1 (edited by Immersion 2013-04-11 02:53:03)

Topic: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Yes I wonder, since my monitor controller Dangerous audio source have no word clock input but AES input,  but rest of my system for my AES inputs converter, effect processor etc have word clock. 

Is it possible, to set up the RME HDSPe AES to use steadyclock for the AES output signal to my monitor controller and use Wordclock for input signals only from my converter and effect processor ?

To clarify.

RME AES port 1 (input) = Word clock
RME AES port 2 (output) = Steadyclock.

I want to minimize the amount of jitter in my chain.

For my Input chain, converter, effects etc, I would want to use Word clock just because I think it is better,
and if I have word clock function I should use it If I can that is my opinion.

Regarding my AES output to my monitor controller from RME ,  the Burr Brown chip
in the DAC is said to solve a lot of the jitter problem to so both steadyclock and Burr Brown chip I think will work good.
This is what Dangerous audio told me when I asked why their product had no Word clock

"The AES receiver by Burr Brown does re-time the data stream by 128 samples
to knock the bit jitter down to the block jitter territory. After lots of
listening tests, we have determined that this is a good compromise between
reducing jitter related distortion products and keeping the delay time down
so you can record with our products in real time.""

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

AES/EBU carries clock, so no need for a word clock cable. Sync your converters via work clock, set the RME card to sync to AES input (from converters) and set your Dangerous device to sync to AES input (from RME).

Regards,
Jeff Petersen
Synthax Inc.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Jeff wrote:

AES/EBU carries clock, so no need for a word clock cable. Sync your converters via work clock, set the RME card to sync to AES input (from converters) and set your Dangerous device to sync to AES input (from RME).

I understand also that there is not a "need" to use word clock, however it should be used if possible if sound quality is of your concern.
So I guess it is not possible to use steadyclock and word clock simultaneously for each operate  AES port ?
I invested a lof money in my audio chain such as Lynx converter and soon also the eventide h8000fw effect processor.

BTW I just edited my post above.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

For one, using Word Clock will not improve "sound quality" at all. And I assume you have some misconceptions about Steadyclock - what do you mean when you say you want "to use steadyclock for the AES output signal to my monitor controller"?


Regards
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

5 (edited by Immersion 2013-04-11 15:14:45)

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

RME Support wrote:

For one, using Word Clock will not improve "sound quality" at all. And I assume you have some misconceptions about Steadyclock - what do you mean when you say you want "to use steadyclock for the AES output signal to my monitor controller"?


Regards
Daniel Fuchs
RME

Yes this is my first time I am using AES, and also word clock.

What I mean is.

(Steadyclock) RME AES OUTPUT---->Dangerous audio source
(Word clock) RME AES INPUT<------<Lynx urora------<---Eventide H8000FW 

obviosuly it would be a compromise since I would prefer to use word clock on all digital devices.
But the Dangerous source monitor controller do not have a word clock.

Yes it might not improve sound quality but it might make the signal more clean and less jitter.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Word clock is not the best! It is simply the most universal. Nothing worse then a false sense of safety.
The clock in AES is not inferior to WC. Listen to RME pros....

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
BFpro fs, 2X HDSP9652 ADI-8AE, 2X HDSP9632

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

vinark wrote:

Word clock is not the best! It is simply the most universal. Nothing worse then a false sense of safety.
The clock in AES is not inferior to WC. Listen to RME pros....

It sounds wonderful!
maybe it is myth then, cause when I read on other forums I alwawys here the same
that running the clock through the AES cable will cause jitter and that is why the Word clock is superior.

But if you tell me that steadyclock is not inferior to the word clock it must be some truth in that.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

What's your master clock?  The one and only place Jitter is a concern is at your primary AD Converters - and more and more people "in the know" recommend using your primary ADC's internal clock as your system master (assuming you can get that clock reliably distributed to the remainder of your system).

Your Dangerous DAC should easily handle any reasonable jitter as it very likely re-clocks anyways (any good DAC generally will), and any digital-to-digital transmissions (AES>AES, etc) can not be affected by jitter unless it approaches 1/2 a sample's worth of jitter or more (not generally possible unless you have a malfunctioning unit somewhere).

