Ext Word Clocks - No point unless syncing other digital gear

[James Steele]

I just went to the article to check on some things. I will say that on closer read, the author limited his tests to A > D conversion. I guess that explains why he was able to quantifiably compare the results with graphs so easily. With all sorts of expensive clocks he was not able to find a single one that improved the A > D conversion. The best case was that it was about the same.

What wasn't discussed was comparing D > A converters, which actually seems like what I, and the rest of us, assumed we were talking about. I don't know how you can actually measure that. Seems like that is something you'd have to be in the room in order to hear for yourself.

I would love to see an audio publication have the courage to do a double-blind listening test with all these clocks, in a good listening environment and see if anybody could consistently pick out when a quality D>A converter is externally or internally clocked. And just to be on the safe side, I'd like them to use some really esoteric expensive clock in a test. I'm not a betting man, but if I had to put money on something like that, would I be willing to bet that the results would be better than random? It would be cool if they could get some industry "golden ear" types to participate, but I suspect some might decline... or maybe not?

Seems like a listening experiment would be the only way to get to the bottom of it. Otherwise, it's a heck of a lot tricker to make qualitative scientific measurements of the D>A stage I would think.

[tomeaton]

James... it's really easy to try this for yourself. You can trust the article, I'm sure they did the tests right, but you should also trust your ears. When I got my Isochrone (maybe 4 years ago? maybe more) I tried every which way to connect my gear in terms of clock. No matter what, the Isochrone simply sounded better than any other option. My wife, who is a musician but not an engineer, could clearly hear the difference listening to a solo acoustic guitar (which she had just played) as I changed from the internal clock in a 1296 unit to the same unit clocked to the Isochrone. It was not a small difference. There was a tremendous amount of detail being lost using the internal clock.

I preferred the Isochrone to the internal clocks of my RADAR and CraneSong HEDD, too.

I have a bunch of digital gear, and need the clock... but even if you don't, borrow or rent a master clock from someone and use your ears to tell you what you prefer. It will take less than 10 minutes.

tom

[James Steele]

Thanks Tom. I actually have a master clock. It's the Lucid GENx192. I'm not sure I hear a difference or not. It's one of those things where I'd need someone doing the switching for me quickly to see if I could tell. All this article said is that measurement-wise, for A>D, there was no improvement.

[bongo_x]

...there were some pretty strong arguments, essentially saying that it's impossible for a decently-designed modern internal clock to actually perform better when it is having to sync to an external clock source...

That's what I was referring to when I compared it to a tape machine chasing SMPTE. I can't read the SOS article, sorry.

What effect does it have on the finished product? I'm completely uninterested in things that make it sound better to me that no one else will ever hear. I'm definitely not in the audiophile club of engineers.

I think it's an interesting argument that some of these things actually make the sound technically "worse", but that people like it better for legitimate reasons, or are fooling themselves because they can hear a difference and assume that means it's better.

I've said it before, but one of the best pieces of advice I got when I was learning was
"don't A/B; if it sounds good, move on"

bb

[stephen1212b]

Dan is right but he was speaking from a theoretical design point of view. As has been pointed out most of the MOTU gear is designed to work in adaptable ways, slaving to external digital sources, at multiple sample rates, even time code. A result of all of this flexibility, certainly the Traveler and the 2408 don't have a great dedicated single reference TCXO local clock which is the design goal that is suggested can't be bettered by only made different by applying an external clock.

This flexible cost effective design of MOTU interfaces has the positive aspect that these devices can and are substantially improved by a good external clock. The Aardsync and now the Antelope are really good clocks. Look at the absolute ppm drift as well as the phase noise and jitter and also consider the power and other build quality factors. Not all external clocks are as good your mileage may vary.

That said MagicD is also correct that when choosing between similar clocks you should set the most local clock to the AtoD or DtoA if just playing back, as master. Just don't be afraid to try a better external word clock reference with the MOTU interfaces as in these real world designs it has the potential to bring a smile to your face.

[tomeaton]

What effect does it have on the finished product? I'm completely uninterested in things that make it sound better to me that no one else will ever hear.

That's an interesting working model. I guess if I can find a way to make things sound more like I think they should... I make it happen, precisely because I want the audience to hear it that way. I'm not sure how you could make something sound better to you and then not have anyone else hear it.

Clocking your system is SO basic and important. You're literally throwing away part of your gear investment if you run on an "average" clock, or with bad clock practices... whether you want to compare or not is up to you, but when people pay me money to record them I tend to think they are paying me to have thought about this stuff, considered the options, and made a choice that helps (usually by shooting for maximum transparency) get their art from mic to cd. And for me that does mean a/b comparisons (never during a session, mind you)... regularly!

tom

[stephen1212b]

Oh bye the way, since it can take from several seconds to well over a minute for something like the Traveler to slew and stabilize when switched to an external clock reference, depending on the design settings of the phase locked loop and the difference between the two clocks both absolutely and in relative phase relation at the moment of switching, fast switching back and fourth to try and hear the difference is out of the question. You can't A/B clocks that way.

