You've got to ask yourself whether a better clock would bring an improvement in sound in the first place. You seem to all be assuming that only by having a more stable clock would the sound be perceivably better...and yet is this not the world where everyone still to a large extent prefers the 'sound' of analogue?

Maybe the jitter is not reduced, however the effect of the external clock does bring around a change which happens to be pleasing to *your* ear

rob

Regardless of who is right or wrong, giving someone who is trying to constructively add to the discussion the fuuck is disrespectful and uncalled for.

He said, "You either try it yourself and see what im talkin about
or dont but if you have no exp on the unit then you have
nuthing to really contribute to this Thread because you dont
know what your talking about Nuff said."

You don't consider that rude? If so, then your manners are lacking.

People always seem to get so upset about wordclock and whether it improves things. I also have Big Ben and see bigger improvements on an old 882 than on my Rosetta 800. However, none of it matters really. If g-netic hears a big emprovement it can only improve his situation and inspire him to better work.

G-Netic doesn't owe me an apology. I only would like to know exactly how his listening tests were performed. I suggested that since the RME is an extraordinarily immune piece, the likelihood that there would be a "huge improvement" is so tiny that any sane person would have to question G-Netic's results. Unfortunately, it's not enough to say, "we heard a huge difference". We have to know the intimate details of his testing procedur. Please, G-Netic, go into detail about how you performed the tests. Were they performed consecutively (one listen after another?) or synchronously (switching back and forth between two different, synchronous files?

If he performed the tests consecutively, then I'm very very very sorry to say that they are likely invalid! Anyone can convince themselves that one listen is better than another. And there are issues with phase locked loop locking. As the converter goes into lock, its sound changes! It's like trying to tarck a moving target. It's easy to convince yourself you hear a huge differnce when things are changing right before your ears and have not settled down. It probably takes a good 30 seconds to a minute for the secondary phase locked loop in the converter to settle down.

Which is why a consecutive test, especially of converters and clocking, is fraught with problems. The only way G-Netic could prove to the world that his consecutive listening test is valid would be to do it with many trials, the whole double-blind ABX thing. That's such a pain in the ass that most people (including me) try to avoid it.

But synchronous switching has a much greater chance of increased reliability, even when done less formally. Though it is still fraught with problems because music is a constantly changing thing. But it is a lot more reliable test to perform, the human ear is making instant comparisons. You do have to make several trials, and you do have to do it blind.

So if G-Netic performs a blind synchronous switching test and hears a meaningful (big???) difference, say, 8 times out of 10 in blind trials, then I think it's time to do some further investigation! Because his results go against theory, AND PRACTICE. Plenty of reliable listeners have reported that great converters like the Prism, the Lavry, and several others, simply do not benefit from external clocks. So G-Netic is the one who's under fire, and he has to report his testing methods, that would be the first step.

My listening tests with the TC Electronic System 6000, which has EXTRAORDINARY jitter rejection, on internal versus external sync, demonstrate that the difference is so small as to be "inaudible", and most people would fail a blind test on it. In other words, a converter of that quality should not be affected at all. So, either, the RME is not as good as they say, OR, and not to offend G-netic, he performed his tests improperly. To be totally honest, most people perform their listening tests informally, and rely hugely on these invalid tests. A valid listening test that can stand scrutiny is EXTREMELY difficult to perform. In fact, it is IMPOSSIBLE to peform a valid consecutive test on a converter while changing clock! The time it takes for the phase locked loop to settle is enough to totally invalidate the test! And if you wait the 30 seconds to a minute for the secondary PLL of the RME to settle, your listening memory of what it sounded like completely goes to pot.

To be totally honest, and not meaning to offend G-Netic, in my opinion, with converters of TC, Prism, or RME quality, you would need an extraordinary reproduction system and room acoustics to hear any kinds of differences. And your listening tests have to be performed impeccably, not at all informally. Been there, done that, got the T-Shirt, as they say. I've been listening to and studying jitter for years, and I think I have a very strong handle on what jitter sounds like, the various "flavors" of jitter, and their influence on the stereo image, the tonality, the depth. And the kinds of listening tests you have to go through to prove there is a difference.

