Prior to the inspiration obtained in this thread, I found USB so aversive both musically and otherwise (cf. CA) that I will not go back to it. I will wait for another paradigm shift even if Thunderbolt 3. I'll still be interested in any relevant first hand experience and look forward to your reflections astro. Meantime, I have installed an Optane Drive for both o/s and music and have a RedNet PCIeR card coming. I am re-establishing my PC with both at once and, so, won't be able to report "scientifically" about the relative contribution of each. I recommend Optane anyway - see here
https://www.computeraudiophile.com/...q-nirvana/?page=7&tab=comments#comment-789200 and here
http://www.usaudiomart.com/forum/viewtopic.php?f=35&t=1172&start=2850#p8582. The only audiophile report of the RedNet PCIe(R) card I know of is here
https://www.head-fi.org/threads/aud...-computer-audio.806827/page-219#post-13862319 and I look forward to adding to it.
Hi Iving!
I think this can come down to the exact system specs. My current DAC has an SPDIF as input, no USB. It is an old school tube DAC NOS, though excellent sound. My spare AMR DAC has both USB and SPDIF, but the built in USB card is rubbish. So all my tests were based on my Audio Note DAC 5 which is purely SPDIF. My previous USB debacle thus had to go Mac Mini > USB out > USB to SPDIF convertor > DAC input SPDIF. The Rednet goes Mac Mini > Ethernet out > Rednet 3 > SPDIF to DAC. So the Rednet is using ethernet which could be better than USB for data transfer, but also converting to SPDIF in the Rednet and feeding with it's own clock to my DAC which then uses that clock for it's timing (jitter). Remember SPDIF HAS to follow the incoming clock. In other words my DAC 5 is old school and a bit thick, it has no internal clock and has to use the CDPs clock. At Audio Note they swear by their own CDPs that are crazy money of course, but use a Philips Pro mech which has it's own board with it, and I suspect a decent on board clock on there.
Now my new DAC has a well designed galvanically isolated USB board, with good power supply to it. It has a clock on the input, and clock on the output which is i2S format to the chip array. The clocks are super close to the digital board. So in that situation we have digital stream from my Mac Mini to the USB interface then i2S and 2 x clocks. The Rednet is 1 x clock and conversion to SPDIF, then fed to my DAC 5 with no clock. I can understand how the USB board in this new DAC 'could' be better than the Rednet. It also can do 384 with no manual adjustment for clock rates.
So IMO the Rednet could beat a USB chain but may depend on the DAC input itself and how that DAC is designed. We all know how many DACs had terrible USB inputs a few years back, almost an after sight. Those DAC used a weak USB to SPDIF internal convertor board, some with USB power! I am wondering if things have moved on a bit. My DAC is the Aries Cerat Kassandra Ref II, which I think uses the Amenero 384 board like the Lampizator Golden Gate, but need to check that to be sure.
Another reason many manufactures push USB and tend to drop interest in SPDIF is the sample rate, and it's support for DSD input rates. There is a lot of users who like the sound of PCM to DSD in HQ Player and that means they have to use USB.
I think Ethernet to i2S is the way ultimately, but what do I know?
I'll post here with my findings guys.