I will release more firmware upgrades on Rev one that has a crazy idea on clock divider . You will increase the N (noise) to 5uV (from 3.7uV) but reduces the THD again . This time at -127and especially it consists of only H2 and H3 (rest are at -140 around).
And, does it sound better
?Which version was sent to ASR? N-3.7uV I suppose?
Cdsgames, Some new small profile clocks are starting to appear in China, Accusilicon AS318 going for as little as 28 cents each on Tabao. They are same clocks Topping used in their AK4499-based D90 (only they used the version in larger cans). Very good sounding clocks after they warm up for a couple of days. There are also some USB boards starting to appear using the same clocks. IMHO, they clearly sound better than off the shelf NDK SDA. Probably drop right into Allo dacs too. Might want to check them out. http://www.accusilicon.com/docs/AS318BM.pdf
Hmmm I find it harder and harder to distinguished improvements on my regular setup.
Even on the best setup I am wondering if some of the perceived improvements are psychological or real...however
The THD on the clock divider is very closely aligned with theory of euphonics (mostly 2 and 3 harmonic) and a lack of higher harmonics , in a sense closer in FFT (and better) to AKM DAC ICs. Yes it beats even the AKM4499 in FFT (THD but not in N)
The hardware stays same , its just a new firmware. ASR has the firmware that everyone receives .
Even on the best setup I am wondering if some of the perceived improvements are psychological or real...however
The THD on the clock divider is very closely aligned with theory of euphonics (mostly 2 and 3 harmonic) and a lack of higher harmonics , in a sense closer in FFT (and better) to AKM DAC ICs. Yes it beats even the AKM4499 in FFT (THD but not in N)
The hardware stays same , its just a new firmware. ASR has the firmware that everyone receives .
Last edited:
What's your regular setup, including cables? You ever compare to the sound of really good R-2R dacs? Yes, they have worse looking FFTs but they sound different in some good ways. That something is what we're still missing. One part of it is being able to hear long decaying reverb tails from the room in which music was recorded. Those details are still getting lost, and its not for reasons that show up on FFTs. (of course, one can still hear the strongest reverb sounds, its the subtle stuff that I am talking about...)
Two things that stand out with Accusilicon clocks are full and detailed bass, and wide stereo soundstage. They don't show up on FFTs because they are not due to stationary nonlinear distortion.
Last edited:
Know how it was done in Katana. Don't know if you have changed Rev since the pre-production one. It sounded like NDK SDA to me.
Point the test equipment at the wave and test until you find the proper termination for your particular PCB
Hmmm. Probably easier to hear differences with some other gear.
What about cables, they can have audible effects (worth some study: https://www.edn.com/loudspeakers-effects-of-amplifiers-and-cables-part-6/ ).
Also, what about equipment used for checking dacs with headphones? Headphones are what I use for checking some things, and big electrostat speakers for checking other things.
Last edited: