On my main rig I have a headless WS2012R2 machine running in core mode with JRiver and I’d like to incorporate Tidal, which JRiver doesn’t support, and I reading great things about Roon so I’m considering going with HQP and optimizing around it.
I don’t do DSD, but I do use Acourate to generate digital crossovers and room correction to get 4-way stereo. So 8 channels. Right now at 24/192 the CPU runs at about 15% load.
I also have a Squeezebox Touch which is fed from a std old computer from LMS. So this would be another end point for Roon.
I’m wondering what’s the configuration that would yield best sound in my main rig: use one machine for everything and connect to the DAC thru USB, use a PC as server and attach a NAA such as mR or SOSE, add a new PC as server and use my existing as NAA?
Been reading around and I’m aware Chris Connaker verified the sound out of a ethernet-connected SOSE or mR was not affected by the front end computer. But is computer + NAA achieving the best possible sound?
I’m contemplating how to use my existing PC:
Should I use this one machine for everything, connected to the DAC thru USB? Like I do now.
Use it for processing and use a NAA connected to the DAC? A microRendu, for example?
Or build a workhorse machine as front end and use my existing machine as NAA?
What is the best configuration for the best sound?
This is my current machine: Intel S1200KPR, Xeon E3 1265lv2, Crucial V4 SSD for OS, PPA USB card v2, PPA wide range picoPSU, PPA SATA cable for SSD, Streacom FC8 Evo case, lab linear PSU for picoPSU, Windows Server 2012 R2, AudioPhil’s Optimizer.
You can run everything in one PC. You do not need much horsepower when no DSD. I have HQPlayer on an old fanless homemade HTPC.
I personally think USB cable from PC to DAC is the worst way to do computer audio. Use a dedicated Roon endpoint like the SBT or MicroRendu.
Later you can use a dedicated Roon server. QNAP or the Sonic Transporter i5, or almost “anything”.
I suggest you purchase the UltraCap LPS-1 to power your SBT and maybe even consider FMC’s, where one of them could be powered with a Y-split from the LPA-1.
Later you can buy the MicroRendu (If your DAC supports USB), and replace the SBT, but I think The LPS-1 will make the biggest change in SQ.
If your DAC support 24/192, do EDO on your SBT.
There are so many possibilities. I think you just have to try, according to what your wallet allows
The general advice is to split Roon and HQPlayer, but you don’t do anything wrong trying with one PC, and then evolve.
As you see, I did not think any of your 3 suggested options was the way to go, but others may think different.
(About 8 channel, I don’t have sufficient knowledge). Does you DAC support 8 channel FLAC?
That will get you as many answers as their are people to answer I think. Any of the ways you listed to use your pc will work well try them and see which sounds best for you. I personally like having an ethernet feed directly into the endpoint. But, those DACs are rare and on the too much for me to spend on a DAC end of the spectrum.
Be aware, that Roon is only stereo and does not do multichannel at the moment.
Yeah. I’m aware I won’t get one single and clear answer. That is the nature of fora… But over the years I’ve learnt to segregate.
The software architecture would be Tidal stereo into Roon stereo into HQP that will convolve into 4-way stereo, then send to NAA (should there be one) and to the DAC thru USB. I guess a microRendu doesn’t care if it’s sending 2 or 8 channel through the USB as long as the DAC can take it.
Sorry, but this sounds like involving rather than evolving. I had my SBT as front end for a couple years. Eventually it was replaced in the main rig and my current computer with the linear power supply is A LOT better sounding (in core mode, with AO on, etc., etc.) than the SBT with the same linear power supply. As much as I love the SBT, it’s not the best sound anymore in my case. Let alone it not being able to do multichannel, which opened the door to room correction and active digital crossovers and active system, which is yet another two steps further.
I’m open to options I have not considered, but this one I don’t see reason to go after.
Not many people have tried all the options you’ve described and compared them. I’ve moved from a direct USB connection to an NAA by several steps.
I use an i7 BRIX to run Roon and HQP, upsampling to DSD128, convolving Room EQ for two channels and sending Output to an NAA on a mR with the UltraCaps PS.
I’m very happy and it sounds better than when I used a direct connection USB. I was surprised at the improvement the UltraCaps PS made to the mR. Dynamics increased to the point where it was blowing a B+ fuse on an overly sensitive power amp.