You might have read this "Do I need a master clock" article on SOS - but they generally agree most converters sounded best on their internal clocks:
http://www.soundonsound.com/sos/jun10/a … clocks.htm

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

In my research, I have seen most people recommend using the Converter as masterclock,  but then I can't use RME steadyclock though but Lynx Syncroclock.
But I read also some one yesterday said that it will sound better using the steadyclock and rme As Masterclock is it true ?
When reading about clocking otherwise almost all people will tell you use the your converter as masterclock since it is the center in the system but in the case of RME and steadyclock it might be en exception of the rule?

I have been reading the article.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

The Wordclock BNC Connection plus necessary PLL circuits inherent in a Wordlock connection increase jitter by their nature - so even though Steady Clock will lower the jitter the RME device uses internally and also re-transmits, any devices at the receiving end will have the unavoidable jitter from the required PLL's and BNC Cabling.

Using the ADC's internal crystal is about the only way to avoid the PLL's and associated BNC Cabling and connections which will always increase jitter at the receiving end.  You can have a master clock with theoretically ZERO jitter (not even possible), but send it down a BNC cable and through a PLL circuit to a converter and it will have all sorts of jitter...

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Randyman... wrote:

The Wordclock BNC Connection plus necessary PLL circuits inherent in a Wordlock connection increase jitter by their nature - so even though Steady Clock will lower the jitter the RME device uses internally and also re-transmits, any devices at the receiving end will have the unavoidable jitter from the required PLL's and BNC Cabling.

Using the ADC's internal crystal is about the only way to avoid the PLL's and associated BNC Cabling and connections which will always increase jitter at the receiving end.  You can have a master clock with theoretically ZERO jitter (not even possible), but send it down a BNC cable and through a PLL circuit to a converter and it will have all sorts of jitter...

cool

Alright, thanks for the explaination. Very instructive.
Even though all good talk about steadyclock, you suggest I will use the LYNX aurora Synchrolock as master clock rather then using the more highly regarded RME Steadyclock as master clock ? My gut feeling is telling me to use Rme steadyclock, rather then Lynx Synchrolock, but that is just a feeling and cause I have more confidence in RME steadyclock, but a gut feeling is all what it is.

Is it possible to run some sort of program to test what is the most jitterfree solution in reality with my setup ?

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

It's not about how much jitter is at the transmitting end per se - it's about how much jitter you end up with at the receiving end and at the actual ADC chip.  As mentioned, the cabling and PLL (Phase Locked Loop) required to transmit and receive the Wordclock signal degrade the signal by their nature.  So even if RME's SteadyClock gave you a perfect clock at the RME's output, by the time it reaches the other device and passes through the PLL, you will have additional jitter.

Chances are the Lynx is not using the Synchrolock when running on its internal crystal - rather it likely uses Synchrolock to clean up any incoming clock signals when run as a slave (similar to Steadyclock from what I gather).  I'm assuming the Lynx uses the internal crystal w/o any additional re-clocking when run on internal clock - this should give you the least jitter at your ADC chip from what I gather.

For "testing" - you'd need some pretty sophisticated equipment - and even then, the "sound" of measured jitter is highly subjective.  As we know - lots of people prefer external clocks, and this generally means more jitter vs internal crystal - so these people like the sound of the additional jitter when clocked externally (odd, isn't it!).  Best way is to use your ears - but you probably won't hear much difference IMNSHO (certainly not night and day as some claim - more like swapping a mic cable than like swapping a mic or moving a mic 1/2")...

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Randyman... wrote:

It's not about how much jitter is at the transmitting end per se - it's about how much jitter you end up with at the receiving end and at the actual ADC chip.  As mentioned, the cabling and PLL (Phase Locked Loop) required to transmit and receive the Wordclock signal degrade the signal by their nature.  So even if RME's SteadyClock gave you a perfect clock at the RME's output, by the time it reaches the other device and passes through the PLL, you will have additional jitter.

Chances are the Lynx is not using the Synchrolock when running on its internal crystal - rather it likely uses Synchrolock to clean up any incoming clock signals when run as a slave (similar to Steadyclock from what I gather).  I'm assuming the Lynx uses the internal crystal w/o any additional re-clocking when run on internal clock - this should give you the least jitter at your ADC chip from what I gather.

