This review page is supported in part by the sponsors whose ad banners are displayed below

But first I had questions: Could you explain how your machine handles volume control? If digital—which would involve A/D conversion of the analog inputs—how do you avoid resolution decimation, bit stripping and compression of dynamic range at high attenuation rates? How much attenuation is possible before resolution suffers? If analog I assume a ladder chip from Cirrus Logic or JMC? Relative to DSD, does your DAC chip have DSD-direct mode without any further data processing?

Removing the cover is highly inconvenient. It requires a long screw driver to remove 6 screws through holes in the bottom. Getting those screws reseated afterwards is tricky. SOtM has thus incorporated a belly hatch. Here the batteries can be accessed from the bottom if they need replacing. Simply remove four easy bolts and pop the two outer push-on connectors.
I read that certain delta-sigma chips convert to 352.8kHz PCM for volume control before reverting back to delta-sigma. There's much misunderstanding about what exactly native DSD playback entails when pure 1-bit chips are no longer made except for Meitner's and Playback Design's Xilinx FPGAs coded to operate as such. The better delta-sigma chips today are 2.5-bit '5-level' designs. Could you shed some light on how this relates to pure DSD streaming? Will the sDP-1000 support DSD128 with a firmware upgrade? Relative to amp-direct drive, what is the output impedance on RCA/XLR, slew rate, current delivery and max voltage-swing potential?

SOtM replies: "The sDP-1000 uses digitally actuated analog volume to avoid A/D conversion for its analog inputs and compression of dynamic range at low volumes. Our chosen converter chip is the AK4399 from the Japanese Asagi Kasei Microdevices company. This chip accepts a direct DSD input via the DoP standard over USB. The DSD signal path bypasses the chip's interpolator so there are no modifications to the DSD stream from it. It's possible that the chip's digital volume would modify the DSD signal but we cannot find any technical information about this. Never mind, with its on-chip digital volume bypassed at 0dB which is how we've fixed the sDP-1000, both DSD and PCM data paths remain perfectly pure.


"And we're still investigating the possibility of adding DSD128. The issue could be inherent limitations of the AK4399. It's presently unclear whether it'll support 2 x DSD or not. The output impedance is 47Ω on both RCA and XLR. The latter's slew rate is set by its buffer chip to 50v/μs. For the unbalanced chip that becomes 11v/μs. Obviously those specs degrade a bit with the EMI filter of the output stage and a few other parts in the circuit. The current delivery of the balanced output is 85mA. That becomes 45mA for RCA. Max voltage swing is 6Vp-p relative to a 10KΩ input impedance."


To recap, the sDP-1000 is one of very converters in its price class to offer proper analog volume (others include the Antelope Audio Zodiac Gold and Platinum). Despite claims to the contrary—that modern 32-bit digital controls are lossless—steep -60dB to -80dB attenuation routinely sounds significantly better when accomplished in the analog domain with a quality linestage. Could SOtM's decision to use an analog control truly replace a quality preamp? With three analog inputs and remote control for input switching, volume, balance and mute, this machine on paper really did have the makings of a preamp eliminator.


When I was offered the optional sPS-1000 linear power supply as well, I agreed but immediately had more questions. Did this simply replace the small stock SMPS wall-wart charger for the internal batteries? Or did the full-size box take over certain supply voltages to limit the feeds of the built-in batteries to specific circuit junctions? Charging two small batteries with one big linear PSU did seem overkill after all.

But that's precisely what this one does to "prevent switching noise from the stock switching supply and its unavoidable impact on the AC power feeding other equipment. The umbilical cable between sPS-1000 and sDP-1000 can be silver or copper. The client decides which one they want." As a mere charging connection one wouldn't expect any difference. Unlike other schemes though, SOtM power-cycles its two cells for uninterrupted power. One battery charges whilst the other one plays, then auto-switches without interruption. The machine thus never disconnects itself from the AC power grid. And as the next photo caption explains, the production sPS-1000 will be able to power three SOtM devices simultaneously. That eliminates three small nasty charging SMPS which not only hog outlets but inject plenty of ultrasonic garbage into the power line. Now overkill becomes just right.

The hand-written 220V/60Hz reflects my loaner's prototype status. The two charger outputs with blue 4-stage trim pots are meant for the digital music server where settings 1-4 correspond to 18-21V (output 1, 80-100W); and for the sDP-1000 (output 2) where the same settings correspond to 9-12V or 20-25W. An output 3 which my loaner didn't have yet will supply SOtM's small units like the tX-USBexp, dX-USB HD to output 5, 5.5, 6 and 7V or 10W. Asking whether the variable trim-pot values compensated for various line voltages—and how would a customer know what to use—I was told that instead they'll accommodate various equipment from other manufacturers. For SOtM gear, each power output's trim pot will be factory preset for its proper value. Attention! Connecting the sDP-1000 to the wrong outlet 1 "will destroy the device". Given that risk, it might be highly advantageous to in production label each input with the device names it is meant for. This should avert both catastrophic circuit failures and any potential fire hazard from frying batteries.