Digital Processor Reviews

Kal Rubinson loaned me his 8-channel Okto dac8 PRO so I could compare its measured performance with that of the dac8 Stereo. From the front, the PRO looks identical to the Stereo. On the back, there are four AES/EBU inputs and one USB Type B port. Apple's USB Prober utility identified the multichannel DAC as "dac8PRO" from "Okto Research" with the serial number string "000032." The firmware version was 1.3. Like the dac8 Stereo, the PRO's USB port operated in the optimal isochronous asynchronous mode, and Apple's AudioMIDI utility revealed that the dac8 Stereo accepted 16- and 24-bit integer data sampled at all rates from 44.1kHz to 192kHz. The AES/EBU inputs accepted data sampled at rates up to 192kHz.

The dac8 PRO's maximum level at 1kHz was 4.24V at both the balanced and headphone outputs, and the processor preserved absolute polarity (ie, was noninverting) from both outputs. The balanced output impedance was higher than the Stereo's 50 ohms, at 200 ohms from 20Hz to 20kHz. The headphone output impedance was lower, however, at 0.5 ohm across the audioband. The dac8 PRO offers the same seven reconstruction filters for PCM data. These filters all behaved identically to the two-channel processor's regarding impulse response, ultrasonic rolloff, and frequency response, so I haven't shown them. Channel separation was also the same. The PRO's rejection of jitter via its USB and AES/EBU inputs was identical to the Stereo's, with USB clean but AES/ EBU introducing some very low-level, supply-related sidebands.

I did find a small difference between the processors when I examined the multichannel DAC's resolution. An increase in bit depth from 16 to 24, with dithered data representing a 1kHz tone at –90dBFS, dropped the PRO's noise floor by 25dB compared with the Stereo's 30dB. This is shown in fig.1, which shows spectral analyses with the 24-bit tone reproduced by the dac8 Stereo (blue and red traces) and the dac8 PRO (cyan, magenta). The PRO offers just over 20 bits of effective resolution compared with the Stereo's 21 bits. The difference is very likely inconsequential.

Fig.2 shows the dac8 PRO's linearity error plotted against absolute level. Like the two-channel processor, the error is superbly low, but compared with fig.10 in the dac8 Stereo's measurements, the higher level of the multi-channel DAC's noise floor introduces slightly greater error below –130dBFS.

To examine the dac8 PRO's harmonic distortion signature, I used Audio Precision's high-performance APx555. The result with a 50Hz signal at 0dBFS is shown in fig.3. The total THD+N was 0.000189% compared with the Stereo's 0.00016%. The spectrum looks very similar to fig.12 in the dac8 Stereo's measurements, with the third harmonic highest in level at –120dB (0.0001%). The second harmonic is about 6dB higher in level than with the two-channel DAC, but as this is still at –124dB (0.00006%), the difference is inconsequential. Intermodulation distortion was also extremely low in level.

The dac8 PRO's measured performance indicates that almost no compromises were made to squeeze eight D/A channels into its slim chassis.—John Atkinson