For "testing" - you'd need some pretty sophisticated equipment - and even then, the "sound" of measured jitter is highly subjective.  As we know - lots of people prefer external clocks, and this generally means more jitter vs internal crystal - so these people like the sound of the additional jitter when clocked externally (odd, isn't it!).  Best way is to use your ears - but you probably won't hear much difference IMNSHO (certainly not night and day as some claim - more like swapping a mic cable than like swapping a mic or moving a mic 1/2")...

cool

You seem have a lot of knowledge regarding clocking.

That word clock is said to be better and superior must be some old myth of some kind ?
I guess technology has gone forward.

All you are saying regarding PPLs and jitter feels very logical to me.

After all I have heard so far it seem to be the best choice to use Lynx as masterclock.  I will try to ask the guy who said he got much better sound
using steadyclock with this rme/lynx cobo.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Immersion wrote:

You seem have a lot of knowledge regarding clocking.

I'm certainly not the be-all end-all.  Just one guy with lots of digital gear to sync on a budget :-)  Dan Lavry and others have posted plenty of good info about this on the Web, and the SOS article also lends some un-biased credence to these claims (and they likely pissed off a few of their advertisers in the process).

Immersion wrote:

That word clock is said to be better and superior must be some old myth of some kind ?
I guess technology has gone forward.

All you are saying regarding PLLs and jitter feels very logical to me.

After all I have heard so far it seem to be the best choice to use Lynx as masterclock.  I will try to ask the guy who said he got much better sound
using steadyclock with this rme/lynx cobo.

Sound is always subjective, and there is plenty we don't know.  Lots of people swear by external clocking - and I'm not suggesting they are incorrect.  I tend not to blindly fall into memes and know that any humans' perception is hardly a concrete measurement tool (especially when you are trying to justify a $1200 master clock purchase wink ).  I operate from what seems logical to me - and running your primary ADC off its internal crystal (and avoiding as much cabling and associated circuitry as possible between the crystal and the ADC chip) makes 100% logical sense to me.

As long as you can get that clock reliably distributed to all of your other devices - it is very likely the best and cheapest way to sync a system IMO.

Also consider if the ADC's internal clock is not the best, then how good should we expect its BNC wordclock and PLL circuitry to be? wink  Seems like a dog chasing its tail!


cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Immersion wrote:

Even though all good talk about steadyclock, you suggest I will use the LYNX aurora Synchrolock as master clock rather then using the more highly regarded RME Steadyclock as master clock? My gut feeling is telling me to use Rme steadyclock, rather then Lynx Synchrolock, but that is just a feeling and cause I have more confidence in RME steadyclock, but a gut feeling is all what it is.

Your basic misconception here is the idea of "using Steadyclock as master clock" (as opposed to Word Clock). That's not what Steadyclock is about. Please read this:
http://www.rme-audio.de/support_techinf … teadyclock

Regards
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

RME Support wrote:
Immersion wrote:

Even though all good talk about steadyclock, you suggest I will use the LYNX aurora Synchrolock as master clock rather then using the more highly regarded RME Steadyclock as master clock? My gut feeling is telling me to use Rme steadyclock, rather then Lynx Synchrolock, but that is just a feeling and cause I have more confidence in RME steadyclock, but a gut feeling is all what it is.

Your basic misconception here is the idea of "using Steadyclock as master clock" (as opposed to Word Clock). That's not what Steadyclock is about. Please read this:
http://www.rme-audio.de/support_techinf … teadyclock

Regards
Daniel Fuchs
RME

I have been reading the manual also. But I did understand it like this, if you do the clock in cable and set RME as master clock you will use  steadyclock.
If you clock in cable and use Lynx as masterclock you will use syncroclock.  Are you telling me that Steadyclock is only used when using word clock ?

