LABORATORY TEST RESULTS
All measurements made on the DCD-2500NE SACD player were obtained using a standard Philips test CD (SBC-429) so all data was 16-bit/44.1kHz. Using this disc, Newport Test Labs measured distortion at 0dB recorded level, the results of which are shown in Graph 1. You can see that all distortion components are at or more than 100dB down (0.001% THD), so detailing the levels of the individual harmonics is essentially an exercise in technicalities, with the second harmonic coming in at –105dB (0.00056%), the third at –100dB (0.001%), the fourth at –120dB (0.0001%), the fifth at –113dB (0.00022%), the sixth at –135dB (0.00001%) and the seventh at –128dB (0.00003%).
Obviously, the results at 0dB were excellent, but musical discs are not recorded at 0dB, because there needs to be some headroom for peaks, so it’s instructive to look at the distortion results at lower recorded levels that reflect what would be on a music CD. Graph 2 shows the Denon DCD-2500NE’s performance at –10dB recorded level and you can see there are just two harmonics visible in the output—a second harmonic at –135dB (0.00001%) and a third harmonic at –111dB (0.00028%). This is an exceptionally good result.
But exceptionally good got even better at a recorded level of –20dB, where Newport Test Labs measured just one single harmonic distortion component in the DCD2500NE’s output at –121dB (0.00008%). I don’t recall any CD player—or SACD player—or DAC, delivering such good performance at this level.
The performance at –60dB was about standard for a high-quality player, with all the ‘grass’ on the noise floor being due to the fact that the –60dB signal on the test CD is not dithered, unlike all music CDs, which are dithered. Nonetheless, all the ‘grass’ was more than 120dB down (0.0001%). The remarkable effect on distortion of dithering is shown graphically in Graphs 5 and 6, which both show performance with a 1kHz signal at a recorded level of around –90dB, but with one test signal dithered and the other not. In Graph 5, when the test signal is not dithered, you can see third, fifth, seventh and additional harmonics. The second harmonic is at a real level of –110dB
(0.00031%), but since the level of the fundamental is at –90dB, the relative level of the second is at –20dB (10%). Graph 6 shows the same signal, with dither, and you can see all the harmonic distortion components have vanished below the noise floor at –120dB (0.0001%).
If you’re interested in dither, there’s a good article called ‘The Magic of Dither’ written by Stephen Dawson, at www.tinyurl.com/dither-magic
Intermodulation distortion (IMD) was also excellent, as you can see from Graphs 7 and 8, where Newport Test Labs has measured both CCIF twin-tone distortion (Graph 7) and SMPTE IMD (Graph 8). In Graph 7 you can see the two test signals towards the centre of the graph. The two closest high-frequency sidebands (at 18kHz and 20kHz) are well down, at –108dB (0.00039%), while the next are both more than 130dB down (0.00003%). As for the unwanted difference signal re-generated at 1kHz, it was 122dB (0.00007%) down, which is excellent. SMPTE results were also excellent. The thickening around the base of the 7kHz test signal is some 60Hz components, but they’re all more than 110dB down (0.00031%).
Frequency response, as with all digital products that have not had their response deliberately adjusted to deliver specific sonic attributes, was ruler flat below 3kHz, after which it ‘rolled off’ to be 0.22dB down at 20kHz.
I put ‘rolling off’ in inverted commas because the roll-off is so slight it would be completely imperceptible to the ear. Indeed it’s only visible as a roll-off on Graph 9 because of the magnified vertical scaling. If normal vertical scaling had been used, you’d have seen only a ruler-flat horizontal line along the 0.00 scale mark. Expressed the usual way, the Denon DCD-2500NE returned a response of 20Hz to 20kHz ±0.11dB. Channel separation was outstandingly good, with Newport Test Labs measuring a best result of 119dB at 1kHz, but even at the frequency extremes the results were outstanding: 117dB at 16Hz and 115dB at 20kHz. Channel balance was almost exact at 0.056dB. Channel phase was also outstandingly good, with absolutely no inter-channel phase errors at 16Hz or 1kHz, and only a 0.11 degree error at 20kHz. Group delay was typical for a TI PCM1795 8× filter, which is being used in the DCD-2500NE.
Linearity error, as tabulated in the accompanying test result sheet, was very low from –60dB right down to –90.31dB and also uniform in its errors, which is excellent. De-emphasis errors were also vanishingly low, just 0.007dB at 1kHz, 0.013dB at 4kHz and 0.2dB at 16kHz.
The digital output signal was very clean and stable, with very low jitter, with the only eye-raising result being the eye-narrowing result at 200mV of 9.9nS, which is a little higher than I’m used to seeing.
To see Denon’s AL32 processing in action you can do no better than look at the impulse and square wave responses. The impulse response is almost perfect, with almost no preor post-ringing. The same is true of the square wave response, which also shows almost no ringing at all, with the penalty appearing to be slight slopes on the leading and trailing edges of the wave. This is the best pulse and square wave performance I’ve seen from any digital player or DAC. Output voltage was around 2.3-volts for a 0dB recorded signal, which means the Denon DCD-2500NE will interface perfectly with any ancillary components you might care to connect. Power consumption on standby is less than 1-watt and increases to only 22-watts when the player is operating.
The performance of the Denon DCD2500NE as measured by Newport Test Labs was outstandingly good in every respect, and unbelievably good with respect to its time domain responses.
It receives my highest and unconditional recommendation. Steve Holding
Readers should note that the results mentioned in the report, tabulated in performance charts and/or displayed using graphs and/or photographs should be construed as applying only to the specific sample tested.