Originally Posted by MikeJazz
Superdad, can you confirm that your solution requires that the source computer has to be on all the time?
Yes, it does. Here is how it will be used:
a) The computer (anywhere in the house, and playing any locally stored, network stored, or web streamed files) gets a small dongle (about the size of your thumb) plugged into its USB port. An Ethernet cable--yes, a second one separate from the one already connecting your computer to your LAN--runs from the dongle to your Ethernet switch/router, or if you want, could run directly to your DAC.
b) The dongle looks like an async-USB2.0 XMOS-based 32/384 (and DSD128 viaDoP) sound card--because that is what it is! Just as always for XMOS, no drivers needed under OS X, licensed Thesycon drivers provided for Windows. So just like with any USB-connected DAC, you can use ANY player s/w you like, and ALL streaming services, web browser, whatever can send audio out through it.
c) At your audio system end--anywhere else on the same LAN (can't cross Ethernet bridges without some special configuring) a DAC with another tiny module (consisting of Ethernet PHY, another XMOS chip, clock for those, ultra-low noise regulators, and special format isolators) pairs with the dongle via MAC address and not via IP address (we did not want to have to include a web page into the device for setting IP address, etc.--keeping it totally simple). This module (which can include a USB input if the DAC licensee wants that as well) accepts the master clock from the DAC and outputs to the DAC's main board a special format signal (sorry, important part of secret sauce, having broad implications for future), where it is received by isolators, run through a flip-flop and finally results in a I2S signal for the DAC.
d) Remote control of the user's chosen player s/w (since keeping the computer in a separate room is a typical application) can be via whatever of the many available means the user chooses. Anything from tablet/phone remote app (Apple's Remote app is fine on iOS if one is just using iTunes), to screen sharing/VNC, to whatever is available specifically for chosen player s/w.
We are staying out of the software business. The whole idea is to offer 99% compatibility with whatever s/w you use, and to not have any s/w support headaches.
And no, it is not multi-room, multi-channel, or anything fancier than what I have stated.
I hope you can all see how the UpTone/Swenson USB>Ethernet Audio Bridge solution will be very different from DLNA or other server/renderer models. There is no OS, no big giant processing chips, no s/w system, etc. We are just breaking apart the USB stack, putting the data into an Ethernet packet stream, receiving at the other end and finishing the USB protocol (not nearly as simple as I make it sound; lots of serious code written). And of course, this being a John Swenson affair, isolating, clocking (from DAC master clock), and PS regulation are all state-of-the-art. We, and our prospective OEM licensees, will be carefully comparing SQ of the Ethernet link to USB (both on the same DAC-side module, and versus the client's existing USB input). If it is a winner, then 2015 will be a very big and busy year for us.
To make this post the last word for a while about what is still a work in progress, let me preempt the obvious question about broader availability: After we get a few OEM licensees under our belt, we intend to offer a DIY version that capable users can install in their DACs in much the same way as they do with a typical USB>I2S board. Such a kit version will most likely include USB input in addition to the Ethernet (why not? the XMOS is there; just needs another PHY chip and a USBjack), and low jitter/phase-noise audio clocks--while still offering the line in to run from DAC master clock. And of course our required USB>Ethernet dongle will be included for the computer end (I guess if someone wanted our board set just for the USB input we would knock something off the price to omit the dongle from the kit).
A stand-alone, external box retail version does not make sense (unless we output I2S), but we have some radically different other ideas for when we get there--a long way down the road.
Thanks all for the interest. But please, no more questions on this for a while. We must first get it out there and prove the concept.
HAPPY NEW YEAR,