Distributing word clock: Big ben vs Lynx vs distro with clock in AD16

[Mario]

(I posted this 3 days ago but it must of not gone through b/c I cannot
find it after 20 minutes of searching)
Hello Again
I have bought a AD-16 and have a swissonic a/d for the other 8 channels
going to a mx-2424. I have a Kramer word clock distribution amp. In
theory I have everything I need to make things work (Running the ad-16
clock as master to the kramer and out to all other gear) but I am
concerned about jitter and am considering buying a external word clock
generator.

Now, I know some suggest T-Connectors or a distribution amp like the
one I have and I would love to hear from you but I want to focus my
question on this: SHOULD I BUY A BIG BEN OR A LUCID GENX96 considering
that if things go well I will upgrade and eventually get a ad16x or
better that would have an onboard word clock just as good (or better)
as the one in the big ben, and then I would have to distribute that?
In other words, will the lucid beat the intelliclock (Not the c777 on
the ad16x) on the ad16 especially considering it will have to be
distributed through the Kramer unit? Would it be a better investment
(though slightly lower in quality) than the big ben because it could
then serve as a distribution point later (though I am aware I could do
the same with the Big Ben), using it to distribute the clock on the
ad16x or better unit? I have no problems with the limitations of 96k
the lucid has because I probably won't make the jump to 192 for a long
time.

Or is it best just to make the jump to the Big Ben? Is it (the BB) that
much better than the genx96? Will the genx96 not make much of a
difference over the intelliclock? Can the Kramer handle the job until I
can afford the BB? Also what cable/connector do you suggest? Are the
apogee cables worth it?

I know these are a lot of questions. Any answers to any of the
questions, any suggestion or comments would be greatly appreciated.
Thanks!
Mario

[[email protected]]

Mario wrote:

> question on this: SHOULD I BUY A BIG BEN OR A LUCID GENX96 considering

One data point: I've been using a GenX6 (the original) and have not
heard a compelling reason to 'upgrade'. All outputs are used driving
digital mixers, outboards, converters, etc. Have had a GenX6-192 in for
evaluation and it does have some interesting features that I may want
in the future but I can still live with the original. The clock is
stable and it has been working daily for years.

bobs

Bob Smith
BS Studios
we organize chaos
http://www.bsstudios.com

[Mario]

THanks Bob.
Here are the stats on the Krame distribution box. It is made of
composite video, but I assume it is the same.
INPUT: 1 video, 1Vpp/75Ω on a BNC connector.
OUTPUTS: 5 video, 1Vpp/75Ω on BNC connectors.
MAX. VIDEO OUTPUT: 2 Vpp.
VIDEO BANDWIDTH (-3dB): 280MHz.
DIFF. GAIN: 0.1%.
DIFF. PHASE: 0.1 Deg.
K-FACTOR: 0.05%.
S/N RATIO: 80dB.
CONTROL: Up to +6dB.
COUPLING: AC.
POWER SOURCE: 12 VDC, 100mA.
DIMENSIONS: 12cm x 7.5cm x 2.5cm (4.7� x 2.95� x 0.98�, W,
D, H.).
ACCESSORIES: Power supply, mounting bracket.
OPTIONS: Model VA-50P power supply with six 12VDC outlets. RK-T1,
RK-T3 19� rack adapters.

[animix]

I have been using the original GenX6 for years. I have it clocked to my
Mytek A/D converter and it can serve as a standalone clock or a distro amp
(similar to the Kramer, I would guess).

"Mario" > wrote in message
oups.com...
> (I posted this 3 days ago but it must of not gone through b/c I cannot
> find it after 20 minutes of searching)
> Hello Again
> I have bought a AD-16 and have a swissonic a/d for the other 8 channels
> going to a mx-2424. I have a Kramer word clock distribution amp. In
> theory I have everything I need to make things work (Running the ad-16
> clock as master to the kramer and out to all other gear) but I am
> concerned about jitter and am considering buying a external word clock
> generator.
>
> Now, I know some suggest T-Connectors or a distribution amp like the
> one I have and I would love to hear from you but I want to focus my
> question on this: SHOULD I BUY A BIG BEN OR A LUCID GENX96 considering
> that if things go well I will upgrade and eventually get a ad16x or
> better that would have an onboard word clock just as good (or better)
> as the one in the big ben, and then I would have to distribute that?
> In other words, will the lucid beat the intelliclock (Not the c777 on
> the ad16x) on the ad16 especially considering it will have to be
> distributed through the Kramer unit? Would it be a better investment
> (though slightly lower in quality) than the big ben because it could
> then serve as a distribution point later (though I am aware I could do
> the same with the Big Ben), using it to distribute the clock on the
> ad16x or better unit? I have no problems with the limitations of 96k
> the lucid has because I probably won't make the jump to 192 for a long
> time.
>
> Or is it best just to make the jump to the Big Ben? Is it (the BB) that
> much better than the genx96? Will the genx96 not make much of a
> difference over the intelliclock? Can the Kramer handle the job until I
> can afford the BB? Also what cable/connector do you suggest? Are the
> apogee cables worth it?
>
> I know these are a lot of questions. Any answers to any of the
> questions, any suggestion or comments would be greatly appreciated.
> Thanks!
> Mario
>

