danlavry wrote on Mon, 25 April 2005 23:12 |
Tomas Danko wrote on Mon, 25 April 2005 19:47 | Dear Mr. Lavry,
I'm not suggesting anything, but merely discussing if by chance this method would be the 'super-secret' one found in the Apogee Big Ben. Until they make it public, we will never know, and I doubt it that they ever will publish this information.
For whatever it is worth, I am firmly seated on your side of the reasoning (both the clocking-issue as well as the overkill sample rate-ditto) and have already followed the related threads for ages by now.
Sincerely,
Tomas Danko
|
Hi Tom,
Thanks for the comments. I am curious to hear your comments about the jitter thread (jitter, internal vs external...) at:
http://www.lavryengineering.com/forum/phpBB2/viewforum.php?f =1&sid=6acfbf0f60559d11c998c2cdc5abc673
I did my best to say it all in one post, and I hope the "picture" does help...
|
Hi Mr. Lavry,
To me, the sensible way to approach this on a practical level is to first ask the question 'how bad are the internal clocks of today?'. Can they really be so bad, they need to be 'fixed' by an external clocking device? Now, that would mean they are very bad.
Or perhaps the bad clocking designs is only to be found in less expensive converters, and that leads to the second sensible question: Why spend $1,500 to 'fix' something that costed much less? Instead of spending $1,500 - $2,000 to buy something that doesn't need the 'cure'.
If these questions stated above are valid, I don't see the point to even justify an expensive external clocking device. However, let's pretend that professional converters are lacking in stable clocking and that one has to cough up more than $3,500 to solve the 'problem'. Given this scenario, I'm willing to extend the reasoning further...
I don't think the skin-effect would yield a lesser rise time severe enough to disturb the eye-pattern of the clocking circuit in the converter. I do however believe that funky cable termination can work wonders to sabotage an otherwise healthy connection. In the end, any kind of IC or OP-amp between the master- and slave-clock that has to be at both ends of the cable surely will be a bigger culprit unless the RF interference and ground loops are present (although they shouldn't be in a professional environment).
It's ridiculous to even try and claim less jitter with external clocking, all things considered.
Trying to se the forest among all the trees, or however I should put it, while reading all the correspondence from Apogee at REP it seems to me they fessed up saying that their external clock would not improve a normally-functioning converters internal clock. So that's one discussion not needed to be taken further, in my opinion.
This leaves the topic at hand down to psychoacoustics, and taste. Those are topics better discussed elsewhere, from what I understand with the purely technical focus you are trying to maintain here.
Regardless, I am starting to wonder if this is a take on noise shaping in the time-domain. As far as ordinary noise-shaping goes, even though it's supposed to shift the focus away from the areas where our ears are the most sensitive, I rarely like the result as much as plain dither only. This is my subjective opinion. Maybe one can noise-shape the jitter artifacts in a similar way, shifting the problem away from the midrange frequencies and so on, and maybe this is what's going on here. If so, it will have to be as subjective a result as different noise-shapers, and the mystery would be solved.
Best regards,
Tomas Danko - Sticking to 24/96, all converters running on internal clocking, dither applied last in the chain and no noise-shaping whatsoever. YMMV