Viking raid? As Singxer say, "when used as an independent master clock, it can output a word clock of 44.1K-384K or a master clock of 22.5792MHz to 49.152MHz. It supports up to 768 PCM and DSD 1024. Using a large FPGA as processor core, we still decentralized some of our flagship SU-6's FPGA algorithm tech with a 150Mbps ground isolator chip for zero PC interference; source synchronization and FPGA shaping tech to reshape the post-isolation I²S signal and eliminate additive jitter; and our in-house developed clock system." It's definite geek stuff. Our ears must sort what it actually means. Just because oscilloscopes can measure it needn't mean that our system or hearing resolution can tell a difference.
To set the stage, Andrew Startsev's Slovakian cubes are single crossover-less SB Acoustics drivers in sealed Oak boxes. They externalize their XLR-only class D amp whose DSP embeds the designer's proprietary correction in the frequency/time domains. So yes, the amp only takes an XLR analog signal. Why then prattle on about USB cables, 705.6kHz upsampling, 24/192 Qobuz streams or on-the-fly PCM-to-DSD1028 resampling? Shouldn't it all get wiped out by the speaker amp's ADC? Categorically not! Things that with lesser speakers mattered not—like deciding for instance between DirectSound, Wasapi or Wasapi Exclusive output mode in Qobuz Sublime—now do. Thorsten Loesch's USB 3.0 cable made a significant difference over my prior Vermöuth Audio. Whether I upsample inside Audirvana Origin or on iFi's Cassiopeia FPGA isn't lost in translation either. Visibility now penetrates a few extra layers because the opaque top layer is gone.
The time-domain correction Andrew baked into his control monitor seems to have evaporated a thick layer of temporal blur which my prior desktop actives suffered by contrast. Whatever theoretical losses his A/D conversion incurs are obliterated by real gains in resolution and transparency. Now things matter that never did before. It's why I've been on a small tear with my nearfield remodel. Here the combatants are few: PC ⇒ USB ⇒ DAC ⇒ XLR analog ⇒ speakers.
What can you really do to/for them unless you change hardware? I did just with the USB cable. Then it was new player software of Audirvana Origin, then upsampling with it not the DAC to enroll my PC's greater computational powers for the gig. I don't use my DAC's RJ45 port. I haven't worried about an 'audiophile' network switch. All our Ethernet cabling is industrial CAT8a straight to computer network card. Would re-inserting a USB bridge to enter the DAC S/PDIF become another matter of really mattering? Or would it take the oars out of the water?
At left the guts of Singxer's €179 XU208, a basic USB⇒coax converter. Opting for the SU-2, I rather hoped for a sprinkling of fairy dust beyond just the same basic functionality expanded across parallel ports. Finally the iFi's backside for its clock i/o should that experiment net anything worthwhile. As with the Singxer's BNC output which can convert to a clock output, iFi's coaxial BNC input can convert to a clock input. Since the iFi wouldn't see an asyncronous USB signal but S/PDIF, using its clock output to sync up the source clock seemed more promising. It's how my upstairs Soundaware transport slaves to the Denafrips Terminal Plus clock in I²S mode. That causes a fine resolution uptick. Could I do the same on my desktop?