If you feel that your external clock isn't improving what you are hearing it may not be. There are cabling impedance and power and grounding and interference issues that demand that the basic clock quality be quite a bit better in order to outweigh these detriments.

Having upgraded clocks in numerous devices I think it is fair to say that something like the Black Lion internal clock mod would or could be superior to slaving to their similar external clock box. Of course their external box may still bring benefit over the stock built in clock, while brand X external clock may only provide the functionality needed for various equipment to work together at the cost of transparency, actually sounding worse than the built in clock would.

Finally I want to bring in the concept that digital audio assumes that the sampling clock and the playback clock are the same. Have you ever noticed that even listening through the A/D D/A converters of cuemix during recording everything sounds great. That is because in that one instance the record and playback clock are as close to the exact same as they will ever be. They don't have to be perfect as much as perfectly the same. Playing back what you just recorded later, even with the same clock and it isn't as good. Jitter is partially random. Use a different clock as will be the case for anyone else listening to your recording and things are worse yet. The more accurate to a standard your clock is the closer a better playback clock will be to the record clock and the better the whole process will work, so it isn't just what you are hearing, there is benefit that can be passed to the final listener if you both provide a good clock close to the ideal standard.

Not to mention hearing more leads to better choices on our part at every stage of the process.

Finally look at the extreme cases of recording on location where the temperature change alone can cause a large drift in the clock frequency of a non temperature compensated oscillator, or the mechanical vibration from being too close to powerful bass at a concert, or bouncing along in a moving vehicle adding jitter. Once you get the principle it is just a matter of where to draw the line and practically that becomes a cost issue he says, longing for a $6K atomic clock.

[msmith92]

I use the Black Lion Micro Clock with my 2408mk3 and 1296. It sounds much better when tracking with it vs. the pci-424 clock.

My 2 cents!

[James Steele]

Right... well this is all very interesting. In that situation are you using your 1296 internal clock as the master? I guess I'll have to open the article and start quoting sections within reason. It made a lot of sense. I know there was the example of the sample rate being off someone mentioned with their Traveler. But the gist of what was written was that if you're interface is having to slave to external word clock it's more or less impossible for it to performer better during recording that when running on internal clock. There were plenty of charts showing increased noise floor when external clocking.

[richardein]

James,

". But the gist of what was written was that if you're interface is having to slave to external word clock it's more or less impossible for it to performer better during recording that when running on internal clock. There were plenty of charts showing increased noise floor when external clocking."

Bob Katz, in Mastering Audio, makes much the same point about, I believe, playback and external clocking.

[Proklaim Recording]

Well the bottom line to me is this. I don't get out my osilliscope when I record something into my computer, and I don't get out any other piece of electrical signal measuring equipment. What I do is listen. That is what we all do, we listen. From the first recording article I, or any of us have ever read, to the latest one we have read, they have always said the same thing. Go with your ears! If I have learned anything in recording, it is this, Go with what sounds best.

To my ears, and my wife's, who is not involved in making music or recording, the sound is much much better with my Aardvark Aardsync ll hooked up than with any other clock I have. Now what this clock does to my audio does not really matter, what matters is it sounds better, a lot better! I understand everyone may not get the same results that I have gotten from using a real good external clock, and that may be due to different equipment setups, I don't know. All I know is, as recording engineers we are supposed to go with what sounds right, or best. And to me that is definitely using a very good external clock!!

PR

[James Steele]

I was merely mentioning the article where at least in the A>D stage the factual science is what it is. That said, some people may subjectively prefer the sound when it is clocked externally. It may be similar to people preferring analog for its pleasing "imperfections."

[msmith92]

I was merely mentioning the article where at least in the A>D stage the factual science is what it is. That said, some people may subjectively prefer the sound when it is clocked externally. It may be similar to people preferring analog for its pleasing "imperfections."

Good point!

[Proklaim Recording]

I am glad you did mention the article James, and it is very interesting to know what the science does say. I have not read the article yet, but I am going to make it a point to do so. Plus it will give me a reason to have to go over to the bookstore for a coffee. If the external clock sounds better due to imperfections, as in analog, then I guess that is fine, as long as it sounds better to me, that is what I will go with. Remember though, I am clocking several pieces of digital equipment here and that could be a big difference for me. Where you or someone else may only be using one or two pieces of gear. Each person will have to use what sounds best to them, whether it be an external clock or not. Main thing is we all just want to record great sounding audio, we may all get there different ways, but in the end, it's all about the sound.

PR

[labman]

I was merely mentioning the article where at least in the A>D stage the factual science is what it is. That said, some people may subjectively prefer the sound when it is clocked externally. It may be similar to people preferring analog for its pleasing "imperfections."

Hiya James. Just to clarify a tad more about what I posted way earlier. When we are actually cutting thru A/D's here, the A/d being used always becomes the house clock - via the Antelope distribution, unless we are cutting thru 'multiple' brand interfaces (A/D's), then they all get locked to house clock. For D/A and mix etc, everything is then locked to external.