Quote:
Originally posted by tangent
The $300 level cards really don't have much better specs than the $150-200 level ones. Instead, they tend to have more outputs, be more professionally...
|
That was my point about not wanting to spend $300 or more, and not get in return what really counts (better performance, not more gee-gaws)
Quote:
an M-Audio Audiophile 2496 is every bit as good as an M-Audio Delta 44.
|
And still not good enough! Besides, as the excellent folks at the RightMark Audio Analyzer
site point out, it seems that most sound cards trumpet the performance of their codec(s) (or separate ADC/DACs),
not the unit as a whole! Obviously, this is deceptive because the level of, for example, SNR that can be achieved is much more dependent on the board layout and support components than what the codec can
theoretically achieve. And putting the card inside a computer's case - especially if that case is a laptop?! - well, you can forget about getting the -120dB noise floor a 24bit codec is theoretically capable of (it's actually a bit better than -120dB, but physics precludes achieving better than -120dB more or less, ymmv.)
Also, the actual cost of components inside even the most exotic, premium sound card (even those ridiculous contraptions I've seen with tubes in the output stage?!) is minimal compared to the engineering and, probably more than anything, the perceived clout of the design. Well, I couldn't give two shakes about the perceived clout and I can deliver very adequate engineering services to myself for free. So, I'd much rather take a really cheap USB based external sound card and only keep the USB to isochronous data bridge + driver software part of it, as I'd rather give myself paper cuts on the lips than design a device driver.
Quote:
Or, you can help me lean on ppl to test his modded Audigy-1 (4558 -> AD8620) against his stock Audigy and publish the results. |
Given Creative's track record (and boy did they pick the right name, at least when it comes to inventing performance metrics!), I'm inclined to think that replacing its 4558s with AD8620s would amount to little more than "casting pearls before swine"