[gunnar]

You should be aware that jitter is not quite as simple as it might seem
at first. There is a lot of magic and marketing going into placing that
clock tower in the mind of people. Not all of that is translated into
the real world, and not all of that is appearant as reduction in
jitter.

According to several authorities, one example is Dan Lavry, any quality
AD will have more jitter driven by an external clock. The reasoning is
that the clock circuits inside that box is soo much better at creating
a good clock than all the stuff involved in transferring word clock
between boxes. So if the jitter is reduced, the only way it could
happen is that you started with a really bad AD from the start. Now,
less jitter may not sound better, but that is a completely different
discussion. In that respect, as they say, your mileage may vary.

So if least jitter is what you are after, let your best AD converter
run on internal clock and run your most important signals through that
one. This means that you would take the word-clock from that box and
distribute. Distributing word-clock on T-connectors is perfectly valid
and does not add jitter as long as you follow the design (exactly one
termination at the end of the chain, not too long cable, not too many
units) .

There might be completely different reasons why you want an external
word-clock generator. It has a lot to do with utility and functions,
such as easily changin sample rater or synching to external sources.

And of course, from some reason, one of the manufacturers work hard in
making you believe differently. It is your ears though. And your money.

Gunnar

[Mario]

Gunnar,
Thanks for your input. Here is the problem. I am using multiple ADs
(2 at least). Should they not be all be synced? I also have a
multitrack recorder and a digital mixer. Should not all these devices
be synchronized? If so I have to have a word clock distro? If I am
incorrect, please correct me.
Thanks,
Mario
gunnar wrote:
> You should be aware that jitter is not quite as simple as it might seem
> at first. There is a lot of magic and marketing going into placing that
> clock tower in the mind of people. Not all of that is translated into
> the real world, and not all of that is appearant as reduction in
> jitter.
>
> According to several authorities, one example is Dan Lavry, any quality
> AD will have more jitter driven by an external clock. The reasoning is
> that the clock circuits inside that box is soo much better at creating
> a good clock than all the stuff involved in transferring word clock
> between boxes. So if the jitter is reduced, the only way it could
> happen is that you started with a really bad AD from the start. Now,
> less jitter may not sound better, but that is a completely different
> discussion. In that respect, as they say, your mileage may vary.
>
> So if least jitter is what you are after, let your best AD converter
> run on internal clock and run your most important signals through that
> one. This means that you would take the word-clock from that box and
> distribute. Distributing word-clock on T-connectors is perfectly valid
> and does not add jitter as long as you follow the design (exactly one
> termination at the end of the chain, not too long cable, not too many
> units) .
>
> There might be completely different reasons why you want an external
> word-clock generator. It has a lot to do with utility and functions,
> such as easily changin sample rater or synching to external sources.
>
> And of course, from some reason, one of the manufacturers work hard in
> making you believe differently. It is your ears though. And your money.
>
> Gunnar

[gunnar]

Hi Mario,
yes they should all be syncronized. But in many cases simple
daisy-chaining of the word-clock is what the word-clock system was
designed for and what will work and what you can use. There is not
necessarily any need for any other word-clock distribution than T
connectors, cables and a termination resistor.

So here is what I would do as a starting point -- this is going to be
the reference setup. If you want to try with distributors or external
word-clocks do compare against this.

* Your main AD should be master clock source, running on its internal
clock.
- take the signal from the output WC connector, use a short cable of
the right sort. Remember wc is 50 Ohm (not the more common 75 Ohm)
* Put a T-connector on the input of your number 2 AD converter.
- one end of the goes to the main AD, the other end to next unit in the
chain.
- make sure WC termination is set to off on (there might be a small
switch)
*Now continue putting T connectors on the rest of the units. Use fairly
short cables.

* The last unit should have a termination resistor. Either turn on the
in-built one or use a T plus termination resistor.

