Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Fernando Lucchesi]

You can.

Set the switch EXACTLY like this:

2 opto cables to connect the console to the ADA8000. No BNC cable required. You should get a solid sync.

-Ray

Hey, Ray!

How are you doing?

So, how should I set my 01V's WORD CLOCK?

[Scott Holtzman]

Fernando,

just keep 01v as master, switch ada to ADAT IN on the back and you're ok.

Briad Maddox brought this up in a previous point.

I think it is far more important, rather than simply giving tactical suggestions is to offer strategic advice.  If you are going to work with digital audio you need to understand some basics.  Then when confronted with the interconnection of a new system you can utilize the principals to create your clocking plan instead of hacking away until each device gets time from a possibly degraded source.

All synchronous digital signals require clocking.  A few terms need to be learned.  This clock is called a BITS clock, basically the same as a World clock but more key.  The BITS clock make sure that the receiving device is looking for the bit at the right place in the stream.  When timing is poor the bitstream "slips" and bits are lost to timing errors.  Significant artifacts result in the recovered analog signal.

A discrete clock is the best way to accomplish this but many don't want to go through the trouble.  One digital circuit (even a dry circuit without a signal) can be used as the master clock.  T1 lines are used to distribute clocking to devices in telephone central offices.  This is the exact same principal.  Those BITS clocks use nuclear decay as the standard.  Even the Apogee Big Ben is a PLL clock.  (for a bit of techno babble accuracy of clocks is expressed in Parts Per Million).

Assuming you don't want to designate your mixer or another device as the master clock and then distribute that clock signal to all connected devices (a great way to lower Jitter) or drop a large on Big Ben then you have to use what is called a recovered clock.

A recovered clock is a compromise.  While easy to use it provides sufficient clocking to recover the signal but would degrade if reproduced.  You can't take a device that uses a recovered clock as it's master then use it as a clock source for a downstream device.

Here is the last fifty cent word of the post,  plesiochronous (for close time in Latin) even the word root shows it is a less than optimal solution.  Embedded solutions use bit markers to recover timing.  Simply hashing or hamming code is used to find bit errors in the stream.  If you are using embedded timing at least use the device with the most stable clock as the upstream source of the embedded signal.

The more complex your digital system the more important clock design becomes in ensuring accurate transmission of the baseband data.

Once you understand the principals these tactical tips won't be needed.  You will be able to strategically design a clocking architecture that is accurate and reliable.  Your peers will be amazed with your mastery of a very important topic and most importantly your audience will thank you because jitter is contributory to listener fatigue.

I found this article, it is a nice summary:  http://www.presonus.com/community/Learn/Digital-Audio-Connections-and-Synchronization

[Ray Aberle]

Hey, Ray!

How are you doing?

So, how should I set my 01V's WORD CLOCK?

I BELIEVE it will stay the same to however you have it set now-- it should be the master, tho. When you had your switch set to "44.1KHz" - it's in the master side, so the ADA8000 was thinking IT was the master, as was the 01V -- hence the sync error.

-Ray

[Fernando Lucchesi]

Briad Maddox brought this up in a previous point.

I think it is far more important, rather than simply giving tactical suggestions is to offer strategic advice.  If you are going to work with digital audio you need to understand some basics.  Then when confronted with the interconnection of a new system you can utilize the principals to create your clocking plan instead of hacking away until each device gets time from a possibly degraded source.

All synchronous digital signals require clocking.  A few terms need to be learned.  This clock is called a BITS clock, basically the same as a World clock but more key.  The BITS clock make sure that the receiving device is looking for the bit at the right place in the stream.  When timing is poor the bitstream "slips" and bits are lost to timing errors.  Significant artifacts result in the recovered analog signal.

A discrete clock is the best way to accomplish this but many don't want to go through the trouble.  One digital circuit (even a dry circuit without a signal) can be used as the master clock.  T1 lines are used to distribute clocking to devices in telephone central offices.  This is the exact same principal.  Those BITS clocks use nuclear decay as the standard.  Even the Apogee Big Ben is a PLL clock.  (for a bit of techno babble accuracy of clocks is expressed in Parts Per Million).

Assuming you don't want to designate your mixer or another device as the master clock and then distribute that clock signal to all connected devices (a great way to lower Jitter) or drop a large on Big Ben then you have to use what is called a recovered clock.

A recovered clock is a compromise.  While easy to use it provides sufficient clocking to recover the signal but would degrade if reproduced.  You can't take a device that uses a recovered clock as it's master then use it as a clock source for a downstream device.

Here is the last fifty cent word of the post,  plesiochronous (for close time in Latin) even the word root shows it is a less than optimal solution.  Embedded solutions use bit markers to recover timing.  Simply hashing or hamming code is used to find bit errors in the stream.  If you are using embedded timing at least use the device with the most stable clock as the upstream source of the embedded signal.