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

I have responded to this same clocking question from you over on the Gearslutz forum. Please have a look over there for further details, and also take the time to read through both your RME manual, your Lynx manual (which explains SynchroLock) and the manual that comes with your Dangerous Audio DAC/monitor controller. In brief, use your Lynx Aurora as your Master clock source (as it is doing the initial A/D conversion), then set up the RME as a slave (deriving it's clocking from the incoming AES stream), and your Dangerous Audio DAC also as a slave (once again deriving it's clocking from its AES input).

If you have a look at the technical link provided by Daniel you will see that Steadyclock is simply RME's method for extracting the clock signal from an external source, suppressing any incoming jitter it finds, and passing this cleaned up clock signal onto other devices. It uses a digital synthesis technique as distinct from other approaches which rely on multi-stage PLL's (Phase-locked loops) such as the Lynx SynchroLock approach. This is also decribed in the RME technical paper. The advantage of the RME approach is its quick locking time and it's ability to suppress jitter in varipitch situations. If you look at the Lynx manual, the analogue SynchroLock approach takes one to two minutes to achieve final lock, using a two-stage phase-lock loop system. These are simply two approaches to the same problem, and the decision as to which unit should be the master does not having anything to do with this problem, but rather should be based on which unit is doing the main analog to digital conversion, as this is where it is most important to use the internal low-jitter crystal oscillator!

So in summary, Steadyclock (RME) and SynchroLock (Lynx) are simply methods used to derive clock from an external source when either device is used as a slave. Because your initial A/D is being provided by the Aurora, it should be the clock master (in which case the Lynx SynchroLock is not required nor is it being used). Your RME AES card will be therefore be the slave (set to derive it's clock from it's AES inputs) and it will use it's Steadyclock technology to synchronize itself to the incoming stream of digital audio data. This clock will also be passed on to your Dangerous Audio DAC/Monitor controller (which is also slaved via its AES input). The Dangerous will do the final digital to analog conversion, which is then sent to your active speakers.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

Bravo! Well said! 100% correct! (just trying to ensure Immersion reads it)

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
BFpro fs, 2X HDSP9652 ADI-8AE, 2X HDSP9632

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

SuburbanStudio wrote:

I have responded to this same clocking question from you over on the Gearslutz forum. Please have a look over there for further details, and also take the time to read through both your RME manual, your Lynx manual (which explains SynchroLock) and the manual that comes with your Dangerous Audio DAC/monitor controller. In brief, use your Lynx Aurora as your Master clock source (as it is doing the initial A/D conversion), then set up the RME as a slave (deriving it's clocking from the incoming AES stream), and your Dangerous Audio DAC also as a slave (once again deriving it's clocking from its AES input).

If you have a look at the technical link provided by Daniel you will see that Steadyclock is simply RME's method for extracting the clock signal from an external source, suppressing any incoming jitter it finds, and passing this cleaned up clock signal onto other devices. It uses a digital synthesis technique as distinct from other approaches which rely on multi-stage PLL's (Phase-locked loops) such as the Lynx SynchroLock approach. This is also decribed in the RME technical paper. The advantage of the RME approach is its quick locking time and it's ability to suppress jitter in varipitch situations. If you look at the Lynx manual, the analogue SynchroLock approach takes one to two minutes to achieve final lock, using a two-stage phase-lock loop system. These are simply two approaches to the same problem, and the decision as to which unit should be the master does not having anything to do with this problem, but rather should be based on which unit is doing the main analog to digital conversion, as this is where it is most important to use the internal low-jitter crystal oscillator!

So in summary, Steadyclock (RME) and SynchroLock (Lynx) are simply methods used to derive clock from an external source when either device is used as a slave. Because your initial A/D is being provided by the Aurora, it should be the clock master (in which case the Lynx SynchroLock is not required nor is it being used). Your RME AES card will be therefore be the slave (set to derive it's clock from it's AES inputs) and it will use it's Steadyclock technology to synchronize itself to the incoming stream of digital audio data. This clock will also be passed on to your Dangerous Audio DAC/Monitor controller (which is also slaved via its AES input). The Dangerous will do the final digital to analog conversion, which is then sent to your active speakers.

THanks this post was to a big help for me. It got read a lot of the confusion I had regarding this.

Re: Can HDSPe AES mix Steadyclock and wordlock for input and outputs ?

No problem. Hope things are working out well for you with your new setup!