Now a bit of theory:
A word-clock distribution is setup so that the master sends a signal
down the cable. There can be many slaves listening to the signal. At
the end of the cable the signal is "eaten" by the terminating
resistance in order to not bounce back and create interference. Every
word-clock cable should be terminated at the end, and only there (well,
the master sender has termination built-in, but now we are getting real
technical). Often though the system is inherently so stable that it
works without the termination, not recommended though. It is not a
good idea to have several terminations on the same WC distribution,
although in practice two seems to work fine, again not recommended.

The practical problem with T connectors is that they might not fit at
the back of many kinds of equipment. This may be a reason to invest in
a distribution unit. You may elect to use really short runs of cables
from the stem end of the T. Another reason may be if your main AD does
not switch easily to the sample rates you want, then it might be more
convenient with a dedicated wc generator. Yet another reason is if you
tend to change your rig often. That WC generator will probably stay in
the rack even when you change AD-s and other units. You pay a bit of
money for a dedicated WC generator though.

So with a bit of knowledge you may be able to save some money and get
things working and not drop any quality. Once there you may do a proper
A/B test with whatever WC generator and see if it really is worth it in
terms of improved sound. Some people claim to hear the difference every
time, some not, it seems to be a bit of religion involved, but there
are many factors.

I will add a little twist to this. Remember that a purely digital piece
of equipment is not affected by jitter. Jitter only comes into play at
AD or DA conversion. Say if you have an ADAT recorder without AD and
DA. This simply records bits and is very immune to even extreme amounts
of jitter, they simply do not modify the bits. So I tend to simply
clock them on the ADAT signal. The ADAT signal is selfclocking, but the
jitter can be a little bit higher than a dedicated WC connector. It
does perfectly for synchronizing though. Now give the AD-s and DA-s a
good word-clock and the ADAT may jitter along all it wants.

What you should avoid doing is going through any box. I have seen
people connect to the IN terminal and then chaining from the OUT
terminal. This degrades the signal by addind jitter to the WC (but it
still keeps things in synch) But again, if the receiving end is a
purely digital thing, it really makes no difference if there is a bit
of jitter. In my rig I synch the purely digital stuff from one of these
OUT connectors on a WC slave unit, no problems.

Mario wrote:
> Gunnar,
> Thanks for your input. Here is the problem. I am using multiple ADs
> (2 at least). Should they not be all be synced? I also have a
> multitrack recorder and a digital mixer. Should not all these devices
> be synchronized? If so I have to have a word clock distro? If I am
> incorrect, please correct me.
> Thanks,
> Mario

[Mike Rivers]

Mario wrote:

> I am using multiple ADs
> (2 at least). Should they not be all be synced? I also have a
> multitrack recorder and a digital mixer. Should not all these devices
> be synchronized?

Yes, as long as they're interconnected. It's not only a good idea, it's
the only way to assure that you won't get clicks when playing one into
the other.

> If so I have to have a word clock distro?

Yes, or the equivalent - a master clock generator that has multiple
outputs. The difference between that and a "distro" is that the
distribution box has one input and multiple outputs where the master
clock (which may be equipped with an input) has a built-in clock
source.

As has been pointed out, there's no reason why, unless you're working
with real junk, an external word clock should make a device sound
better than its internal clock, but with a good design, it shouldn't
sound noticably worse. I don't recall hearing any reports that someone
connected an outboard word clock and found that their A/D converter
didn't sound as good as with the internal clock, but I suppose these
things are pretty subtle.

There are those who have reported that things have sounded better with
an external clock, however. This is one of the Big Ben's major selling
point - that it improves the sound of everything to which it's
connected. This is a pretty bold claim, but famous ears and even the
wife of an infamous ear who was listening in another room said it
sounded better.

When I asked an Apogee engineer how this could be, I got a bit of what
might be mumbo-jumbo about how they recognize that there's no getting
away from jitter, so they control theirs and its randomness is what
makes things sound so good. Now I can see that if there was a
consistent or cohernet modulation to the clock, this could work its way
into changing the audio output. I'm not sure about the "controlled
randomness" explanation though. It might just be fuggin' magic. Or
fuggin' bull****. You gotta listen. Most dealers of products like this
will let you try it without committing to a purchase (either as a demo
loaner or willingness to give a full refund), so get one and listen to
it.

