External Clock or No External Clock, that is the question

Mike

Mike

Audioshark
Staff member
Joined
Apr 2, 2013
Messages
30,489
Location
Sarasota, FL
I have read so many conflicting reports, theories, etc. on going with an external clock vs internal clock. I had a chance to sit down with Ted Smith at RMAF for about an hour to discuss the topic. Of course, he believes an external clock is detrimental to the sound because of so many things that can be introduced by having another box, PSU, cable, etc. whatever.

But I read review after review of people raving about the performance improvement they are receiving from an external clock.

What's your view? What has been your experiences. I for one have never had a digital component with an external clock. I'm curious.
 
The short answer is: try it and see.

The long answer, well here goes.

In general, clocks should be physically located as close to the DAC as possible. If you look at the PCB's of most DAC boards, you will see that the designers usually attempt to put the clock as close to the DAC as they can, with nice wide PCB traces so that as little of the clock signal is corrupted as possible. The further away the clock is, the greater the potential for timing errors.

If you locate the clock externally and rely on a cable to transmit the clock signal, this is very bad. The cable needs to be high quality and exactly 75 Ohm, otherwise the signal will reflect when it encounters a difference in resistance and corrupt or degrade the clock signal. The usual way to terminate such a cable is with a T-piece BNC connector, with the cable on one end, the BNC connector on the other, and a 75 Ohm resistor on the other end. Even if the clock signal survives to the BNC connector, it has all the internal cabling to worry about.

Furthermore, how well your DAC responds to the clock signal depends on the design of the Phase-Lock Loop (PLL). The PLL is a comparator that compares the timing of its internal clock with that of the incoming audio signal, it then goes through a feedback loop so that it can adjust itself to provide proper timing. A good PLL will obtain lock of the audio signal whilst ignoring as much jitter as possible.

A "fast" PLL reacts too quickly to incoming jitter - it effectively re-creates the jitter it is fed. On the other hand a "slow" PLL is much more effective at jitter rejection. However, the type of PLL a designer chooses depends on whether he expects his product to be clocked internally (i.e. most DAC's) or externally (pro audio equipment).

Now, I am sure you will realize that if you have the wrong design of the PLL and tell your PLL to use a severely degraded external clock signal to reclock your audio stream, you might make things worse. On the other hand, if you have the correct type of PLL and the clock signal has survived, you will have made an improvement.

In short - whether you experience a benefit from an external clock depends on a whole bunch of variables that have been baked into your hardware by the designer, of which you have no idea (when was the last time you saw a DAC manufacturer publish details of their PLL?) and no control over. The quality and jitter performance of the external clock is not the only variable!

THIS is why people have so many different opinions about clocks.

So after saying all that - it all boils down to this: try it and see.
 
Mike.......I also had some interest in an external clock generator, specifically the Esoteric G-02 for one of my Esoteric K-03 CD/SACD/DAC players. Once I dug into it I discovered that with the K-03 the external clock only reclocked the transport, not the external digital inputs (coaxial, optical, USB). On the new Esoteric K-03x ,with the DAC that decodes DSD, the external clock does work on external digital inputs. At any rate, I decided it was not worth the price of admission in my case to add the Esoteric G-02 to my K-03.

I have read many testimonies about how a more accurate external clock connected to a DAC helped recreate a more musical presentation. I have not had the opportunity to audition an external clock so I don't have a clue if the difference is worthwhile.

I have discovered the Cybershaft external clocks, ordered direct from the Japanese manufacturer through their Amazon store. These clocks are more reasonably priced and may be worthy of experimenting with. http://www.cybershaft.jp/
 
My understanding is an external clock is indeed detrimental to sound if the DAC's internal clock is located close and have ultra low jitter. The external clock is really meant for the pro audio world where multiple devices needs clock synchronization. For home audio if an external clock is helping with the sound, then the DAC's internal clocks are riddled with jitters and most of the time these jitters are driven from PLL circuits to support different sampling rates. If I were to pick from a ultra low jitter clock with no PLL circuits based DAC than a very high quality external clock, I would pick the former.
 
Keith_W said:
The short answer is: try it and see.

The long answer, well here goes.

In general, clocks should be physically located as close to the DAC as possible. If you look at the PCB's of most DAC boards, you will see that the designers usually attempt to put the clock as close to the DAC as they can, with nice wide PCB traces so that as little of the clock signal is corrupted as possible. The further away the clock is, the greater the potential for timing errors.