The more complex your digital system the more important clock design becomes in ensuring accurate transmission of the baseband data.

Once you understand the principals these tactical tips won't be needed.  You will be able to strategically design a clocking architecture that is accurate and reliable.  Your peers will be amazed with your mastery of a very important topic and most importantly your audience will thank you because jitter is contributory to listener fatigue.

I found this article, it is a nice summary:  http://www.presonus.com/community/Learn/Digital-Audio-Connections-and-Synchronization

Thanks for the enlightment Scott. I'm not sure I understood everything you said, but I'll check that link you sent me and look more about it!

[Fernando Lucchesi]

I BELIEVE it will stay the same to however you have it set now-- it should be the master, tho. When you had your switch set to "44.1KHz" - it's in the master side, so the ADA8000 was thinking IT was the master, as was the 01V -- hence the sync error.

-Ray

This is how it's set right now... The message isn't showing up anymore, so I guess that was it...

[Scott Holtzman]

Thanks for the enlightment Scott. I'm not sure I understood everything you said, but I'll check that link you sent me and look more about it!

You are welcome. 

[Jano Svitok]

I don't understand Scott's post either (and I'm going to read that article) but on a more simplified level (that's what I've learned so far):

- when you connect devices in digital domain, you need to synchronize them so the receiver knows where the bits start in the incoming stream. the synchronization signal is called word clock
- one device is master, that is clocked to it's internal timer, and transmits the clock to the rest of the devices that are called slaves
- you can send WC over specialized coax cables with BNC connectors, or it's possible to recover clock signal from ADAT or SPDIF ("2TRD") signal (so in your case ADAT output from 01V carries clock signal to ADA8K, that is used to properly sync A/D converters so the signal from ADA8K is in sync with 01V)

- you can run the system without synchronization but you need to expect noise/clicks/lower quality (When our 01V96 was full and we needed to add CD, we used SPDIF connection even though the CD didn't have any WC input, and we didn't want to clock 01V to CD. It worked quite ok, but it was a special case)

[Fernando Lucchesi]

I don't understand Scott's post either (and I'm going to read that article) but on a more simplified level (that's what I've learned so far):

- when you connect devices in digital domain, you need to synchronize them so the receiver knows where the bits start in the incoming stream. the synchronization signal is called word clock
- one device is master, that is clocked to it's internal timer, and transmits the clock to the rest of the devices that are called slaves
- you can send WC over specialized coax cables with BNC connectors, or it's possible to recover clock signal from ADAT or SPDIF ("2TRD") signal (so in your case ADAT output from 01V carries clock signal to ADA8K, that is used to properly sync A/D converters so the signal from ADA8K is in sync with 01V)

- you can run the system without synchronization but you need to expect noise/clicks/lower quality (When our 01V96 was full and we needed to add CD, we used SPDIF connection even though the CD didn't have any WC input, and we didn't want to clock 01V to CD. It worked quite ok, but it was a special case)

I don't know if it is related, but I used my ADA8K last friday and recorded the whole gig in 16 channels through the USB port on my 01v. In some points of the recording everything gets out of tune really quickly (autotunes style) then mutes for around 0.5 second and come back... Would this be related to this Clock issue?

[Jano Svitok]

if all all channels are out of tune/muted (i.e. those from ADA8K and also those from 01V preamps) the problem is not in clocking. If only channels that come from ADA are bad, then there is clocking problem.

My guess is that it's a problem of USB/computer/HDD latency/overload. Either USB bus was saturated, or another device caused it to run at lower speed (when USB 1.1 device communicates, the whole bus is running at low speed), or the computer was not able to process (copy from USB to memory and then to hdd) incoming data fast enough.

Make sure you have connected to one USB root only 01V, and your mouse/keyboard/whatever else is on another USB root. You can check it in Device Manager in Windows ("Show devices by connection"). I don't know how can you affect where the device is connected -- you may try to replug to another USB port.

Other than that, look up any computer optimization guide targeted at audio/low latency (for start: disable antivirus, disable NTFS access times, make sure you have enough RAM, defragmented disk,...)

[Riley Casey]

Just to skip past all the other useful information, if you want this to work reliably day in and day out use the BNC clock connections and set the ADA8000 accordingly.  Clocking via ADAT works right up til when it doesn't.  Reliable clocking is worth a few dollars or Euros spent on a locking connector and a robust clock signal.  Want  to run multiple outboard devices all clocked from the console?  Invest in a clock distribution amp, they are relatively cheap unless you fall for the boutique clock voodoo.

wouldn't I need the WC IN cable?

[ProSoundWeb Community]