On the Bench
I used the Philips SACD test disc, which has
test tones and sweeps. The same test tones are recorded on a second layer
for CD players. The left channel RCA unbalanced output was utilized for
measuring the signals. I calibrated the analysis software using a 1 kHz sine
wave, recorded at 0 dB on a separate disc, such that the peak registered at
0 dB on the graph. The SMPTE/DIN test signal was at - 9 dB, while the others
were at - 3 dB. I recorded results with the 24/96 upsampling turned
off, then with it turned on, to compare the effects of the upsampling.
Using 60 Hz and 7 kHz, IMD came in at less
than 0.003%. Notice that the noise floor is lower with upsampling turned on,
although there is no effect on the IMD itself.
At 1 kHz, THD+N is less than 0.002%. It is a
bit higher with upsampling turned on.
At 10 kHz, on the other hand, upsampling not
only lowered the noise floor, it reduced the THD+N.
Click Here to Go to Part IV.