Sign in with
Sign up | Sign in
Your question

ADAT Lightpipe transfer

Last response: in Home Audio
Share
Anonymous
August 31, 2004 1:31:35 AM

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

I am recording to 2 ADAT XT-20s through an M-3700 console. I would like
to transfer tracks to a PC running cubase SX. Does anyone have
experience with the steinberg VSL-2020 card. Does anyone else transfer
between computer and ADAT. I would like this to happen in sync so my
transfers back to adat are in time. Any advise or comments would be helpful.
Thanks
Dave O.
August 31, 2004 1:31:36 AM

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

I don't know your card but I am sure it works like others, The synch
part happens through the light pipe. The actual clock is part of the
signal. You either slave your computer to the adat, your adat to the
computer, or do it the simplest way which is to push record on your daw
and play on your adats and let the sound roll on in. Even if you go back
out through the analog and back into digital you won't notice any real
lose.

If you card has light pipe though, it is a piece of cake. set your
computer to slave and push record, then push play on the adat.

Noizman wrote:

> I am recording to 2 ADAT XT-20s through an M-3700 console. I would like
> to transfer tracks to a PC running cubase SX. Does anyone have
> experience with the steinberg VSL-2020 card. Does anyone else transfer
> between computer and ADAT. I would like this to happen in sync so my
> transfers back to adat are in time. Any advise or comments would be
> helpful.
> Thanks
> Dave O.
Anonymous
August 31, 2004 1:31:37 AM

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

I've read that the lightpipe connections carry a sync signal but some
cards advertise ADAT sync and some (like the VSL2020Z) just have word
clock. I know there is a difference but will the word clock be as
accurate as the "sample accurate" sync of the ADAT format.

Dave
noizman

Danny wrote:
> I don't know your card but I am sure it works like others, The synch
> part happens through the light pipe. The actual clock is part of the
> signal. You either slave your computer to the adat, your adat to the
> computer, or do it the simplest way which is to push record on your daw
> and play on your adats and let the sound roll on in. Even if you go back
> out through the analog and back into digital you won't notice any real
> lose.
>
> If you card has light pipe though, it is a piece of cake. set your
> computer to slave and push record, then push play on the adat.
>
> Noizman wrote:
>
>> I am recording to 2 ADAT XT-20s through an M-3700 console. I would
>> like to transfer tracks to a PC running cubase SX. Does anyone have
>> experience with the steinberg VSL-2020 card. Does anyone else transfer
>> between computer and ADAT. I would like this to happen in sync so my
>> transfers back to adat are in time. Any advise or comments would be
>> helpful.
>> Thanks
>> Dave O.
>
>
Related resources
August 31, 2004 1:31:38 AM

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

as i understand it, a word clock counts the words (bits) going through
and is the same thing as adat sync. If you read a bunch of E-bla bla bla
on a computer it means firewire, mac's technology, but only mac can call
it firewire. I think the same is true with adat and how they license the
technology.

Owned wrote:

> I've read that the lightpipe connections carry a sync signal but some
> cards advertise ADAT sync and some (like the VSL2020Z) just have word
> clock. I know there is a difference but will the word clock be as
> accurate as the "sample accurate" sync of the ADAT format.
>
> Dave
> noizman
>
> Danny wrote:
>
>> I don't know your card but I am sure it works like others, The synch
>> part happens through the light pipe. The actual clock is part of the
>> signal. You either slave your computer to the adat, your adat to the
>> computer, or do it the simplest way which is to push record on your
>> daw and play on your adats and let the sound roll on in. Even if you
>> go back out through the analog and back into digital you won't notice
>> any real lose.
>>
>> If you card has light pipe though, it is a piece of cake. set your
>> computer to slave and push record, then push play on the adat.
>>
>> Noizman wrote:
>>
>>> I am recording to 2 ADAT XT-20s through an M-3700 console. I would
>>> like to transfer tracks to a PC running cubase SX. Does anyone have
>>> experience with the steinberg VSL-2020 card. Does anyone else
>>> transfer between computer and ADAT. I would like this to happen in
>>> sync so my transfers back to adat are in time. Any advise or comments
>>> would be helpful.
>>> Thanks
>>> Dave O.
>>
>>
>>
>
August 31, 2004 1:32:14 AM

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

Danny wrote:
> I don't know your card but I am sure it works like others, The synch
> part happens through the light pipe. The actual clock is part of the
> signal. You either slave your computer to the adat, your adat to the
> computer, or do it the simplest way which is to push record on your daw
> and play on your adats and let the sound roll on in. Even if you go back
> out through the analog and back into digital you won't notice any real
> lose.
>
> If you card has light pipe though, it is a piece of cake. set your
> computer to slave and push record, then push play on the adat.

Lightpipe does not carry a time code reference. It does carry work clock
info. Since Noizeman has 2 Adat machines, and his interface has 2
Lightpipe ins, he will be fine. His tracks will be in sync in the time
domain as long as the ADAT sync cable between the 2 ADATs is connected.
If he adds a third ADAT machine, and transfers that in on a second pass,
the third machine's tracks will not line up with the first 2 machine's
tracks.

You need to understand the difference between "Word Clock" sync and
"Time Code" sync. They are different.

Take a look at something like RME's Hammerfall card. That has 3 sets of
Lightpipe ins and outs. It also has ADAT Sync (the 9 pin connector). As
long as you have the ADAT sync 9 pin connected, you could transfer one
machine at a time, and the tracks would line up.

--
Eric

Practice Your Mixing Skills
Multi-Track Masters on CD-ROM
www.Raw-Tracks.com
August 31, 2004 1:32:15 AM

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

EricK wrote:

> Danny wrote:
>
>> I don't know your card but I am sure it works like others, The synch
>> part happens through the light pipe. The actual clock is part of the
>> signal. You either slave your computer to the adat, your adat to the
>> computer, or do it the simplest way which is to push record on your
>> daw and play on your adats and let the sound roll on in. Even if you
>> go back out through the analog and back into digital you won't notice
>> any real lose.
>>
>> If you card has light pipe though, it is a piece of cake. set your
>> computer to slave and push record, then push play on the adat.
>
>
> Lightpipe does not carry a time code reference. It does carry work clock
> info. Since Noizeman has 2 Adat machines, and his interface has 2
> Lightpipe ins, he will be fine. His tracks will be in sync in the time
> domain as long as the ADAT sync cable between the 2 ADATs is connected.
> If he adds a third ADAT machine, and transfers that in on a second pass,
> the third machine's tracks will not line up with the first 2 machine's
> tracks.
>
> You need to understand the difference between "Word Clock" sync and
> "Time Code" sync. They are different.
>
> Take a look at something like RME's Hammerfall card. That has 3 sets of
> Lightpipe ins and outs. It also has ADAT Sync (the 9 pin connector). As
> long as you have the ADAT sync 9 pin connected, you could transfer one
> machine at a time, and the tracks would line up.
>

And see - I knew that but the beer was hiding it from me. Bad beer, bad
beer! Now I need another beer
Anonymous
August 31, 2004 3:27:06 AM

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

So if I transfer only a few tracks to the computer for editing and then
back to ADAT, I need to have ADAT sync on the card to keep it all in time?

Noizman

EricK wrote:

> Danny wrote:
>
>> I don't know your card but I am sure it works like others, The synch
>> part happens through the light pipe. The actual clock is part of the
>> signal. You either slave your computer to the adat, your adat to the
>> computer, or do it the simplest way which is to push record on your
>> daw and play on your adats and let the sound roll on in. Even if you
>> go back out through the analog and back into digital you won't notice
>> any real lose.
>>
>> If you card has light pipe though, it is a piece of cake. set your
>> computer to slave and push record, then push play on the adat.
>
>
> Lightpipe does not carry a time code reference. It does carry work clock
> info. Since Noizeman has 2 Adat machines, and his interface has 2
> Lightpipe ins, he will be fine. His tracks will be in sync in the time
> domain as long as the ADAT sync cable between the 2 ADATs is connected.
> If he adds a third ADAT machine, and transfers that in on a second pass,
> the third machine's tracks will not line up with the first 2 machine's
> tracks.
>
> You need to understand the difference between "Word Clock" sync and
> "Time Code" sync. They are different.
>
> Take a look at something like RME's Hammerfall card. That has 3 sets of
> Lightpipe ins and outs. It also has ADAT Sync (the 9 pin connector). As
> long as you have the ADAT sync 9 pin connected, you could transfer one
> machine at a time, and the tracks would line up.
>
Anonymous
August 31, 2004 6:04:54 AM

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

Fire wire is a western digital thing.

Bob

"Danny" <Keep@Your.com> wrote in message
news:10j7ldcl56glh76@corp.supernews.com...
> as i understand it, a word clock counts the words (bits) going through
> and is the same thing as adat sync. If you read a bunch of E-bla bla bla
> on a computer it means firewire, mac's technology, but only mac can call
> it firewire. I think the same is true with adat and how they license the
> technology.
>
> Owned wrote:
>
> > I've read that the lightpipe connections carry a sync signal but some
> > cards advertise ADAT sync and some (like the VSL2020Z) just have word
> > clock. I know there is a difference but will the word clock be as
> > accurate as the "sample accurate" sync of the ADAT format.
> >
> > Dave
> > noizman
> >
> > Danny wrote:
> >
> >> I don't know your card but I am sure it works like others, The synch
> >> part happens through the light pipe. The actual clock is part of the
> >> signal. You either slave your computer to the adat, your adat to the
> >> computer, or do it the simplest way which is to push record on your
> >> daw and play on your adats and let the sound roll on in. Even if you
> >> go back out through the analog and back into digital you won't notice
> >> any real lose.
> >>
> >> If you card has light pipe though, it is a piece of cake. set your
> >> computer to slave and push record, then push play on the adat.
> >>
> >> Noizman wrote:
> >>
> >>> I am recording to 2 ADAT XT-20s through an M-3700 console. I would
> >>> like to transfer tracks to a PC running cubase SX. Does anyone have
> >>> experience with the steinberg VSL-2020 card. Does anyone else
> >>> transfer between computer and ADAT. I would like this to happen in
> >>> sync so my transfers back to adat are in time. Any advise or comments
> >>> would be helpful.
> >>> Thanks
> >>> Dave O.
> >>
> >>
> >>
> >
>
August 31, 2004 6:04:55 AM

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

it was mac's big thing. Whoever made it doesn't count, mac owns it. It
was big mac news and all of us mac addicts were quite proud of it when
it came out. :-)

sycochkn wrote:

> Fire wire is a western digital thing.
>
> Bob
>
> "Danny" <Keep@Your.com> wrote in message
> news:10j7ldcl56glh76@corp.supernews.com...
>
>>as i understand it, a word clock counts the words (bits) going through
>>and is the same thing as adat sync. If you read a bunch of E-bla bla bla
>>on a computer it means firewire, mac's technology, but only mac can call
>>it firewire. I think the same is true with adat and how they license the
>>technology.
>>
>>Owned wrote:
>>
>>
>>>I've read that the lightpipe connections carry a sync signal but some
>>>cards advertise ADAT sync and some (like the VSL2020Z) just have word
>>>clock. I know there is a difference but will the word clock be as
>>>accurate as the "sample accurate" sync of the ADAT format.
>>>
>>>Dave
>>>noizman
>>>
>>>Danny wrote:
>>>
>>>
>>>>I don't know your card but I am sure it works like others, The synch
>>>>part happens through the light pipe. The actual clock is part of the
>>>>signal. You either slave your computer to the adat, your adat to the
>>>>computer, or do it the simplest way which is to push record on your
>>>>daw and play on your adats and let the sound roll on in. Even if you
>>>>go back out through the analog and back into digital you won't notice
>>>>any real lose.
>>>>
>>>>If you card has light pipe though, it is a piece of cake. set your
>>>>computer to slave and push record, then push play on the adat.
>>>>
>>>>Noizman wrote:
>>>>
>>>>
>>>>>I am recording to 2 ADAT XT-20s through an M-3700 console. I would
>>>>>like to transfer tracks to a PC running cubase SX. Does anyone have
>>>>>experience with the steinberg VSL-2020 card. Does anyone else
>>>>>transfer between computer and ADAT. I would like this to happen in
>>>>>sync so my transfers back to adat are in time. Any advise or comments
>>>>>would be helpful.
>>>>>Thanks
>>>>>Dave O.
>>>>
>>>>
>>>>
>
>
Anonymous
August 31, 2004 12:40:46 PM

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

In article <5oQYc.45315$%n4.39550@bignews6.bellsouth.net> owend@bellsouth.net writes:

> I've read that the lightpipe connections carry a sync signal but some
> cards advertise ADAT sync and some (like the VSL2020Z) just have word
> clock. I know there is a difference but will the word clock be as
> accurate as the "sample accurate" sync of the ADAT format.

The ADAT optical format allows for a word clock signal to be embedded
in the data stream, and a way for extracting it on the other end. Real
ADATs do this and make use of it for synchronizing the data rate. Some
other products using the ADAT data format use the embedded word clock
(generate or extract it, depending on whether it's a gozinta or a
gozouta), others don't, and depend on an external word clock
connection.

This synchronizes the data rate so you don't get clicks when the
device's clock doesn't match up with the incoming data, but it doesn't
synchronize the sample number or time position. All it guarantees is
that SOME sample will be present when the clock says it's time to read
a number.

To get "sample accurate" sync, you need more information. This is
what's carried by the ADAT Sync (9-pin D-sub) connection, and that's
the only way to assure it. But the concept of sample accurate sync is
a bit overrated. There are only special cases where you actually need
it. Time sync via SMPTE or MIDI Time Code is good enough for music.


--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
August 31, 2004 2:29:44 PM

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

Owend <owend@bellsouth.net> wrote in
news:L6SYc.45839$%n4.34610@bignews6.bellsouth.net:

> So if I transfer only a few tracks to the computer for editing and
> then back to ADAT, I need to have ADAT sync on the card to keep it all
> in time?
>
> Noizman
>

Correct
Anonymous
September 1, 2004 12:01:55 PM

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

Word clock means your oscillators are running at the same speed, ADAT sync
means your system (host/slave) knows the time code. Two different things.
Word clock is to reduce transfer jitter, the other is to make certain your
tracks line up correctly.

