital source, Pilot and BiPhase also don't
apply, so the choices are LTC and Internal/VSO. It's not happy in LTC
mode, probably because SMPTE was not originally striped with any Black
reference, plus it's playing back unreferenced.
Which leaves Internal/VSO mode, which as Scott says is working
acceptably. I can't find my manual on the USD though, and I was hoping
someone who uses a USD more frequently than me could confirm that this
is indeed the correct setting.

Scott wrote:
the machine is never going to run at exactly
the same speed twice in a row. It'll always vary a little bit.

Hi Scott,
The machine varies a lot with each pass, way more than 49 samples.
But the SMPTE I'm locked to stays identical with each pass. That's
why I'm puzzled. As you say, it's working and I'm over-thinking this,
but still I'm curious as to why the drift? If it was a steady 49
samples from beginning to end it would be slightly more understandable
.
Anyways, thanks guys!
Cheers, Rick.

As Scott said, a 49 sample total drift is very good. Assuning your
sampling rate is 44.1 or 48K, 49 samples is only about 1ms. Halve that
amount if it's 88.2 or 96K. Many, many great records were made with 24
track 2' machines locked with a Lynx or equivalent, using a blackburst
generator as a sync master, and those setups came *nowhere near* that
kind of lockup accuracy or consistency. That was the state of the art
for quite a while. Aside from not splitting stereo tracks across two
passes of transfers in this setup, as others have mentioned, forget
about the insignificant "drift" and be happy. I guarantee you *no one*
will ever hear a problem.

Ted Spencer, NYC
tedspencerrecording.com

Thanks guys!
Again, it's not that I'm worried about this small amount of drift.
It just irks me that I don't understand why there's any drifting at
all. Probably comes from my 8 years of working at Technicolor, syncing
stuff up everyday. The problem with accepting a small amount of
unexplained error is that it leaves you vulnerable for more significant
amounts in the future.
I'll ask some of my old Technicolor buddies about this situation,
and let y'all know if I learn anything new.
Cheers, Rick.

In article ,
"Adam Calaitzis" wrote:

In this case I would want to set the USD's Clock Reference to "LTC" as
that's the sync master, i.e., the tape, the drifting tape, USD should
follow the drift and keep it locked.

its not the perfect way, but it should work.

The perfect way is as you said, both 2" and USD get a master house sync. for
this you need a synchroniser that is controlling the capstan of the 2",
like a adams smith or zeta 3 and so on.

the synchroniser looks at the smpte, compares it to the house sync and
speeds up or slows the 2" to be more correct.

in your case the 2" is free to do as it pleases, and the USD must follow it
and vary the WC or superclock or whatever it is in PT to follow along, this
has a negative side effect on the AD's quality. but all the same it can and
does work.

I use the MOTU HD192's ability to set any audio input to resolve to smpte.
and it works great as it uses ASIO positioning protocol to sample accurate
sync Nuendo, but its still one step down from a full sync situation where
everyone is locked to one clock.

Did I make any sence, its been a long day.


Adam Calaitzis
Toyland Recording Studio
www.toyland.com.au







"rickymix" wrote in message
ups.com...
Hi everyone!
I'm in the process of transferring a bunch of songs from 2" 24
track tape to ProTools. I'm doing each song in two 12-channel passes
with ProTools chasing SMPTE on the unresolved 2" tape using a Universal
Slave Driver. In transferring the first song I was surprised to see
that the second 12 channel pass had drifted 49 samples off from the
first pass.
There were actually less than 24 total tracks, so I recorded the
kick drum track with both passes to ProTools in order to verify sync.
At the top of the song the two kick tracks in PT were 3 samples off
each other, but by the end they were 49 samples off.
I have the USD set to Internal/VSO for Clock Reference, LTC for
Positional Reference, with the correct Frame Rate and the USD solidly
locked to the SMPTE. Although musically 49 samples is close enough,
this still strikes me as strange. Could it be because I'm unresolved
to any type of House Black? I'd still expect sample accuracy since PT
is locked to the SMPTE on the source. Or should I perhaps set the
USD's Clock Reference to "LTC"? It didn't seem to want to lock that
way, probably because the 2" machine is unresolved.
Any insights? Am I doing something wrong or is this as close as
it's going to get, unresolved? I can't get my head around why Lack Of
Black would make a difference, although I suspect that's what the
problems is.
Thanks, Rick.


What Adam says is correct. When slaving Protools to an unresolved tape
machine, the Clock Reference for the USD MUST be set for LTC or it WILL
drift.

