Clock source doesn't matter?

[Rob0214]

So, just for the hell of it, I took some stuff I was working on, recorded it a stereo track with the 003 using internal clock. Then, I did the same thing using my TC Finalizer as my external clock. Inverted one track, and the result was a perfect null.

So, does this mean that clock source doesn't matter, and those who claim anything different are prejudiced by the experiment? Empirical results say yes. Haven't tried it with my Rosetta as the clock source, but somehow I feel it will yield the same result.

[daeron80]

If both transfers are occurring in the digital domain, then the results will always null regardless of clock source. It makes no difference to the integrity of a digital stream how it's clocked, as long as it remains digital. You could even make a digital transfer at a different sample rate with no loss.

But if one or both of the transfers were done in the analog domain, it would be surprising if they yielded a complete null. That's because clocking affects the accuracy of encoding and decoding (A/D and D/A). Tests have shown that the use of external clocks in DAW environments results in greater jitter. But to the ears of many, if the external clock is more consistent over the long term, the low end sounds tighter. Whatever sounds good to you.

[CME]

Yeah your test was a moot point. Since no audio was actually being converted for A/D or D/A the clock would not make a difference. The idea of a better clock counts when there is conversion going on. So, in theory, a better clock would produce a better sound into the digital domain when tracking, and then allow you to hear it better as it transfers back to analog to go to the monitors. Once it's "inside the computer" the only difference a clock would make is in you hearing what is coming out. You might hear things differently and cause you to mix differently but wouldn't actually affect the audio that's already been recorded. Hope that makes sense.

[Rob0214]

Ok, I sort of get it.

I did the test because I wasn't really hearing any difference in the monitors. Given what you guys are saying, would a better test be to put a loopback conversion (ie put a hardware insert on the master fader to outputs 3-4, and wire output 3 to input 3, output 4 to input 4)?

I use a lot of hardware inserts for mixing (lots of outboard stuff), so it would effect that too, right?

And given my choices of clocks (003 internal, finalizer, Rosetta 800) my guess that the order of preference should be Rosetta, Finalizer, internal. Unfortunately, I don't know of a way to use just the Rosetta's word clock at 88.2.

[CME]

Originally Posted by Rob0214

Ok, I sort of get it.

I did the test because I wasn't really hearing any difference in the monitors. Given what you guys are saying, would a better test be to put a loopback conversion (ie put a hardware insert on the master fader to outputs 3-4, and wire output 3 to input 3, output 4 to input 4)?

I use a lot of hardware inserts for mixing (lots of outboard stuff), so it would effect that too, right?

And given my choices of clocks (003 internal, finalizer, Rosetta 800) my guess that the order of preference should be Rosetta, Finalizer, internal. Unfortunately, I don't know of a way to use just the Rosetta's word clock at 88.2.

To start of with the test. Yes looping through the 003 would put it through a D/A and then a A/D conversion. I also think you could use the Rosetta's clock at 88.2 with a BNC cable but you couldn't actually use the Rosetta as an ADAT interface at 88.2 as the ADAT input is limited to 48khz. And honestly I wouldn't expect much difference in between the Finalizer and internal. And honestly the Rosetta may not be a big jump either. But def try it out and see what you think. IMO a more interesting test would be looping a session through the 003 and through the Rosetta. You could even do it with both clocks. So in other words you'd could try

A. 003 as master clock, playback looped through 003
B. Rosetta as master clock, playback looped through Rosetta
C. 003 as master clock, playback looped through Rosetta
D. Rosetta as master clock, playback looped through 003

However IMO "improvements" in clocking and converters is additive. The more tracks you have recorded through a setup the more the difference become apparent. So just routing a stereo (2-channel mix) would be much more subtle than a full session recorded through different clocks/converters. Just IMO though.

[daeron80]

Haven't tried that kind of test, but my instincts tell me that it would be very difficult to hear any differences. A better test might be to split the output of a mic pre 3 ways and record it on three identical systems being clocked differently. Record a wide variety of instruments and voices through it - preferably a moderately large arrangement of the sort that could be built one mono track at a time. Mix all three identically with the same clock, and see if they sound different. A lot of trouble. And would still only be anecdotal evidence, not a rigorous scientific test. I think probably several months of regular work experience with each clock source would be more telling - a cumulative case kind of "test."

[CME]

Originally Posted by daeron80

Haven't tried that kind of test, but my instincts tell me that it would be very difficult to hear any differences. A better test might be to split the output of a mic pre 3 ways and record it on three identical systems being clocked differently. Record a wide variety of instruments and voices through it - preferably a moderately large arrangement of the sort that could be built one mono track at a time. Mix all three identically with the same clock, and see if they sound different. A lot of trouble. And would still only be anecdotal evidence, not a rigorous scientific test. I think probably several months of regular work experience with each clock source would be more telling - a cumulative case kind of "test."

Yeah. I wouldn't expect much "audible" difference per se. But it would be interesting to see what his opinions were. I was part of the BLA Mod test thread that did about as good a job as anyone can do of comparing different rigs, and it was rather interesting. To me the only true test would be like you said. Splitting the source to different systems simultaneously to record to. However mixing it becomes the hard part. How do you mix something on each system like you've never heard it before? Seems to me you'd take something you learned in the first mix and pass it along to the second. And then pass from the second to the 3rd. Ultimately giving the third mix an advantage over the second and the second mix over the first. Unless you had separate people mix it, but then everyone mixes differently. Guess maybe a blind listen test to a mix that doesn't include any plugs. Just level adjustment. Still no way to be truly unbiased.

[daeron80]

I would mix it on one system, then make copies of the session and replace the audio with the versions recorded via the other two clocks. That would be the only way to get a head-to-head comparison.

[smpkeys]

Sound on Sound had an interesting article on clocking. It seems that the writer determined that they "expensive Master clocks" don't make the difference that they are hyped to suggest. Here's a quote, and a link below that for the whole article.

"But in any of these cases, the use of a master clock will not improve the audio quality achieved by the converters in any technical sense — and the most expensive clocks fare no better in this regard than the least expensive. "

http://www.soundonsound.com/sos/jun1...sterclocks.htm

[Rob0214]

I figured out how to get the Rosetta to be the master clock about 5 seconds after I posted :)

I have a session I mixed with 5 channels of HW outboard inserts. Everything's still set up. I can bounce 3 different tracks with the different clock sources and post them for a blind test if anyone's interested.