--


Roger W. Norman
SirMusic Studio

"Danny" <Keep@Your.com> wrote in message
news:10j7ldcl56glh76@corp.supernews.com...
> as i understand it, a word clock counts the words (bits) going through
> and is the same thing as adat sync. If you read a bunch of E-bla bla bla
> on a computer it means firewire, mac's technology, but only mac can call
> it firewire. I think the same is true with adat and how they license the
> technology.
>
> Owned wrote:
>
> > I've read that the lightpipe connections carry a sync signal but some
> > cards advertise ADAT sync and some (like the VSL2020Z) just have word
> > clock. I know there is a difference but will the word clock be as
> > accurate as the "sample accurate" sync of the ADAT format.
> >
> > Dave
> > noizman
> >
> > Danny wrote:
> >
> >> I don't know your card but I am sure it works like others, The synch
> >> part happens through the light pipe. The actual clock is part of the
> >> signal. You either slave your computer to the adat, your adat to the
> >> computer, or do it the simplest way which is to push record on your
> >> daw and play on your adats and let the sound roll on in. Even if you
> >> go back out through the analog and back into digital you won't notice
> >> any real lose.
> >>
> >> If you card has light pipe though, it is a piece of cake. set your
> >> computer to slave and push record, then push play on the adat.
> >>
> >> Noizman wrote:
> >>
> >>> I am recording to 2 ADAT XT-20s through an M-3700 console. I would
> >>> like to transfer tracks to a PC running cubase SX. Does anyone have
> >>> experience with the steinberg VSL-2020 card. Does anyone else
> >>> transfer between computer and ADAT. I would like this to happen in
> >>> sync so my transfers back to adat are in time. Any advise or comments
> >>> would be helpful.
> >>> Thanks
> >>> Dave O.
> >>
> >>
> >>
> >
>
Anonymous
September 1, 2004 12:03:50 PM

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

Firewire is an APPLE thing, not Western Digital. It was ratified by IEEE
under 1394 and is a MAC layer device protocol, hence why you can run
proprietary upper layer protocols over it such as MLan or even ethernet if
one wanted to write the code.

--


Roger W. Norman
SirMusic Studio

"sycochkn" <sycochkn@earthlink.net> wrote in message
news:a1RYc.3326$8d1.1739@newsread2.news.pas.earthlink.net...
> Fire wire is a western digital thing.
>
> Bob
>
> "Danny" <Keep@Your.com> wrote in message
> news:10j7ldcl56glh76@corp.supernews.com...
> > as i understand it, a word clock counts the words (bits) going through
> > and is the same thing as adat sync. If you read a bunch of E-bla bla bla
> > on a computer it means firewire, mac's technology, but only mac can call
> > it firewire. I think the same is true with adat and how they license the
> > technology.
> >
> > Owned wrote:
> >
> > > I've read that the lightpipe connections carry a sync signal but some
> > > cards advertise ADAT sync and some (like the VSL2020Z) just have word
> > > clock. I know there is a difference but will the word clock be as
> > > accurate as the "sample accurate" sync of the ADAT format.
> > >
> > > Dave
> > > noizman
> > >
> > > Danny wrote:
> > >
> > >> I don't know your card but I am sure it works like others, The synch
> > >> part happens through the light pipe. The actual clock is part of the
> > >> signal. You either slave your computer to the adat, your adat to the
> > >> computer, or do it the simplest way which is to push record on your
> > >> daw and play on your adats and let the sound roll on in. Even if you
> > >> go back out through the analog and back into digital you won't notice
> > >> any real lose.
> > >>
> > >> If you card has light pipe though, it is a piece of cake. set your
> > >> computer to slave and push record, then push play on the adat.
> > >>
> > >> Noizman wrote:
> > >>
> > >>> I am recording to 2 ADAT XT-20s through an M-3700 console. I would
> > >>> like to transfer tracks to a PC running cubase SX. Does anyone have
> > >>> experience with the steinberg VSL-2020 card. Does anyone else
> > >>> transfer between computer and ADAT. I would like this to happen in
> > >>> sync so my transfers back to adat are in time. Any advise or
comments
> > >>> would be helpful.
> > >>> Thanks
> > >>> Dave O.
> > >>
> > >>
> > >>
> > >
> >
>
>
Anonymous
September 1, 2004 4:03:53 PM

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

Danny <Keep@Your.com> wrote in message news:<10j7davq6iu7355@corp.supernews.com>...

> or do it the simplest way which is to push record on your daw
> and play on your adats and let the sound roll on in.


Arrghh! NO! DO NOT DO IT THAT WAY! Do not record any digital source
asynchronously into a digital recorder!

If you do, your dick will fall off.
Anonymous
September 1, 2004 4:12:32 PM

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

In article <4135b8c8$0$19713$61fed72c@news.rcn.com> rnorman@starpower.net writes:

> Word clock means your oscillators are running at the same speed, ADAT sync
> means your system (host/slave) knows the time code. Two different things.
> Word clock is to reduce transfer jitter, the other is to make certain your
> tracks line up correctly.

Word clock isn't just to reduce jitter. Synchronization is necessary
to make sure that the receiving end is ready to receive a bit when the
sending end is sending it. The ADAT sync cable carries word clock and
takes care of both data and time synchronization in one connection.
Otherwise you have to do them separately.

Since the ADAT sync knows both when the data transmissions occur and
the exact time for each sample, time synchronization can be "sample
accurate." While you can synchronize the time code rate with the
data clock, you can't do that to the resolution of a single sample, so
this technique will never give sample accurate sync.


--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
September 1, 2004 5:20:31 PM

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

Buster Mudd wrote:

> Danny <Keep@Your.com> wrote in message news:<10j7davq6iu7355@corp.supernews.com>...
>
>
>>or do it the simplest way which is to push record on your daw
>>and play on your adats and let the sound roll on in.
>
>
>
> Arrghh! NO! DO NOT DO IT THAT WAY! Do not record any digital source
> asynchronously into a digital recorder!
>
> If you do, your dick will fall off.

I was wonder were it went..... yes I have been corrected but I have done
it and it worked, unless I did something else and forgot. Well anyway, I
have been corrected
Anonymous
September 1, 2004 6:32:58 PM

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

Sorry, Mike, but word clock is to synchronize all clocks to the same
oscillation, that way when the bytes come in they are in sync the best they
can be. ADAT has a sync, but it's not word, it's ADAT sync, which is audio
sync. In fact, a word clock master doesn't even think about audio, so it
has no idea of what audio sync is. It only knows digital sync in the realm
of having one or more oscillators running at the same speed so that there's
no fluctuation, hence establishing a lower jitter rate because each
individual device isn't re-establishing sync (and having pops and clicks).

I quote:

Music Dictionary: Word Clock
The precise and accurate timing of digital audio samples is critical to the
correct operation of interconnected digital audio equipment. The 'metronome'
that governs sample timing is called the Word Clock (sometimes conjoined to
'Wordclock', or abbreviated to 'Wclk'). However, word clock does more than
merely beat time; it also identifies the start and end of each digital word
or sample, and which samples belong to the left or right channels. Digital
interfaces such as the AES-EBU and S/PDIF embody clock signals within the
data stream, but it is often necessary to convey a discrete word clock
between equipment as a square wave signal running at the sampling rate.
Dedicated word clock inputs and outputs on digital equipment generally use
BNC connectors (the kind of terminals commonly used for video).