If you locate the clock externally and rely on a cable to transmit the clock signal, this is very bad. The cable needs to be high quality and exactly 75 Ohm, otherwise the signal will reflect when it encounters a difference in resistance and corrupt or degrade the clock signal. The usual way to terminate such a cable is with a T-piece BNC connector, with the cable on one end, the BNC connector on the other, and a 75 Ohm resistor on the other end. Even if the clock signal survives to the BNC connector, it has all the internal cabling to worry about.

Furthermore, how well your DAC responds to the clock signal depends on the design of the Phase-Lock Loop (PLL). The PLL is a comparator that compares the timing of its internal clock with that of the incoming audio signal, it then goes through a feedback loop so that it can adjust itself to provide proper timing. A good PLL will obtain lock of the audio signal whilst ignoring as much jitter as possible.

A "fast" PLL reacts too quickly to incoming jitter - it effectively re-creates the jitter it is fed. On the other hand a "slow" PLL is much more effective at jitter rejection. However, the type of PLL a designer chooses depends on whether he expects his product to be clocked internally (i.e. most DAC's) or externally (pro audio equipment).

Now, I am sure you will realize that if you have the wrong design of the PLL and tell your PLL to use a severely degraded external clock signal to reclock your audio stream, you might make things worse. On the other hand, if you have the correct type of PLL and the clock signal has survived, you will have made an improvement.

In short - whether you experience a benefit from an external clock depends on a whole bunch of variables that have been baked into your hardware by the designer, of which you have no idea (when was the last time you saw a DAC manufacturer publish details of their PLL?) and no control over. The quality and jitter performance of the external clock is not the only variable!

THIS is why people have so many different opinions about clocks.

So after saying all that - it all boils down to this: try it and see.
Click to expand...

Great post. Spot on.
 
Keith_W said:
In short - whether you experience a benefit from an external clock depends on a whole bunch of variables that have been baked into your hardware by the designer, of which you have no idea (when was the last time you saw a DAC manufacturer publish details of their PLL?) and no control over. The quality and jitter performance of the external clock is not the only variable!
Click to expand...
Well, maybe an indirect indication would be whether or not there's an input for an external clock and/or the manufacturer's recommendations regarding external clocks.
 
Certainly clocks that feed DACS need some consideration, thanks Keith_W for an excellent post, that's the crux of the matter.

For AES3 and S/PDIF clocking, the field is still wide open, if the data from one device to the other is re-clocked properly, the level of jitter tanks, like for this older Panasonic SV-3700 DAT recorder's output.

attachment.php
and then cleaned up by reclocking & isolation
attachment.php


This is utilising pro audio technology for us audiophiles. Re-clocking benefits outputs from 44.1kHz s/PDIF signals from CD players, 48kHz tuner outputs, and of course 192/176.4/96 AES3, AES3id and RCA signals. Given USB requires a few more steps to achieve a similar result, why bother with USB, but USB can carry DSD, I really like DSD. I'm very happy with conversion of <DSD256 to PCM to AES3 and running this through a re-clocker. You get to keep your favourite DAC since most of them have an AES3 or RCA S/PDIF input, although RCA coax is a shade above TOSLINK for resolution.
 

Attachments

  • j-test-AMI-DDH-1-2448-coax.jpg
    j-test-AMI-DDH-1-2448-coax.jpg
    122.3 KB · Views: 121
  • LowBeats-J-Test-2448-Linnenberg-pre-2k-coax.jpg
    LowBeats-J-Test-2448-Linnenberg-pre-2k-coax.jpg
    110.7 KB · Views: 118
Interesting link. I have not thought about external clocks in my system. Appreciate the info above. CH precision have a new T1 clock out for the C1 DAC. Would love to be able to compare the impact of adding the T1 to my C1 and see if there is a noticeable improvement.



Sent from my iPhone using Tapatalk
 
tunes said:
Interesting link. I have not thought about external clocks in my system. Appreciate the info above. CH precision have a new T1 clock out for the C1 DAC. Would love to be able to compare the impact of adding the T1 to my C1 and see if there is a noticeable improvement.
Click to expand...

External clocking is very useful if the source can be synchronised to the same clock as the DAC. Typically this would be a DAW sound card something like this RME. The situation is the clock would have to be changed every time the sample rate of the source material is changed. This can be a challenge for a playlist containing say 96, 44.1 and 48 sample rate tracks. The big plus is, since the source and DAC have the same clock, the transmission of data has very low jitter and it also depends on the quality of the clock that's in use.
 
Mike said:
I have read so many conflicting reports, theories, etc. on going with an external clock vs internal clock. I had a chance to sit down with Ted Smith at RMAF for about an hour to discuss the topic. Of course, he believes an external clock is detrimental to the sound because of so many things that can be introduced by having another box, PSU, cable, etc. whatever.