Positional reference tells Protools where to start playing, but unless
the clock is resolved to the SMPTE, once Protools starts playing, it
will just run at it's own speed, ignoring the speed of the timecode once
it's begun playing (or recording in this case).

This explains why your PT rig "locks up" so quickly when you begin
playing. When it's correctly setup, it will usually take 2-3 seconds
(sometimes longer, depending on the disk buffering/number of tracks in
play/number of tracks in record/phase of the moon/your client's
attitude/the color of your walls...

What's amazing is that you're getting only 46 samples worth of drift
over the course of a song....that's pretty spectacular speed accuracy
for any 2" machine. What's the machine?

BTW, once you set the Clock Reference to LTC, if your rig has trouble
locking up, you might need to adjust the disk buffering....it's been a
while since I've used the USD, I forget the details/tricks....I think
smaller buffers is better in this case.

Some will say that slaving the clock to LTC SMPTE is EVIL (causes
impotency or something...). While it certainly isn't the ideal way of
doing it, it's often the most expedient way. If you don't have a
synchronizer for your tape machine, and a clock source to resolve both
it and the USD to, it's the ONLY way to do it. But if you have the
hardware, then resolve the tape machine to your house sync reference
(typically blackburst), and set your USD Clock Reference to the same
source. Positional reference would still be LTC.


Chestek

--
Anti-Spam email address in effect.
My real email should be pretty obvious to an actual human being.

Jeff wrote:
When slaving Protools to an unresolved tape machine,
the Clock Reference for the USD MUST be set for LTC or it WILL drift.

Thanks Jeff, that's what I needed to know.

What's amazing is that you're getting only 49 samples worth of drift
over the course of a song....that's pretty spectacular speed accuracy
for any 2" machine. What's the machine?

An MCI JH24 that I got for next to nothing. Tweaked by Charlie Bolis
though. Having owned Studers and Ampex's (Ampexi?) in the past, I had
always looked down my nose a bit at MCIs, but this one continues to
impress me. Sounds great too, like an Ampex but more so.
Maybe I just got lucky with the one song I've done so far. I'll
try some more manana with the clock locked to LTC.
Thanks again to all,
Rick.

In article .com,
"rickymix" wrote:

Jeff wrote:
When slaving Protools to an unresolved tape machine,
the Clock Reference for the USD MUST be set for LTC or it WILL drift.

Thanks Jeff, that's what I needed to know.

What's amazing is that you're getting only 49 samples worth of drift
over the course of a song....that's pretty spectacular speed accuracy
for any 2" machine. What's the machine?

An MCI JH24 that I got for next to nothing. Tweaked by Charlie Bolis
though. Having owned Studers and Ampex's (Ampexi?) in the past, I had
always looked down my nose a bit at MCIs, but this one continues to
impress me. Sounds great too, like an Ampex but more so.
Maybe I just got lucky with the one song I've done so far. I'll
try some more manana with the clock locked to LTC.
Thanks again to all,
Rick.

I couldn't begin to count the number of records I've cut on JH-24s. When
I was a young-un I spent half an hour a day (or more) just aligning the
machine for each session (...we kneel now for our daily prayer to the
"washing machine"!...).

Workhorses...

They cut really good tracks, but don't seem quite able to actually
reproduce everything that they put onto the tape. Play back the tapes on
a really nice Studer, and you'll be happily surprised!

Keep your power supplies clean.

And keep a couple of spare transistors for the reel motor drivers
(2N3055 H...make sure you get the much harder to find "H" version!).
They go shorted pretty commonly, making one reel take off at full speed
despite being in "stop". Only thing to do is let the tape wind off, and
shut off the power just after it runs off the reel. Don't even THINK of
trying to stop the tape before it runs out! (don't ask...)

They don't make terribly good slave machines, you need to be VERY
particular about the tape path and capstan alignment and tension
adjustments, and it'll still want to crinkle the tape as it slews up and
down in speed while synchronizing to the master. It's much happier as
the Master machine! Of course, it's not very common anymore to see
people slaving ANY 2" machines. And just resolving the JH24 to house
sync (as we'd discussed earlier) isn't bad at all.

I'm thrilled that someone's still getting good use from these guys!
We've got two that haven't seen tape in months (even the Studer only
gets used once in a while :-{ . They're going for dirt cheap these days,
I might have to buy one of the studio's to go with my MXP console.

Chestek

--
Anti-Spam email address in effect.
My real email should be pretty obvious to an actual human being.