Since we have to discrete units of time differentiation and one isn't
running audio, then word clock has nothing to do with audio, but rather that
ability to time each device's clock to an incoming signal from a discrete
device. Hence, word clock times all devices to a specific frequency, and it
keeps the digital signals coming in from audio/video at that specific
frequency. No oscillator runs exactly at the rate of another at it's native
frequency. Therefore, it's necessary to reclock in order to make data
transmission as reliable as possible.

Glenn Meadows and I argued about this for a long time back in 1989 because
audio people were just starting to talk about "jitter" and I already had
jitter problems with digital transmissions on token ring when moving up to
16 Mb/s and a "wild" chipset. He said it wasn't the same thing and I said
it was. I was right. An external clock was established to correct the
problem and it worked fine, both in token ring and in digital audio.

Try it yourself. Measure what's coming out of a word clock BNC and the sync
signal coming from ADAT. They not only will not be the same thing, but
quartz chips can vary quite extremely between different chips specifically
set at the same frequency. Voltage plays a part, but impurities in the
quartz plays a part too. Hence, the idea of having one chip be the master
and the rest be slaves means that all "beats" are exactly the same. A good
master word clock gives back a tone that any ADAT sync looses because each
device operates at it's own natural frequency and reclocks, which can end up
in jitter.

Now when you talk about 9 pin ADAT sync, that's different because it
incorporates real "word clock". ADAT lightpipe does not incorporate ADAT
sync, nor any other type of word clock.
--


Roger W. Norman
SirMusic Studio

"Mike Rivers" <mrivers@d-and-d.com> wrote in message
news:znr1094044894k@trad...
>
> In article <4135b8c8$0$19713$61fed72c@news.rcn.com> rnorman@starpower.net
writes:
>
> > Word clock means your oscillators are running at the same speed, ADAT
sync
> > means your system (host/slave) knows the time code. Two different
things.
> > Word clock is to reduce transfer jitter, the other is to make certain
your
> > tracks line up correctly.
>
> Word clock isn't just to reduce jitter. Synchronization is necessary
> to make sure that the receiving end is ready to receive a bit when the
> sending end is sending it. The ADAT sync cable carries word clock and
> takes care of both data and time synchronization in one connection.
> Otherwise you have to do them separately.
>
> Since the ADAT sync knows both when the data transmissions occur and
> the exact time for each sample, time synchronization can be "sample
> accurate." While you can synchronize the time code rate with the
> data clock, you can't do that to the resolution of a single sample, so
> this technique will never give sample accurate sync.
>
>
> --
> I'm really Mike Rivers (mrivers@d-and-d.com)
> However, until the spam goes away or Hell freezes over,
> lots of IP addresses are blocked from this system. If
> you e-mail me and it bounces, use your secret decoder ring
> and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 1, 2004 6:58:27 PM

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

ADAT does NOT establish word clock. It, like Tascam, resolves to Absolute
Time, which is only a matter of timing that is acceptable within the
differences between two or more units. Absolute Time is only slightly
better than SMPTE, and SMPTE doesn't provide the clocking reference for
multiple devices as does Word.

Now one has to assume that you mean that the ADAT 9 pin interface provides
what you speak of, but not lightpipe, else there would never be a problem in
clocking, and as some of us know, without word clock, it happens all the
time.

Think of it more as a beam of electricity that is as consistent as a pulsar.
For ADAT we have transport problems that could arise, along with voltage
differences that might change the speed of the transport, but with ADAT's
absolute time sync, it makes no difference with an external word clock
because the word clock is the frequency that all devices speak. That means
that even slightly stretched tape would still clock right. Major problems
would ultimately resolve into a Bler 32 condition if not able to be clocked
correctly, whereas with internal ADAT sync, you'd notice pops and clicks.
I'm not certain which is better because both are bad.

But the imbedded clocking signal within ADAT lightpipe has nothing to do
with Word Clock. It has to do with ADAT sync and they aren't the same
thing.

--


Roger W. Norman
SirMusic Studio

"Mike Rivers" <mrivers@d-and-d.com> wrote in message
news:znr1093948420k@trad...
>
> In article <5oQYc.45315$%n4.39550@bignews6.bellsouth.net>
owend@bellsouth.net writes:
>
> > I've read that the lightpipe connections carry a sync signal but some
> > cards advertise ADAT sync and some (like the VSL2020Z) just have word
> > clock. I know there is a difference but will the word clock be as
> > accurate as the "sample accurate" sync of the ADAT format.
>
> The ADAT optical format allows for a word clock signal to be embedded
> in the data stream, and a way for extracting it on the other end. Real
> ADATs do this and make use of it for synchronizing the data rate. Some
> other products using the ADAT data format use the embedded word clock
> (generate or extract it, depending on whether it's a gozinta or a
> gozouta), others don't, and depend on an external word clock
> connection.
>
> This synchronizes the data rate so you don't get clicks when the
> device's clock doesn't match up with the incoming data, but it doesn't
> synchronize the sample number or time position. All it guarantees is
> that SOME sample will be present when the clock says it's time to read
> a number.
>
> To get "sample accurate" sync, you need more information. This is
> what's carried by the ADAT Sync (9-pin D-sub) connection, and that's
> the only way to assure it. But the concept of sample accurate sync is
> a bit overrated. There are only special cases where you actually need
> it. Time sync via SMPTE or MIDI Time Code is good enough for music.
>
>
> --
> I'm really Mike Rivers (mrivers@d-and-d.com)
> However, until the spam goes away or Hell freezes over,
> lots of IP addresses are blocked from this system. If
> you e-mail me and it bounces, use your secret decoder ring
> and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 1, 2004 11:06:08 PM

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

In article <4136146d$0$19722$61fed72c@news.rcn.com> rnorman@starpower.net writes:

> Sorry, Mike, but word clock is to synchronize all clocks to the same
> oscillation, that way when the bytes come in they are in sync the best they
> can be.

and I said:

> > Word clock isn't just to reduce jitter. Synchronization is necessary
> > to make sure that the receiving end is ready to receive a bit when the
> > sending end is sending it.

What's to be sorry about?

> ADAT has a sync, but it's not word, it's ADAT sync, which is audio
> sync.

ADAT sync is ADAT sync and, among other things, it synchronizes the
word clocks in the two units connected by an ADAT sync cable.

> In fact, a word clock master doesn't even think about audio, so it
> has no idea of what audio sync is.

What is this audio sync of which you speak? Is that the same thing
that I call time sync? That when one starts playing bar 3 the other
one isn't playing bar 4? That's what time sync does, whether it's MTC,
SMPTE time code, or the time part of ADAT sync. The difference is that
ADAT sync knows that sample 4,728,112 is right at the start of the
third beat of bar 20 in the song (or when Ahnold starts to say
"girlymen"). Word clock doesn't know and doesn't care.

> It only knows digital sync in the realm
> of having one or more oscillators running at the same speed so that there's
> no fluctuation, hence establishing a lower jitter rate because each
> individual device isn't re-establishing sync (and having pops and clicks).

