MOTU audio interface help - CLOCKING ISSUES >>??

[orcasound]

been successfully using set up in the studio for quite some time - now, of course, problems. - hate to be lengthy but - this one has got me puzzled - so I will try to be thorough

Setup is follows: G5 Dual 2.0 w/ PCI-424 / 2408mkIII, 24i/o, & 308. -
2408mkIII - BANK- A Optical i/o from Apogee Rosetta 800
BANK B Optical i from Presonus Digimax 8ch. mic pre
BANK C Analog out - (8ch. patched directly to dangerous Summing box)
24i/o MIDI gear is patched into INPUT 1-16
Ensoniq DP4 4chs INput and 2 Chs. Output are patched
308 AES I?O from 2 TCM3000's, 2 Eventide Eclipses
----
Apogee Rosetta 800 is the master clock and there is a BNC running out of Apogee to WCIn on 2408mk III, then OUt to the Presonus Digimax mic pre (the 24i/o and 308 subsequently receive WC from the PCI-424 card)
-----
All of the MOTU pieces, as well as the digimax indicate that they are receiving clock.
-------
I route stereo stems out from DP into a Dangerous Summing Box - the first 8 channels are routed through the Apogee OUTs via a patchbay, and channels 8-16,a s indicated above are directly patched fro the 2408mkIII to the summing box.
- Have recently noticed that when I assign outputs in DP to the summing box through the 2408 analog outs that those channels consistently have a noticeable click,pop kind of thing going on - if I change the assignments to channels 1-8 of the summing box through the Apogee - NO problem.
In addition - now when I am using the DP4 effects - as they are patched into the analog i/o of the 241/o - I am also getting the same type click/pop thing through the 24i/o

I have been using this set up for years wthout any issues like this - it sounds like a clocking issue - however all the units as i mentioned, react in a way that shows me they are getting appr. clocking -

Any help would be much appreciated - I've got a bunch of session work lined up - and this issue is really tieing my hands down in terms of what I am able to do - I do have audio clips that I made of the click noise - on a track that is passed out via the 2408 to the summing box - and also a clip from a track that is sent through the DP4 effects bus - that is patched into the 24i/lo - if anyone needs to hear what the issue is.

thanks very much - looking forward to your comments

[BradLyons]

SORRY,this is where I disagree about chaining clock or even using "T" connectors. The reality is that each box has it's own load of clock and what one likes, another may not. YOU NEED DIRECT DISTRIBUTION. Sorry, there's no other real world way around it. The fact it was working and now not doesn't mean it was ever TRULY working, but that it was working enough to hide the issues. When chaining, that can happen. IMHO, look into a Big Ben---all of these issues will not only go away, but your gear will even sound better (even the Rosetta).

[1nput0utput]

Current MOTU interfaces regenerate the clock signal to their word clock outputs regardless of what the clock source is. So there's nothing wrong with connecting the 2408mk3 word clock output to the DigiMax's word clock input. The quality of the clock signal at the output of the 2408mk3 should be identical to the signal at the output of the Rosetta 800.

[orcasound]

thanks - I really have to say that I never looked at my set-up as "chaining" the word cock - The Apogee WC OUT goes to the 2408 WC IN / (which is connected via audiowire to the PCI-424 card) that card then makes sure that the MOTU 24i/o and MOTU 308 receive the necessary WC via their "audiowire" cables from the PCI-424 //that is- or at least thats what I've been told is "supposed" to happen - without much issue.
* The only "thru" or chainging that is done - is to the digimaxLT - that comes out of the 2408 and into the digimax - but the Digimax is working fine - Not to mention, I hardly ever use that piece - its only set up to be able to drop a quick 8ch. drumkit set from the live room -

[BradLyons]

Current MOTU interfaces regenerate the clock signal to their word clock outputs regardless of what the clock source is. So there's nothing wrong with connecting the 2408mk3 word clock output to the DigiMax's word clock input. The quality of the clock signal at the output of the 2408mk3 should be identical to the signal at the output of the Rosetta 800.