[Mario]

Thanks guys! I really appreciate the time you took to write your
responses. I have decided to buy the big ben and do some a/b tests
(though that may be hard because I mailly track and don't have lots of
d/a. If the daisy chaining works fine I will sell the big ben (I got
it used for $1025 shipped). I will post the results here. Gunnar, I
did not know that word clock was not a necessity between pieces of gear
joined digitally. SO if I am just running adat in to my mixer and
spdif out to my stereo recorder (to provide a rough mix CD after the
show to complement the 24 track recording), neither of them (the mixer
or the masterlink) will need external word clock? That makes sense
because the masterlink does not have a word clock in.
Regards,
Mario


Mike Rivers wrote:
> Mario wrote:
>
> > I am using multiple ADs
> > (2 at least). Should they not be all be synced? I also have a
> > multitrack recorder and a digital mixer. Should not all these devices
> > be synchronized?
>
> Yes, as long as they're interconnected. It's not only a good idea, it's
> the only way to assure that you won't get clicks when playing one into
> the other.
>
> > If so I have to have a word clock distro?
>
> Yes, or the equivalent - a master clock generator that has multiple
> outputs. The difference between that and a "distro" is that the
> distribution box has one input and multiple outputs where the master
> clock (which may be equipped with an input) has a built-in clock
> source.
>
> As has been pointed out, there's no reason why, unless you're working
> with real junk, an external word clock should make a device sound
> better than its internal clock, but with a good design, it shouldn't
> sound noticably worse. I don't recall hearing any reports that someone
> connected an outboard word clock and found that their A/D converter
> didn't sound as good as with the internal clock, but I suppose these
> things are pretty subtle.
>
> There are those who have reported that things have sounded better with
> an external clock, however. This is one of the Big Ben's major selling
> point - that it improves the sound of everything to which it's
> connected. This is a pretty bold claim, but famous ears and even the
> wife of an infamous ear who was listening in another room said it
> sounded better.
>
> When I asked an Apogee engineer how this could be, I got a bit of what
> might be mumbo-jumbo about how they recognize that there's no getting
> away from jitter, so they control theirs and its randomness is what
> makes things sound so good. Now I can see that if there was a
> consistent or cohernet modulation to the clock, this could work its way
> into changing the audio output. I'm not sure about the "controlled
> randomness" explanation though. It might just be fuggin' magic. Or
> fuggin' bull****. You gotta listen. Most dealers of products like this
> will let you try it without committing to a purchase (either as a demo
> loaner or willingness to give a full refund), so get one and listen to
> it.

[Mike Rivers]

Mario wrote:

> If the daisy chaining works fine I will sell the big ben (I got
> it used for $1025 shipped). I will post the results here.

Did we say anything about daisy-chaining? You did mention T connectors
- that's daisy-chaining, but I didn't think you were seriously
considering it. That can work for a limited number of devices, but you
need to be aware of whether inputs are terminated (most are),
unterminated (rare on new equipment), or switchable (as rare as
unterminated). The rule is that you only want the last device on the
chain terminated. If you have two devices that have terminators that
you can't turn off (or modify) you'll lose too much level and the word
clock signal may not trigger all of the devices on the chain.

Your Kramer should work just fine the way you have things set up now
(assuming all of the word clock inputs are terminated - they probably
are). Using Tees and a daisy chain won't sound any better and will
almost always work worse.

The thing to listen for with the Big Ben is whether it sounds better
than the Kramer distributing the AD-16 word clock output. If it sounds
$1025 better, then keep the Big Ben. Otherwise, keep what you have.

> Gunnar, I
> did not know that word clock was not a necessity between pieces of gear
> joined digitally.

You may have misinterpreted something. Devices with AES/EBU and S/PDIF
(stereo) inputs can derive word clock from their digital input source,
as can some TDIF and ADAT optical devices. But it's really best not to
count on it. That's why we have word clock distribution amplifiers and
word clock generators with multiple outputs. It's a problem-solver, and
without a single clock source feeding everything on separate branches,
some time, you'll have a problem.

> SO if I am just running adat in to my mixer and
> spdif out to my stereo recorder (to provide a rough mix CD after the
> show to complement the 24 track recording), neither of them (the mixer
> or the masterlink) will need external word clock?

If the mixer can get its word clock from the ADAT, and you set the
Masterlink to use the digital input for word clock (I think this is
automatic), then, no, you don't need separate word clock connections.
The Masterlink likes to think it's a master since it doesn't have a
word clock input (which would actually be a very useful thing).

[gunnar]

Mario wrote:
> i did not know that word clock was not a necessity between pieces of gear
> joined digitally. SO if I am just running adat in to my mixer and
> spdif out to my stereo recorder (to provide a rough mix CD after the
> show to complement the 24 track recording), neither of them (the mixer
> or the masterlink) will need external word clock? That makes sense
> because the masterlink does not have a word clock in.

To be strict they need to be word-clock synched. As both adat and spdif
does include embedded word-clock that can be a very valid solution.

