I have been listening to the uRendu for about a week now and to keep it short and simple, the DS has never sounded this good in my system. Though the uRendu is supposed to have galvanic isolation on the USB end, I found that adding another level of isolation with Intona did wonders and extended almost every aspect of the SQ by a little more. I am using the cheap iFI (which I ordered at the same time) to power the uR. The setup was very simple, took couple of minutes and used in HQP/NAA mode. It runs Linux and the DS was automatically recognized with any issues (no especial configuration needed). The Sonore team has pulled an incredible product at a very decent price.
My prior setup was using well tricked out dual pc Jplay setup with WS 2012R2/AO/FidelizerPro/PL with PPA clocks/ram/USB. hdplex liner power supply and bios tweaked to undervolt/clock the CPU/RAM. I was mostly using either BHE or JStreamer/Minimserver on the control PC. With the uRendu setup, I re-purposed one of the PC to run HQP/Roon (this is the first time I am using both) and undid the cpu/ram frequency to run it as full speed. So far the setup has been working very well, stable and the SQ is excellent.
Now only if I could (and has been itching to) get my hands on the Torrey…
Thanks for sharing your thoughts on the microRendu. Once Torrey’s comes out, I’ll be running Roon from a SonicTransporter into the DS via Bridge II. I decided to go that route over microRendu/USB, though I’m still not convinced which route will yield the best sound quality and user experience.
I wonder whether adding HQPlayer to the mix would result in any better sound quality over just running Roon. I understand that HQPlayer takes the DSD conversion work out of the DAC, thus freeing up resources, and that it employs some advanced filtering that generally results in great sound, but it seems as though Ted has done an excellent job programming the DS FPGA to do basically the same thing. Anyone have any thoughts on this? How does the DS sound with just Roon vs Roon+HQPlayer?
As always you’ll just have to try things both ways and pick which you like better. But things are almost the opposite of your hypothesis:
With an FPGA like the one in the DS everything is always running, there’s no concept of saving work by skipping some steps. And further, in the case of the DS, everything is converted to DSD using the same math/same hardware so there’s no computational difference between sending the DS higher rate PCM or DSD.
The bottom line is that you won’t save any resources or time in the FPGA by doing upsampling or DSD conversions outside of the DS. What you are really doing is changing the amount of work done on the PC and that is more likely to affect the noise in your system (for good or bad) than trying to control what the FPGA is doing.
(For those who care: a simplistic but reasonably close model of FPGA programming is that every arm of a CASE or an IF is computed in parallel and then only the answer from the selected arm is used. Similarly if one argument of a general purpose multiply happens to be the equivalent of a 0 or 1 you don’t really save anything compared to multiplying with random bits: all of the multiply circuit is being used all of the time. With more modern FPGAs things are more complicated with power savings features etc., but the DS’s FPGA isn’t doing that. One way you can see this in practice in the DS is that all inputs are always locked all of the time even if they have different clocks and/or different sample rates, there’s no resyncing/relocking/clock switching at all when you change the input selection, the DS just starts paying attention to a different stream of samples - input selection is seamless.)
Ted Smith said
As always you'll just have to try things both ways and pick which you like better. But things are almost the opposite of your hypothesis:
With an FPGA like the one in the DS everything is always running, there’s no concept of saving work by skipping some steps. And further, in the case of the DS, everything is converted to DSD using the same math/same hardware so there’s no computational difference between sending the DS higher rate PCM or DSD.
The bottom line is that you won’t save any resources or time in the FPGA by doing upsampling or DSD conversions outside of the DS. What you are really doing is changing the amount of work done on the PC and that is more likely to affect the noise in your system (for good or bad) than trying to control what the FPGA is doing.
Actually, Ted, that was kind of my point although I didn't make it clearly. One of the standard reasons given for HQPlayer's sound quality is that it takes the DSD processing overhead out of the DAC and allows a higher-powered (than most DAC's) processor in the computer to handle it. But with the FPGA in the DS, I don't see that to really be a factor. So the only question is whether there is some magic in the filters HQPlayer is using vs what you employ in the DS. It just doesn't seem to me that HQPlayer should offer that great of a difference in quality over sending Roon straight to the DS. But I have been surprised before in this hobby, so I would certainly be interested in hearing the experiences of others who have been down this road.
Avadia said
How does the DS sound with just Roon vs Roon+HQPlayer?
I initially configured the uRendu in MPD/DLNA mode and used Minimserver/Kazoo to get an initial feeling on the uRendu sound and also since that's what I had in my prior setup with Jplay. Later I installed Roon and re-configured uRendu in Roon ready mode. Though the sound of Roon was much better than the MPD/DLNA mode, the HQP/NAA along with Roon is something especial I would say. The combination lifted the SQ quiet a bit in my setup. I don't know if its the HQP filters and up-sampling that is doing the trick or not.
Thanks for your response, Dev. It mirrors what I’ve read about HQP in other threads. It does seem to have some special sauce in the filters it employs. I will just have to experiment with it myself, I suppose. Do you by any chance have a Bridge II to compare SQ between that and the microRendu/USB input?
My friend just got a uRendu and a SonicTransporter so I’ll be giving both a whirl soon and comparing the Bridge II to the uRendu as renderers. I’ll probably use my own server for that comparison to keep things simple. Should be fun!
wglenn said
My friend just got a uRendu and a SonicTransporter so I'll be giving both a whirl soon and comparing the Bridge II to the uRendu as renderers. I'll probably use my own server for that comparison to keep things simple. Should be fun!
I'll be interested, I got both also but have had problems with the ST, tonight it developed electrical problem and will probably have to get an RMA, but when working the combo was very nice. Also recently installed three pair of Belkin XLR TPC's which was a major improvement, got rid of a GL hum to boot.
Avadia said
Thanks for your response, Dev. It mirrors what I've read about HQP in other threads. It does seem to have some special sauce in the filters it employs. I will just have to experiment with it myself, I suppose. Do you by any chance have a Bridge II to compare SQ between that and the microRendu/USB input?
No, I don't have BII but have B1 and may not be worth comparing.
wglenn said
My friend just got a uRendu and a SonicTransporter so I'll be giving both a whirl soon and comparing the Bridge II to the uRendu as renderers. I'll probably use my own server for that comparison to keep things simple. Should be fun!
Glenn, if you have the setup try the uRendu in the HQP/NAA mode. I find it best in that mode. Obviously, since BII doesn't support NAA, when you compare the BII with uRendu you will have to revert back to DLNA or Roon ready mode (provided BII support Roon ready) in order to keep the s/w as similar as possible.
wglenn said
My friend just got a uRendu and a SonicTransporter so I'll be giving both a whirl soon and comparing the Bridge II to the uRendu as renderers. I'll probably use my own server for that comparison to keep things simple. Should be fun!
I was taken a back when I read “Belkin XLR TPC” and thought they might have copied similar products from Belden (even the name ) as they always do with Apple
Well I now have the Sonore uRendu in my greedy little hands for a bit. My friend also had trouble with his SonicOrbiter and it had to go back for repair, leaving his uRendu with nothing to do. Quick back and forth between the Bridge II and uRendu shows these two renderers to be extremely close. I’m not going to say much else until after some extended listening but as for a USB input this thing is way way way beyond anything else that I have heard. Wowza.
) as they always do with Apple 