Right, however there are MULTIPLE devices connected digitally and THERE in-lies the problem.

[orcasound]

thanks Brad for the comments - i think i am still not "on board" with what you are saying - ....
for S..and.g.... - I took the presonus digimax out fo the equation - leaving everything as follows:

Rosetta 800 Word Clock out to 2408 MkIII Word Clock In - thats the only BNC Cable in the setup.
The 2408 runs via audiowire to the PCI-424 card -
the 241i/o and the 308 both run via audiowire to the PCI-424.

Interface is set to 2408 Word Clock In as clock source.
According to MOTU - the clock signal is then sent via 424card to the 308 and 24i/o - nothing is chained - and still the same issue.

[BradLyons]

Clocking is something that is involved with EVERY digital piece of gear, but few really understand that or why. I remember back in 1998 with a loaded EMU Paris system (one of the biggest in the world, actually) and had major issues because of word-clock. I thought that had to do with video, boy was I wrong. Within every digital device there is a timing structure because it's digital. For example for every second of 44.1kHz audio there are 44,100 samples or "pictures" taking place. For every second of 96khz audio, there are 96,000 samples and so on. I say "picture" as an illustration that the better the camera, the better the picture. However things like lighting can severely alter the image. I do TV broadcast....when I'm sitting in the room we're recording I can see plain as day without a lot of light, but that same room on my tv monitors is pitch black without extensive lighting. With high-level, bright white lights....everything is illuminated. However without floor lights shooting up, you see shadows under the eyes. The lighting is fine, but it's not perfect.

Where am I going? Well in a digital audio stream there are these samples---but how those samples are quantized within each second can, will, and absolutely DOES have an impact on the audio quality. Every digital device has a clock, very few are good. I've proven this even on a $20,000 Sony DMX-R100 digital console and nearly every other digital converter. VERY few have ideal clocking. The better in-line those samples per second are, the less garbage is in the way. This garbage is known as jitter. Contrary to popular belief, jitter is NOT just pops and clicks. The reality is, there is jitter in your system right now--the question is how much of it do you wish to reduce? When you put a great master clock on the gear you have, you WILL notice a better and more improved sound across the entire frequency. The low-end will have more punch and depth, the highs are more clear with a wider stereo image. Technically a great clock doesn't make your gear sound better, it over-rides the internal clock allowing it to perform to the highest quality that it is capable of. Another illustration..... and a strange one, but it makes sense. IMAGINE....there are 44,100 people walking in sync with each other, right leg forward, one second later left leg forward, another second right leg, etc. Imagine each person was wearing tap shoes so you could hear them walking. Well each person is walking in-time, but how they interperate (spelling?) that time is up to them. Within a second there are many smaller intervals. You would here, for the most part, everyone in time---but not perfect time. Now imagine these 44,100 people had loose chains strapped to their legs. If one moves to fast it will pull the other. Again, relatively the same timing but not perfect. NOW imagine you had 2x4 wood plans attached to each leg one in front to one behind, each the same exact size. This would provide everyone to be walking in PERFECT sync, the sound would be total harmony. THAT is what a great clock does.... it gets rid of all of the timing errors and inconsistencies within the digital stream which, in the end, allows the converter to run to its full potential and provide better audio quality as a direct result. Let's take a MOTU HD192----great interface! It's very clean with amazing top-end clarity and stereo imaging, the lows are quite good as well---not as sterile as some converters, but not as warm as Apogee. Out of the box, this thing sounds great! BUT put an Apogee Big Ben or an Antelope Audio OCX on it and you will hear the full capability of this box. You can even do your own test by recording floor noise and compare the results, it's noticeable.

OKAY, I'm NOT saying one should buy a clock just for better sounding converters...... that is usually the last purchase advice I would ever give. HOWEVER there are times when one needs a clock because so many different digital devices connected. As they say, you can only have one master.... The same is true in digital audio connectivity. There can only be ONE SOURCE of master clock signal just as in syncing tape-heads there can only be one master. Everything else must slave. When you have (2) devices connected, no problem. When you have (3) usually that's not an issue either...but when you get to (4) or more, chances are you need a word clock!

In my current studio I have a ProTools HD3 Accel with a single 192I/O. The converters in this are phenominal, the clocking is pretty good too. Would I see an advantage of one of those above clocks? Yes, but not a lot in this case. HOWEVER I have plans to add another computer that will have 32-channels of ADAT Optical, feeding to this HD system where I'll be adding a 192Digital for 32-channels of ADAT Optical. There will be a third computer that will have 8-channels of TDIF feeding into that 192I/O, likewise I'll have ADAT Optical going from an Apogee Ensemble to one of these interfaces and eventually a MOTU V4HD with AES/EBU connected to the 192Digital. IF I DID NOT ADD A MASTER CLOCK INTO THIS SYSTEM AS WELL, IT WOULD BE TOTAL AND UTTER CHAOS!!!!!!! In my live setup I have the following:

Digidesign Venue DShow console
ProTools HD3 Accel w/96I/O
Presonus DigimaxLT with ADAT out into 96I/O
MOTU 2408 with an Alesis HD24 connected via 24-channels ADAT Optical
Panasonic DVCPro master video recorder x2
(2) HHB CD Burners

EVERYTHING IS CONNECTED DIGITALLY IN SOME WAY SHAPE OR FORM. I have an Apogee BigBen that is distributing clock from the Digidesign SyncI/O while the SyncI/O is also taking house video sync. Again, without that clock---total chaos!! Digital gear requires clocking, even a single device like an 828. However you won't truly run into it until you run into it, if you know what I mean. A company I greatly respect says to use "T" connectors, but I couldn't disagree more. Every device has a different capacitance of load and therefore terminates the signal differently. Likewise if you add another piece of gear in, so you must re-wire. AND further more.....this is dividing signal and you lose the clocking quality. There is a difference between a "Y" and a "SPLIT", big difference. That is why master clocks have multiple outs, to feed a direct connection and load to every device in the chain that requires it. I've never seen a pro studio configured with connectors, only direct connections.

There are other myths about "how far" you can run. Well, my Venue FOH Rack is out in the sanctuary, my Stage Racks are back in the equipment room near me and my video room is about 50ft from there. I have a 250ft cable run between front of house and the equipment room, about 75ft to the video room, and multiple runs of various lengths in the studio itself. I don't have as much as a single glitch, pop, drop-out, etc.

Does this make sense?

[orcasound]

Brad - thanks for the very detailed explanation - it does make sense.
But what doesnt make sense is that I went into the studio tonight to do some troubleshooting - turned everything on - and boom... the issues were gone.
The clicking from the DP/4 through the 24i/o -not there - and the clicking that i was getting on the channels that were being sent out of the 2408 analog outs into the dangerous summing box - were non existent. - this is good - in that I was not looking forward to spending $$$ for the Big Ben, right now.... however it concerns me - and brings up another point - there must be something going on - I would suspect most likely in my Mac or with the PCI-424 card. - I just purchased a mac on ebay - Dual 2.3 G5, as a spare for my Dual 2.0 G5 - My current G5 was bought form Apple in 04 - and in 06 - while we were building the new studio - I brought the Apple home and sat it in my dining room for safe keeping - 2 weeks later - we had a horrific house fire - the mac endured 1000degree heat - and was covered in soot inside and out - my tech guy at the studio cleaned it up - where he could - but if you open it up - the mother board etc is covered in soot - we booted it up in March 0f 07 - and it worked like a charm - just smelled a bit funny - but my 2 UAD-1's and the PCI-424 were covered in soot - and are still quite messy - so perhaps - something happened - BUT - you would have thought that it would have happened a lot sooner than 2 years after the fire - no?

Kevin
orca sound

[BradLyons]

Welcome to the studio world! What doesn't work, will.....what works, won't. They are called GREMLINS!!!!! It's like when I'm mixing live and I go to solo a channel to listen, and THEY STOP PLAYING at the same time....argh!