As you had a good price on the BB do work with that one a bit. It is a
very nice box and adds a lot of utility to rack of gear through its
options and connections. I am a bit sceptical to Apogee-s claims at
times, but regardless it sure is a good box. It might not decrease
jitter in every setup (which you migth believe from the marketing
material) but if you like the sound better, then it is better.

Gunnar

[Steve Maki]

On 20 Aug 2006 00:53:33 -0700, "gunnar" > wrote:

>- take the signal from the output WC connector, use a short cable of
>the right sort. Remember wc is 50 Ohm (not the more common 75 Ohm)

You sure? Most of the info I've seen indicates 75 Ohm, and that's
what I've always used. Granted, on short runs it shouldn't matter
too much.

Steve Maki

[gunnar]

PRESS STOP.

Sorry, I slipped in an earlier message. Word-clock cable should of
course be the 75 Ohm type. I wrote something different.

Gunnar

[Mike Rivers]

Steve Maki wrote:
> On 20 Aug 2006 00:53:33 -0700, "gunnar" > wrote:
>
> >- take the signal from the output WC connector, use a short cable of
> >the right sort. Remember wc is 50 Ohm (not the more common 75 Ohm)

I think that Roger Nichols may have started that rumor many years ago,
back when he was writing a monthly column for EQ. Perhaps back when
word clock I/O on black boxes was pretty rare, he may have measured
something that was terminated in 50 ohms, or perhaps he was basing this
around his own "Digital Atomics" setup which used a Hewlett Packard
cesium beam frequency standard as a word clock source. Being standard
lab equipment, this would have had a 50 rather than 75 ohm output.

> You sure? Most of the info I've seen indicates 75 Ohm, and that's
> what I've always used. Granted, on short runs it shouldn't matter
> too much.

It's rare that you see any info about this, but generally you can
measure inputs with an ohm meter (it may require that the equipment is
turned on), or determine it by looking at the schematic. For an output,
you can send the clock to a meter or scope with a T in line, then put a
terminator on it. The amplitude will drop to half when the terminator
is the same as the output impedance.

Every word clock source and input that I've encountered has been set up
for 75 ohm connections. (With the exception, of course, of
unterminated inputs)

[Mario]

WOW! This thread has been very educational. Thanks for everyone
contributing. It is surprising that this topic can be contentious but
I must give you guys props for you guys not getting nasty, which is a
strange phenomenon on these types of forums. I had 50 ohm cables and
have just ordered the 75 ohm for the WC and the spdif. I hope they get
here in time. Maybe that is why I was getting clicks and pops when I
retired my system a few years back. I hope to give everything a test
run on Thursday. If I do I will let you all know how it goes.
Thanks,
Mario

Mike Rivers wrote:
> Steve Maki wrote:
> > On 20 Aug 2006 00:53:33 -0700, "gunnar" > wrote:
> >
> > >- take the signal from the output WC connector, use a short cable of
> > >the right sort. Remember wc is 50 Ohm (not the more common 75 Ohm)
>
> I think that Roger Nichols may have started that rumor many years ago,
> back when he was writing a monthly column for EQ. Perhaps back when
> word clock I/O on black boxes was pretty rare, he may have measured
> something that was terminated in 50 ohms, or perhaps he was basing this
> around his own "Digital Atomics" setup which used a Hewlett Packard
> cesium beam frequency standard as a word clock source. Being standard
> lab equipment, this would have had a 50 rather than 75 ohm output.
>
> > You sure? Most of the info I've seen indicates 75 Ohm, and that's
> > what I've always used. Granted, on short runs it shouldn't matter
> > too much.
>
> It's rare that you see any info about this, but generally you can
> measure inputs with an ohm meter (it may require that the equipment is
> turned on), or determine it by looking at the schematic. For an output,
> you can send the clock to a meter or scope with a T in line, then put a
> terminator on it. The amplitude will drop to half when the terminator
> is the same as the output impedance.
>
> Every word clock source and input that I've encountered has been set up
> for 75 ohm connections. (With the exception, of course, of
> unterminated inputs)

[Mike Rivers]

Mario wrote:

> I had 50 ohm cables and
> have just ordered the 75 ohm for the WC and the spdif. I hope they get
> here in time. Maybe that is why I was getting clicks and pops when I
> retired my system a few years back.

Sorry, but that's not too likely to be the cause of your problems
unless they were really long, but the standiing waves that result from
the impedance discontinuity (at the boundary between the cable and the
devices) can cause jitter. But again, it's not serious at the lengths
that you're likely to have in a one-room setup.