You may be taking about inter-device jitter, but that's not what we
mean when we say "jitter" when talking about a digital data stream. We
mean the short term stability of the data clock. Every clock has a
little slop, and while the long term average may be 44.100000000 kHz,
the tolerance for every click may be +/- a few picoseconds, some
occurring a trifle early, others occurring a trifle late. But that
drummer just keeps the average tempo at 44.1 kHz. The error of the
timing of each clock tick is what we call jitter.

> Since we have to discrete units of time differentiation and one isn't
> running audio, then word clock has nothing to do with audio, but rather that
> ability to time each device's clock to an incoming signal from a discrete
> device. Hence, word clock times all devices to a specific frequency,

OK so far.

> and it
> keeps the digital signals coming in from audio/video at that specific
> frequency.

Still OK, sort of, but you're drifting.

> No oscillator runs exactly at the rate of another at it's native
> frequency. Therefore, it's necessary to reclock in order to make data
> transmission as reliable as possible.

Nope. Word clock synchronization, at least as it's conventionaly
done, doesn't reclock as reclocking is conventionally done. What it
does is provides the same reference for each clock so that one doesn't
try to run at 44.0995 kHz and another at 44.1005 kHz when trying to
move digital audio from one to the other. The transfer will work OK
for a while, but eventually the two clocks will get far enough out of
phase so that the sender will send when the receiver is still in the
not-ready-to-receive-yet part of its cycle and you'll drop a sample
<click>. After another cycle or two, they'll drift back toward
in-phase and things will be OK for another little while until they get
too far out of phase again.

If both clocks are synchronized, they won't drift so far in phase that
you'll get loss of data.

Re-clocking sounds a bit like the same thing, but it's actually
different. The incoming data is actually re-sampled at the receiver's
clock rate. Typically this is used for sample rate conversion, or for
when you can't synchronize word clocks because you don't have all the
gozintas and gozoutas that you need.

> Glenn Meadows and I argued about this for a long time back in 1989 because
> audio people were just starting to talk about "jitter" and I already had
> jitter problems with digital transmissions on token ring when moving up to
> 16 Mb/s and a "wild" chipset. He said it wasn't the same thing and I said
> it was. I was right. An external clock was established to correct the
> problem and it worked fine, both in token ring and in digital audio.

You were right in that you needed to synchronize the clocks. You were
wrong to call it "jitter" though in 1989 it wasn't that common a term
in audio.

> Now when you talk about 9 pin ADAT sync, that's different because it
> incorporates real "word clock". ADAT lightpipe does not incorporate ADAT
> sync, nor any other type of word clock.

Wait a minute! That's what I've been talking about all along when I've
said "ADAT sync." So we agree after all. At least in part.

--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 2, 2004 12:27:14 AM

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

So if you do it and it works out ok, does your dick get bigger?

--


Roger W. Norman
SirMusic Studio

"Buster Mudd" <mr_furious@mail.com> wrote in message
news:936ad1fd.0409011103.b92fa80@posting.google.com...
> Danny <Keep@Your.com> wrote in message
news:<10j7davq6iu7355@corp.supernews.com>...
>
> > or do it the simplest way which is to push record on your daw
> > and play on your adats and let the sound roll on in.
>
>
> Arrghh! NO! DO NOT DO IT THAT WAY! Do not record any digital source
> asynchronously into a digital recorder!
>
> If you do, your dick will fall off.
Anonymous
September 2, 2004 6:00:49 AM

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

My head is spinning but I am interested in this thread.

I would like to know: If I take my Digi 002 Rack and I connect additional
box of ADC's to its ADAT lightpipes input, will the samples line up or the
time line up or what will happen?

Thanks in advance,

John
Anonymous
September 2, 2004 11:34:36 AM

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

Close enough. Whether it's sample accurate is something else, but Word
Clock doesn't make anything sample accurate anyway, it just makes everything
listen at the same time to whatever is being the master. ADAT will work
fine for most people. I use Word Clock, but even I use it "wrong" in that I
don't have an Aarvark or other "clocking" unit that has a multi-headed
distribution setup. I gang them on "T"s with BNC coming from my Tango 24's
Word Clock output. It's cleaned up some smearing problems I had before, but
it's still not the best of all choices.

As far transfers/tracking, you'll probably never notice the difference.
When you do start noticing the difference, Word Clock can be the answer.
But in your situation with just the Digi 002 you'll be fine. When you add
another set of converters (you can only add one), probably you'll just be
happy. If you're not, you have the answer.

--


Roger W. Norman
SirMusic Studio

"John Phillips" <jsp5646@hotmail.com> wrote in message
news:l9vZc.4774$683.621325@twister.southeast.rr.com...
> My head is spinning but I am interested in this thread.
>
> I would like to know: If I take my Digi 002 Rack and I connect additional
> box of ADC's to its ADAT lightpipes input, will the samples line up or the
> time line up or what will happen?
>
> Thanks in advance,
>
> John
>
>
Anonymous
September 2, 2004 1:40:19 PM

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

In article <l9vZc.4774$683.621325@twister.southeast.rr.com> jsp5646@hotmail.com writes:

> I would like to know: If I take my Digi 002 Rack and I connect additional
> box of ADC's to its ADAT lightpipes input, will the samples line up or the
> time line up or what will happen?

Something has to be the word clock master (it could be the 002 or
the external A/D converter, or something else). Get that set up
correctly and everything will work fine. Ignore it and, while you
won't hear a delay, you'll almost certainly get clicks in your
recording due to slightly asynchronous data rates from the two
devices.



--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 2, 2004 8:54:21 PM

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

Thanks for the information. The 002 does not have a way to add a word clock
that I can see. I understand what the word clock does, I was unsure what
the ADAT lightpipe did.

Thanks again,
John

"Roger W. Norman" <rnorman@starpower.net> wrote in message
news:413703eb$0$19731$61fed72c@news.rcn.com...
> Close enough. Whether it's sample accurate is something else, but Word
> Clock doesn't make anything sample accurate anyway, it just makes
everything
> listen at the same time to whatever is being the master. ADAT will work
> fine for most people. I use Word Clock, but even I use it "wrong" in that
I
> don't have an Aarvark or other "clocking" unit that has a multi-headed
> distribution setup. I gang them on "T"s with BNC coming from my Tango
24's
> Word Clock output. It's cleaned up some smearing problems I had before,
but
> it's still not the best of all choices.
>
> As far transfers/tracking, you'll probably never notice the difference.
> When you do start noticing the difference, Word Clock can be the answer.
> But in your situation with just the Digi 002 you'll be fine. When you add
> another set of converters (you can only add one), probably you'll just be
> happy. If you're not, you have the answer.
>
> --
>
>
> Roger W. Norman
> SirMusic Studio
>
> "John Phillips" <jsp5646@hotmail.com> wrote in message
> news:l9vZc.4774$683.621325@twister.southeast.rr.com...
> > My head is spinning but I am interested in this thread.
> >
> > I would like to know: If I take my Digi 002 Rack and I connect
additional
> > box of ADC's to its ADAT lightpipes input, will the samples line up or
the
> > time line up or what will happen?
> >
> > Thanks in advance,
> >
> > John
> >
> >
>
>
Anonymous
September 2, 2004 8:54:22 PM

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

In article <1fIZc.5087$683.681219@twister.southeast.rr.com> jsp5646@hotmail.com writes:

> Thanks for the information. The 002 does not have a way to add a word clock
> that I can see. I understand what the word clock does, I was unsure what
> the ADAT lightpipe did.

