On the Bench
For the bench tests, I used an XLR input and
XLR output. Bandwidth for THD+N tests was 80 kHz. The input impedance of our
Audio Precision was set to 100 kOhms.
Bryston states that they have lowered the
distortion in the latest version of this product, and it shows. At 1 kHz and
2 volts output, THD+N was only 0.003%.
When the 1 kHz was input as a digital signal
(16/44), THD+N went up a bit, but was still very low, at 0.005%.
At 10 kHz, still very low THD+N.
And using the DAC, although up a bit again,
still very low.
IMD was also low, at 0.002%.
Inputting the signal via the DAC, IMD went up
to 0.008%, but that is still very low. You can see the IMD peaks (red
arrows) compared to no visible peaks in the graph above. So, using the DAC
does add some distortion, but it is, in my opinion, such a low amount of
distortion as to be an inaudible change.
Output through XLR clipped (1% THD+N) at 34
volts, using one channel. It is specified to be 30 volts output maximum with
both channels operating, so this is right on the spec. Via unbalanced
output, it will be 15 volts which is still plenty.
Below is what I consider to be one of the most
important graphs for bench tests: THD+N vs. Frequency. You can see that
distortion is just about a flat line across all frequencies for both 2 volts
and 5 volts output. This means absolute neutrality in sound. If the graph
goes up in the higher audible frequencies, it will add an "edge",
and perhaps a little brightness, to the
sound. Not necessarily bad (as long as the distortion is a low figure), since
many consumers enjoy this, but it is not neutral. Kudos to Bryston here, as
getting a flat response like this is difficult.
The measured frequency response was 20 Hz to
20 kHz, - 0.1 dB.
Conclusions
The Bryston BP-26DA is a superb preamplifier.
The sound is neutral, with very, very low distortion. It is a reference
quality product, and an ideal choice for anyone who wants to have
ultra-high-end sound in their two-channel wireless media server setup (and
even if you don't have a media server . . . yet).
- John E. Johnson, Jr. -