To summarize: The better the converter, the less likely it will EVER need an external clock. Show us how you performed your listening tests, please.

Wait a minute Bob, how dare you interject science into this discussion?

You must know then when you and Dan post these kind of things all you are going to get back is something like, "I don't know how i is doing it and I don't care what the science is. I hear the improvement and I trust my ears."

We live in an age where everyone's opinion is supposed to be given equal weight regardless of the experience or education level. It is egalitarianism run amuck with no intellectual discipline.

But thank you and Dan for educating me and spurring me to conduct my own tests a few months ago that led me to get a LavryBlack DA-10 instead of a Big Ben.

A few milliseconds! *sigh* A millisecond spans more than 20 sampling periods at 48KHz. Your problem has nothing to do with jitter, which is a phenomenon on the scale of a single clock period. Your problem has to do with latency and the synchronizing of digital interfaces etc as far as I can tell from your posts. Is this what you meant when you said it improved the sound of your converters?

-synthoid

Bob,

I'm in total agreement with your last post - double blind tests properly documented are the only way to find the truth in these circumstances.

I glad you have the patience, but I just wonder what the point is of explaining the science again and again and again. What's the point of trying to educate the uneducatable ?

This subject has been hammered to death over the last year or so. Both here and on PSW. It doesn't matter how much you explain the scientific standpoint. how much you point out that 'objectivity' can be very 'subjective' in many situations. There are always people who will believe what they want to believe despite the facts and I believe that Apogee are more than happy for the hype to continue (which is why they didn't send you an eval unit) as hype sells.

As evidence I give you:
Uri Geller
The DaVinci Code
UFOs
The Bermuda Triangle


All of these are REAL phenomenon - I've seen the pictures with my own eyes.

FWIW I've got a big ben for a master clock as I have a load of digital devices to synchronise - and don't hear any difference when clocking internally to externally (I've an RME card too !). Doesn't prove anything either way of course.

si

Good comment, Rob. I've heard that argument. But if you are making listening judgments based on a DAC which is susceptible to jitter, then you're hearing it WRONG, and the world might be hearing it right! You wouldn't want that to happen, you want your reproduction system to be accurate.

Here are some of the effects of jitter. I'll give them in pairs. Please pick the choice which you think would sound better to you:

a) Narrower b) Wider

a) Colder and harsher b) Warmer

a) Resolved and detailed b) vague and undefined

a is the less-jittery presentation. Of the three pairs of symptoms above, I'd almost always pick a, except perhaps for the third pair, where I could see a bit of vagueness might make some masters sound better. But if that's in my DAC, then I'm fooling myself!

If it's in my A/D, I could see some guy wanting to "tune" his A/D by the jitter. It would be mighty hard not to fool yourself, though, since the same clock might be driving both A/D and D/A. How do you know whether the effect you are achieving is in the recording or the reproduction? The only solution is accuracy.

Then you can achieve the "vagueness" you want through mike placement, eq, tubes, analog warmth, all the places where you have a bit more control. An A/D with color is fine, I suppose, but if it's achieved through jitter, it sounds like an uncontrolled situation.

If you're getting offsets of a few milliseconds under one system and not in the other then you have some problem other than clocking issues. Something is broken or patched incorrectly.

-R

I'm not sure if this has any point, but...

In the RME control settings there is a checkbox for the DDS. The DDS is the whole internal clocking system from what I understand and the Fireface will adjust incoming clock signals and clean possible jitter when this is active, just as it cleans up the internal clock signals.

As long as this checkbox is ticked, I do not think it makes any difference if the unit is acting as a slave to an external WC, or pacing with the internal clock, since the RME steady clock system will be involved before digital audio is hitting the DA converter on the Fireface.

So, in order to hear any effects of a Big Ben, shouldn't you tick out the DDS to start with and bypass any jittercleaning within the Fireface? And in that case - won't there be considerable amounts of jitter present at the DA conversion since you say "no" to cleaning it up?