I can't believe they didn't put a word clock connector, either output,
input, or both, on it, but I just looked at a picture on the web site
and it sure doesn't look like it. I guess with Digidesign, it costs
you another $10,000 or more to get that capability.

It may be possible, however, to select the ADAT lightpipe input as the
word clock source for the 002. This would probably be somewhere in its
software setup. If it does, then selecting that would synchronize the
word clock in the 002 with that of the external interface, and that's
what you need to do.



--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 2, 2004 8:58:09 PM

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

Thanks for the additional clarification. Without any syncing (and in
addition clicks and pops), it would seem to me that for a long live
performance that delay could creep in between the two sets of converters?

John

"Mike Rivers" <mrivers@d-and-d.com> wrote in message
news:znr1094125673k@trad...
>
> In article <l9vZc.4774$683.621325@twister.southeast.rr.com>
jsp5646@hotmail.com writes:
>
> > I would like to know: If I take my Digi 002 Rack and I connect
additional
> > box of ADC's to its ADAT lightpipes input, will the samples line up or
the
> > time line up or what will happen?
>
> Something has to be the word clock master (it could be the 002 or
> the external A/D converter, or something else). Get that set up
> correctly and everything will work fine. Ignore it and, while you
> won't hear a delay, you'll almost certainly get clicks in your
> recording due to slightly asynchronous data rates from the two
> devices.
>
>
>
> --
> I'm really Mike Rivers (mrivers@d-and-d.com)
> However, until the spam goes away or Hell freezes over,
> lots of IP addresses are blocked from this system. If
> you e-mail me and it bounces, use your secret decoder ring
> and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 2, 2004 8:58:10 PM

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

In article <BiIZc.5092$683.681222@twister.southeast.rr.com> jsp5646@hotmail.com writes:

> Thanks for the additional clarification. Without any syncing (and in
> addition clicks and pops), it would seem to me that for a long live
> performance that delay could creep in between the two sets of converters?

Well, no. You're recording at only one sample rate, so the playback
will always follow that rate.

--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 2, 2004 11:06:08 PM

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

When I started this thread I was thinking about getting the Steinberg
VSL 2020 card which has 16 in 16 out light pipe and a bnc word clock
connection. After learning a little here about the formats I am thinking
more of getting the MOTU 828mkII which has 8 I/O light pipe with adat
sync and word as well as 8 analog I/O. It will be used on a winXT PC.
Does anyone have a better suggestion for spending my hard earned $$.

Noizman
Gainesville, Fl.

Mike Rivers wrote:
> In article <4136146d$0$19722$61fed72c@news.rcn.com> rnorman@starpower.net writes:
>
>
>>Sorry, Mike, but word clock is to synchronize all clocks to the same
>>oscillation, that way when the bytes come in they are in sync the best they
>>can be.
>
>
> and I said:
>
>
>>>Word clock isn't just to reduce jitter. Synchronization is necessary
>>>to make sure that the receiving end is ready to receive a bit when the
>>>sending end is sending it.
>
>
> What's to be sorry about?
>
>
>>ADAT has a sync, but it's not word, it's ADAT sync, which is audio
>>sync.
>
>
> ADAT sync is ADAT sync and, among other things, it synchronizes the
> word clocks in the two units connected by an ADAT sync cable.
>
>
>>In fact, a word clock master doesn't even think about audio, so it
>>has no idea of what audio sync is.
>
>
> What is this audio sync of which you speak? Is that the same thing
> that I call time sync? That when one starts playing bar 3 the other
> one isn't playing bar 4? That's what time sync does, whether it's MTC,
> SMPTE time code, or the time part of ADAT sync. The difference is that
> ADAT sync knows that sample 4,728,112 is right at the start of the
> third beat of bar 20 in the song (or when Ahnold starts to say
> "girlymen"). Word clock doesn't know and doesn't care.
>
>
>>It only knows digital sync in the realm
>>of having one or more oscillators running at the same speed so that there's
>>no fluctuation, hence establishing a lower jitter rate because each
>>individual device isn't re-establishing sync (and having pops and clicks).
>
>
> You may be taking about inter-device jitter, but that's not what we
> mean when we say "jitter" when talking about a digital data stream. We
> mean the short term stability of the data clock. Every clock has a
> little slop, and while the long term average may be 44.100000000 kHz,
> the tolerance for every click may be +/- a few picoseconds, some
> occurring a trifle early, others occurring a trifle late. But that
> drummer just keeps the average tempo at 44.1 kHz. The error of the
> timing of each clock tick is what we call jitter.
>
>
>>Since we have to discrete units of time differentiation and one isn't
>>running audio, then word clock has nothing to do with audio, but rather that
>>ability to time each device's clock to an incoming signal from a discrete
>>device. Hence, word clock times all devices to a specific frequency,
>
>
> OK so far.
>
>
>>and it
>>keeps the digital signals coming in from audio/video at that specific
>>frequency.
>
>
> Still OK, sort of, but you're drifting.
>
>
>>No oscillator runs exactly at the rate of another at it's native
>>frequency. Therefore, it's necessary to reclock in order to make data
>>transmission as reliable as possible.
>
>
> Nope. Word clock synchronization, at least as it's conventionaly
> done, doesn't reclock as reclocking is conventionally done. What it
> does is provides the same reference for each clock so that one doesn't
> try to run at 44.0995 kHz and another at 44.1005 kHz when trying to
> move digital audio from one to the other. The transfer will work OK
> for a while, but eventually the two clocks will get far enough out of
> phase so that the sender will send when the receiver is still in the
> not-ready-to-receive-yet part of its cycle and you'll drop a sample
> <click>. After another cycle or two, they'll drift back toward
> in-phase and things will be OK for another little while until they get
> too far out of phase again.
>
> If both clocks are synchronized, they won't drift so far in phase that
> you'll get loss of data.
>
> Re-clocking sounds a bit like the same thing, but it's actually
> different. The incoming data is actually re-sampled at the receiver's
> clock rate. Typically this is used for sample rate conversion, or for
> when you can't synchronize word clocks because you don't have all the
> gozintas and gozoutas that you need.
>
>
>>Glenn Meadows and I argued about this for a long time back in 1989 because
>>audio people were just starting to talk about "jitter" and I already had
>>jitter problems with digital transmissions on token ring when moving up to
>>16 Mb/s and a "wild" chipset. He said it wasn't the same thing and I said
>>it was. I was right. An external clock was established to correct the
>>problem and it worked fine, both in token ring and in digital audio.
>
>
> You were right in that you needed to synchronize the clocks. You were
> wrong to call it "jitter" though in 1989 it wasn't that common a term
> in audio.
>
>
>>Now when you talk about 9 pin ADAT sync, that's different because it
>>incorporates real "word clock". ADAT lightpipe does not incorporate ADAT
>>sync, nor any other type of word clock.
>
>
> Wait a minute! That's what I've been talking about all along when I've
> said "ADAT sync." So we agree after all. At least in part.
>
> --
> I'm really Mike Rivers (mrivers@d-and-d.com)
> However, until the spam goes away or Hell freezes over,
> lots of IP addresses are blocked from this system. If
> you e-mail me and it bounces, use your secret decoder ring
> and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 3, 2004 12:07:00 AM

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

Both Mike and I have recorded hours of performances without drift problems,
and I actually mean that we've done so together. No, you won't have
problems with what you have, even if you add a set of converters.