But I read review after review of people raving about the performance improvement they are receiving from an external clock.

What's your view? What has been your experiences. I for one have never had a digital component with an external clock. I'm curious.
Click to expand...

"External vs. internal" is implementation matter only. It is impossibly to give univocal recommendation.

In pro audio may be used dedicated clock interface and separate clock generator box "Word Clock".

If hardware spectrum analyzer available, need check middle level noise at analog output of a DAC, to choose the best decision.

In general, for home using recommended internal clock for avoiding jitter via cables.

For multichannel-multi-dac systems may be considered external clock. Recommended pay attention to digital cables and grounding for this case.
 
I have reached that same decision buying a good 1 box player to eliminate this issue.

Is it better than the manufactures higher multibix solutions, no, i now do not think so.

I have now an esoyeric clock, quartzlock rubidium clock, cybershaft ocxo selected version 16 and now also a Mutech reclocker.

In all instances, given the use of good cabling the improvements are obvious. All different clock seem to be doing things differently and always an improvement.

Ask the dealers of clocks, almost all customers have liked the result of adding a clock.

The clocks even in an Aurender W20 showed nice improvements using an external clock. Most external clocka are still magnitudes better than even typical ocxo used in a player or dac, which are very rare animals to start with. So, even much expensive gear, soumding very very good to start with, sound even better clocked.

The improvements of resolutiin, naturalness, space and realism seems universal.

No matter how good a system is, the clock is the forefront, heart and basis of a good digital system.
 
justubes said:
I have reached that same decision buying a good 1 box player to eliminate this issue.

Is it better than the manufactures higher multibix solutions, no, i now do not think so.

I have now an esoyeric clock, quartzlock rubidium clock, cybershaft ocxo selected version 16 and now also a Mutech reclocker.

In all instances, given the use of good cabling the improvements are obvious. All different clock seem to be doing things differently and always an improvement.

Ask the dealers of clocks, almost all customers have liked the result of adding a clock.

The clocks even in an Aurender W20 showed nice improvements using an external clock. Most external clocka are still magnitudes better than even typical ocxo used in a player or dac, which are very rare animals to start with. So, even much expensive gear, soumding very very good to start with, sound even better clocked.

The improvements of resolutiin, naturalness, space and realism seems universal.

No matter how good a system is, the clock is the forefront, heart and basis of a good digital system.
Click to expand...
Hi and thanks for your post. I was going to try a Mutech reclocker. How does this compare with the others?
 
alto said:
Hi and thanks for your post. I was going to try a Mutech reclocker. How does this compare with the others?
Click to expand...

It does have an improvement for bluesound node2 reclock. I then have my cybershaft ocxo fed esoteric feeding the mutec.

Another leap in naturalness.

As i already have the ocxo fed esoteric feeding my player. It was a no brainer for the mutec reclocking the bluesound.

So i am getting X2 improvements.
 
alto said:
Great thread. Can we have more user reports please? Anybody tried a Mutec?
Click to expand...

We have not tried a Mutec, but the difference in our systemwith an internal clock vs. running with an external one was startling.
Keith_W and others has already eloquently made the argumentas to why internal clocks potentially have the edge, but I will give you an experiencein our system that really hit home.
Our digital front end has an ultra-accurate clock located inthe DAC itself and allows one to use this clock within the DAC to completelyreplace the incoming clock. They callthis option relocking and it is an attempt to remove all jitter from the inputsource by using the onboard DAC clock and internal buffers to decouple theinternal and outgoing streams and clock using a 33fs clock internal to the DAC.
One day I fired up the system, put on some music, and leftthe room to allow for everything to thermally equalize in the system. When Ireturned and sat down to listen, I felt there was something wrong with what Iwas hearing. I asked my wife to come in and she concurredthat something was not right. After somedebugging, we found out that the reclocking feature was off (due to a handshakebug subsequently fixed by the vendor). For whatever reason, when we turned on our system this time it hit thatbug such that it was clocking off the external signal instead of using theinternal DAC clock.
We power/cycled the digital front end and this time we didn’thit the bug so reclocking was on. Weboth immediately felt the “magic” returned and we were hearing what we wereused to hearing.
Given we can at will turn this feature on and off, we ran someblind and sighted tests with the internal and external clocking. When running blind both my wife and I canjudge which mode it is in 100% of the time. When running sighted, we hear the magic return when it is using theinternal clock.
I can’t emphasize enough in our system to our ears how muchsuperior it is running with an internal clock.
 
You must log in or register to reply here.
Back
Top