[evil-zen]

Feeling bored, I decided to make measurements of my equipment using the version 5.4 of RMAA. The stars of show are Revolution 7.1 soundcard, Denon DCD-835 and NEC MultiSpin 6Xe. It is by no means conclusive or statistically significant but I tried my best to make it so. All tests were repeated 3-4 times and I took the median values.

All cables (digital + RCA-Mini) used are the same for all 3 tests. Connection is from Transport -> digital cable -> Ack! dAck! -> RCA-Mini cable -> M-audio Revolution 7.1 soundcard line-in. The wav files are burnt on a CDr and same CDr is used for both NEC 6xe and Denon DCD-835 cd players. For the soundcard, foobar2000 is used. Settings are kernel streaming, no DSP or replaygain involved.

Here are the results
Denon
NEC
Revolution 7.1
Overall comparison

Well the results show that different transports do make a small but significant difference for dacs. Another more interesting thing though, is that testing using 16bit 96khz on the soundcard's digital out produces a much better frequency response although the dac is supposed to be decoding only 44KHz 16-bit stereo. I also tried generating test tones on foobar and felt that resampling to 96khz does indeed produce better response in the higher frequency range. 16 bits 96khz

 

[zimmer]

evil-zen --
Gee, I think the results look quite close to each other - I wonder what the deviation for the individual tests that you took median values for look like....

 

[evil-zen]

Both NEC and the soundcard gave consistent results. maybe 1% differences. Only the Denon cdp gave a wide ranging result for the IMD so I took the median value.

 

[Glassman]

rather meaningless results

what about recording at 24bit resolution instead of 16bit at least? you're not gonna see any significant differences because of just 16/44 source and very limited A/D conversion on Revo anyway.. but quite interesting frequency plot of that Ack!

 

[evil-zen]

Actually all of them were recorded in 24 bits 96khz mode i think. The source files were 16bit 44.1khz.

Glassman, do you have any idea why playing 96khz files have such improvements in the frequency response?