Imaging might become a problem, but like I said, unless you can perceive the
problem, it's not really there, right? <g>

--


Roger W. Norman
SirMusic Studio

"John Phillips" <jsp5646@hotmail.com> wrote in message
news:BiIZc.5092$683.681222@twister.southeast.rr.com...
> Thanks for the additional clarification. Without any syncing (and in
> addition clicks and pops), it would seem to me that for a long live
> performance that delay could creep in between the two sets of converters?
>
> John
>
> "Mike Rivers" <mrivers@d-and-d.com> wrote in message
> news:znr1094125673k@trad...
> >
> > In article <l9vZc.4774$683.621325@twister.southeast.rr.com>
> jsp5646@hotmail.com writes:
> >
> > > I would like to know: If I take my Digi 002 Rack and I connect
> additional
> > > box of ADC's to its ADAT lightpipes input, will the samples line up or
> the
> > > time line up or what will happen?
> >
> > Something has to be the word clock master (it could be the 002 or
> > the external A/D converter, or something else). Get that set up
> > correctly and everything will work fine. Ignore it and, while you
> > won't hear a delay, you'll almost certainly get clicks in your
> > recording due to slightly asynchronous data rates from the two
> > devices.
> >
> >
> >
> > --
> > I'm really Mike Rivers (mrivers@d-and-d.com)
> > However, until the spam goes away or Hell freezes over,
> > lots of IP addresses are blocked from this system. If
> > you e-mail me and it bounces, use your secret decoder ring
> > and reach me here: double-m-eleven-double-zero at yahoo
>
>
Anonymous
September 3, 2004 6:28:42 AM

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

Thanks!

John

"Roger W. Norman" <rnorman@starpower.net> wrote in message
news:4137b440$0$19728$61fed72c@news.rcn.com...
> Both Mike and I have recorded hours of performances without drift
problems,
> and I actually mean that we've done so together. No, you won't have
> problems with what you have, even if you add a set of converters.
>
> Imaging might become a problem, but like I said, unless you can perceive
the
> problem, it's not really there, right? <g>
>
> --
>
>
> Roger W. Norman
> SirMusic Studio
>
> "John Phillips" <jsp5646@hotmail.com> wrote in message
> news:BiIZc.5092$683.681222@twister.southeast.rr.com...
> > Thanks for the additional clarification. Without any syncing (and in
> > addition clicks and pops), it would seem to me that for a long live
> > performance that delay could creep in between the two sets of
converters?
> >
> > John
> >
> > "Mike Rivers" <mrivers@d-and-d.com> wrote in message
> > news:znr1094125673k@trad...
> > >
> > > In article <l9vZc.4774$683.621325@twister.southeast.rr.com>
> > jsp5646@hotmail.com writes:
> > >
> > > > I would like to know: If I take my Digi 002 Rack and I connect
> > additional
> > > > box of ADC's to its ADAT lightpipes input, will the samples line up
or
> > the
> > > > time line up or what will happen?
> > >
> > > Something has to be the word clock master (it could be the 002 or
> > > the external A/D converter, or something else). Get that set up
> > > correctly and everything will work fine. Ignore it and, while you
> > > won't hear a delay, you'll almost certainly get clicks in your
> > > recording due to slightly asynchronous data rates from the two
> > > devices.
> > >
> > >
> > >
> > > --
> > > I'm really Mike Rivers (mrivers@d-and-d.com)
> > > However, until the spam goes away or Hell freezes over,
> > > lots of IP addresses are blocked from this system. If
> > > you e-mail me and it bounces, use your secret decoder ring
> > > and reach me here: double-m-eleven-double-zero at yahoo
> >
> >
>
>
Anonymous
September 4, 2004 4:36:03 PM

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

Missed the word "just" in

> > Word clock isn't just to reduce jitter. Synchronization is necessary
> > to make sure that the receiving end is ready to receive a bit when the
> > sending end is sending it.

although that's really the idea. Otherwise, what purpose to get your
oscillators running together? Any number of sync schemes can time
bytes(bits) but that's not what Word Clock does. It's simply to align
multiple units so that they are are all listening at the same time. ABS or
SMPTE or even ECC provide for the transmission of data within the signal,
and yet the clocks can be freewheeling and still get the job done. Word
Clock does not do this. It says "this is one" and all clocks react "OK,
this is one". If there's a failure of data transmission, as one using SMPTE
may know, it's possible to have dropped frames and still have proper word
clock sync. The likelyhood is far smaller, hence a reduction in jitter.

And

"The difference is that
ADAT sync knows that sample 4,728,112 is right at the start of the
third beat of bar 20 in the song (or when Ahnold starts to say
"girlymen"). Word clock doesn't know and doesn't care."

is exactly what I'm talking about. Word clock may be part of the ADAT
cabling, but it's not part of ADAT lightpipe, it's word clock is just
running over a nine pin connector. But we're arguing the same thing. More
appropriatedly a later point

"Nope. Word clock synchronization, at least as it's conventionaly
done, doesn't reclock as reclocking is conventionally done. What it
does is provides the same reference for each clock so that one doesn't
try to run at 44.0995 kHz and another at 44.1005 kHz when trying to
move digital audio from one to the other."

is an example of my misusing terminology. I didn't mean to say reclocking
as much as thinking I'd already established a system of clocking that
references all others and I thought I'd said that pretty specifically. And
it's the reference that reduces jitter.

But let's face it. ADAT, in terms of lightpipe, does not supply word clock.
ADAT sync, in terms of a 9 pin cable, provides word clock. So a converter
set like my Tango24 going to my RME Hammerfall 9652, or my Motu 2408 tied to
the Tango24 via lightpipe doesn't have Word Clock unless I link up the Word
Clock BNCs with a master. Even the "description" I supplied is incorrect
because Word Clock doesn't have anything to do with marking start/stop of
stream transmissions or sample positions, although the statement implies
that it does. They do, however, use the correct term, which is "metronome"
but any player can still play the song and not be playing along with the
metronome.

--


Roger W. Norman
SirMusic Studio

"Mike Rivers" <mrivers@d-and-d.com> wrote in message
news:znr1094069998k@trad...
>
> In article <4136146d$0$19722$61fed72c@news.rcn.com> rnorman@starpower.net
writes:
>
> > Sorry, Mike, but word clock is to synchronize all clocks to the same
> > oscillation, that way when the bytes come in they are in sync the best
they
> > can be.
>
> and I said:
>
> > > Word clock isn't just to reduce jitter. Synchronization is necessary
> > > to make sure that the receiving end is ready to receive a bit when the
> > > sending end is sending it.
>
> What's to be sorry about?
>
> > ADAT has a sync, but it's not word, it's ADAT sync, which is audio
> > sync.
>
> ADAT sync is ADAT sync and, among other things, it synchronizes the
> word clocks in the two units connected by an ADAT sync cable.
>
> > In fact, a word clock master doesn't even think about audio, so it
> > has no idea of what audio sync is.
>
> What is this audio sync of which you speak? Is that the same thing
> that I call time sync? That when one starts playing bar 3 the other
> one isn't playing bar 4? That's what time sync does, whether it's MTC,
> SMPTE time code, or the time part of ADAT sync. The difference is that
> ADAT sync knows that sample 4,728,112 is right at the start of the
> third beat of bar 20 in the song (or when Ahnold starts to say
> "girlymen"). Word clock doesn't know and doesn't care.
>
> > It only knows digital sync in the realm
> > of having one or more oscillators running at the same speed so that
there's
> > no fluctuation, hence establishing a lower jitter rate because each
> > individual device isn't re-establishing sync (and having pops and
clicks).
>
> You may be taking about inter-device jitter, but that's not what we
> mean when we say "jitter" when talking about a digital data stream. We
> mean the short term stability of the data clock. Every clock has a
> little slop, and while the long term average may be 44.100000000 kHz,
> the tolerance for every click may be +/- a few picoseconds, some
> occurring a trifle early, others occurring a trifle late. But that
> drummer just keeps the average tempo at 44.1 kHz. The error of the
> timing of each clock tick is what we call jitter.
>
> > Since we have to discrete units of time differentiation and one isn't
> > running audio, then word clock has nothing to do with audio, but rather
that
> > ability to time each device's clock to an incoming signal from a
discrete
> > device. Hence, word clock times all devices to a specific frequency,
>
> OK so far.
>
> > and it
> > keeps the digital signals coming in from audio/video at that specific
> > frequency.
>
> Still OK, sort of, but you're drifting.
>
> > No oscillator runs exactly at the rate of another at it's native
> > frequency. Therefore, it's necessary to reclock in order to make data
> > transmission as reliable as possible.
>
> Nope. Word clock synchronization, at least as it's conventionaly
> done, doesn't reclock as reclocking is conventionally done. What it
> does is provides the same reference for each clock so that one doesn't
> try to run at 44.0995 kHz and another at 44.1005 kHz when trying to
> move digital audio from one to the other. The transfer will work OK
> for a while, but eventually the two clocks will get far enough out of
> phase so that the sender will send when the receiver is still in the
> not-ready-to-receive-yet part of its cycle and you'll drop a sample
> <click>. After another cycle or two, they'll drift back toward
> in-phase and things will be OK for another little while until they get
> too far out of phase again.
>
> If both clocks are synchronized, they won't drift so far in phase that
> you'll get loss of data.
>
> Re-clocking sounds a bit like the same thing, but it's actually
> different. The incoming data is actually re-sampled at the receiver's
> clock rate. Typically this is used for sample rate conversion, or for
> when you can't synchronize word clocks because you don't have all the
> gozintas and gozoutas that you need.
>
> > Glenn Meadows and I argued about this for a long time back in 1989
because
> > audio people were just starting to talk about "jitter" and I already had
> > jitter problems with digital transmissions on token ring when moving up
to
> > 16 Mb/s and a "wild" chipset. He said it wasn't the same thing and I
said
> > it was. I was right. An external clock was established to correct the
> > problem and it worked fine, both in token ring and in digital audio.
>
> You were right in that you needed to synchronize the clocks. You were
> wrong to call it "jitter" though in 1989 it wasn't that common a term
> in audio.
>
> > Now when you talk about 9 pin ADAT sync, that's different because it
> > incorporates real "word clock". ADAT lightpipe does not incorporate
ADAT
> > sync, nor any other type of word clock.
>
> Wait a minute! That's what I've been talking about all along when I've
> said "ADAT sync." So we agree after all. At least in part.
>
> --
> I'm really Mike Rivers (mrivers@d-and-d.com)
> However, until the spam goes away or Hell freezes over,
> lots of IP addresses are blocked from this system. If
> you e-mail me and it bounces, use your secret decoder ring
> and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 5, 2004 12:01:12 AM

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

In article <4139ed86$0$19721$61fed72c@news.rcn.com> rnorman@starpower.net writes:

> But let's face it. ADAT, in terms of lightpipe, does not supply word clock.

Actually, it does. Embedded word clock is part of the protocol.
Anything that sends in ADAT Optical protocol (and uses the name) must
embed a word clock in the data stream.

The part that Alesis didn't insist on is that anything that receives
the ADAT Optical protocol has to be able to extract the word clock
from the incoming data. Most inputs do, but the one I'm most familiar
with, the Mackie DIO-8 and OPT-8 implementations, don't. They embed
the word clock in the output data stream like they're supposed to, and
a gen-u-wine ADAT (or something that correctly emulates it) can pull
word clock out and be set to use that rather than its internal word
clock source. But A Mackie ADAT input (with exception of that on the
later generation SDR24/96) does not extract word clock from an
incoming ADAT lightpipe. It must be synchronized with an external word
clock, or the Mackie word clock must be used to synchronize the word
clock of the device feeding it the lightpipe signal.


--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 6, 2004 5:38:19 AM

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

Mike Rivers wrote:
> In article <4139ed86$0$19721$61fed72c@news.rcn.com> rnorman@starpower.net writes:
>
>
>>But let's face it. ADAT, in terms of lightpipe, does not supply word clock.
>
>
> Actually, it does. Embedded word clock is part of the protocol.
> Anything that sends in ADAT Optical protocol (and uses the name) must
> embed a word clock in the data stream.

Strictly speaking, the protocol includes the necessary embedded timing
from which a Word Clock can be generated.

See the Alesis U.S. patent #5,297,181:

http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Se...

As an example, the Alesis AL1402 decodes ADAT Lightpipe and is able to
either generate a Word Clock signal from the incoming stream as an
output signal or synchronize to a Word Clock input (on the assumption
that it is already synchronized to whatever generated the ADAT optical
signal.)

http://www.wavefrontsemi.com/DataSheetsFolder/Wavefront...
Anonymous
September 6, 2004 2:19:18 PM

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

In article <413BBF89.9080304@audiorail.com> gwiebe@audiorail.com writes:

> >>But let's face it. ADAT, in terms of lightpipe, does not supply word clock.
> >
> >
> > Actually, it does.

> Strictly speaking, the protocol includes the necessary embedded timing
> from which a Word Clock can be generated.
>
> See the Alesis U.S. patent #5,297,181:

Understood. Of course the word clock itself is not carried over the Lightpipe,
only enough information for the receiver to derive a functional equivalent
(needed by the receiving device) is.

It may be interesting to read the patent, but it is not necessary in order
to understand, as I stated, that it's necessary to look at device setup
options to find out if using the ADAT optical input as a source for data
clock syncrhonization is available. Usually this is described in terms of
"use ADAT input for word clock."

A parallel would be to a device which can syncronize its data clock
from black burst video or an AES3 (AES/EBU or S/PDIF) input without
audio. What goes on inside the box is the desgner's business. The
function is what the user sees.


--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Anonymous
September 17, 2004 2:34:40 PM

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

Roger W. Norman wrote:

> Firewire is an APPLE thing, not Western Digital. It was ratified by IEEE
> under 1394 and is a MAC layer device protocol, hence why you can run
> proprietary upper layer protocols over it such as MLan or even ethernet if
> one wanted to write the code.

Anyone can implement an IEEE1394 interface, but if you want to call it FireWire you have to license the name from Apple.
!