I suspect that the differences between player apps one would notice on such a highly-resolving system (Ongaku driving Avant Garde) are too dependent on hardware, installation and operation minutae that it would be meaningless, and probably impossible, to address them at the Roon software level. The different OS and hardware resources being used by Roon vs. another app (this driver vs. that driver, that section of the drive vs. this one, etc.) will be differentiated by that amp and speaker combo. Even on the same computer/device, the two apps will be using different hardware and software resources, and that audio equipment will show those differences.
I am far from a computer audio expert (very new to it, actually), but Iāve heard that amp and those speakers (not together). Iām beginning to notice that computer audio has countless ways to affect the sound (even in my far more humble system). Just some examples: ethernet cable, USB cable, power supply, audio hardware driver, video driver, router, modem, network path, operating system, sub-processes, network traffic, solid state vs. spinning discs, RAM specs, RAM quantity, RAM architecture, and on and on and on. An Ongaku driving AG speakers will reveal differences in all those factors.
I think my conclusion is that the best way to approach it is to sneak up on the configuration that sounds best to you. Even then, there will be some motherboard made 15 years ago that will make it sound better. As Joshua learned, sometimes itās better not to play some gamesā¦
I get some of what youāre saying, but on the flipside, the differences do not fluctuate, they are consistent and persistent. Besides, my impressions have been the same as many others - and I am not one to go with the crowd generally.
You are correct that computer audio, like any other source (shall we talk cartridges and phono stages???) do affect the sound. The concept of ābits are bitsā is in theory correct, in practice thereās more to this - for example reflections introduced by not great impedance matching leading to jitter, or making the USB receiver in the DAC work harder to correct that jitter, etc etc. Iām not going to argue those details.
But your point and/or implication that this is random and uncontrollable in a highly resolving system are not correct.
I do agree with your last paragraph: Roon+HQPlayer is my playback of choice. HQP upsampling of redbook to 4x PCM is what I deem sounds best in my system. Higher res upsampling - not too much of a difference frankly.
I should add that I am now using a microRendu, which effectively isolates the computer bits from the rendering bits. It is quite possible that Roon alone playing over RoonReady sounds just as good as Roon+HQPlayer (without upsampling) playing over NAA. I have not tried.
I was previously doing most of my Roon listening with a Bryston BDP-1 and, to be frank, felt the sound was flat - not engaging - compared with the native MPD playback w/ direct attached USB thumb drives.
I purchased a Cubox-I4pro and with Reneās wonderful notes (@RBM ) was able to setup Roon Bridge on that device, with USB out to a Bel Canto REFLink into the same DAC (Bel Canto DAC 2.5).
After a lot of listening, Iāve been impressed with that combination. With an A/B test, Iām not sure I could differentiate between one or the other (i.e. the Bryston w/ MPD and the Cubox w/ Roon Bridge).
I am just throwing this out there because Iāve experienced the feeling that Roon doesnāt sound right and after exploring alternative playback chains, Iāve revised my personal opinion.
I donāt think people argue that a bit perfect player canāt sound different, it is the player after all and has the potential to add imperfections. I think when people argue about bit perfect they are referring to the digital data packets that comes from the server and arrives at the player over a network via a network protocol. I would argue that at this point bits are bits, but when it gets to the player itās up to the player to do its thing. i.e. there is no jitter or noise that would come as a result of the back end software over any other back end software. I would accept that the results could be different if the playback device was attached directly to a PC.
Do us a favour and do the test. You seem to have the ears and the system to be able to differentiate such things, shame if you donāt come full circle after all your observations above.
Assuming youāre asking meā¦ Do you mean redoing all tests as in connecting my DAC to my mini directly and comparing Roon vs Roon+HQP and the like? I am not planning to do that.
Right now my playback future includes the microRendu, and what I will assess is whether I need HQP (Roon+HQP over NAA) or whether Roon alone (Roon over RoonReady) sounds as good. I think the key question is going to be whether Roon upsampling of redbook is as good as HQPās.
Iām at substantially the same point, albeit with a different renderer (exa PlayPoint) and a more modest system (and view).
If Roon 1.3ās upsampling/filtering comes close to HQPās, thatāll really be something. It would be pretty fantastic if I could run Roon and do all my upsampling on one i7-based NUC running Roon Optimized Core Kit. Of course, thatās entirely speculative, and Iām more than a bit skeptical that even the Roon wizards can outdo Jussiās many years of work on their first upsampling outing.
Same here, though I use electrostatic headphones, because my room (family etc.) doesnāt allow my 300b SET/Horns to shine. In fact, I donāt use the speaker system at all these days. But I still prefer the sound A+ to everything else.
Apart from SQ, I donāt prefer anything else about A+. Roon+HQPlayer are very close to A+ in SQ, while the convenience of network players and the fantastic UI of Roon are an offer I canāt refuse.
Nevertheless, Iām looking into upgrading my MacMini so I can have a new main computer and dedicate the mini as an A+ endpoint.
I donāt know, but bear in mind that itās only the PCM upsampling in A+ that comes from iZotope. The DSD upsampling was done by our friend mansr (based on sox, I think), and with all due respect, I donāt think itās at the same level of wonderfulness as Jussiās DSD upsampling.
Yes, understood. However, in my particular case, I have found HQPās upsampling of redbook to 4x PCM sounds better than to DSD64. I gather this might be because my DAC upsamples internally to DSD128, but does not offer that input rate. However, when I play a DSD64 file (Iāve set HQP to not touch this stream) the sound is magical.
So I am not sure what the deal is with my redbook to DSD64 upsampling in HQP, but I can tell you in that case upsampling to 4x PCM sounds more natural to my ears, in my system. I have tried DSD5v2, DSD7, ASDM7 upsampling algos. It is a bit of a mystery to me, but whatever, I play what sounds best.
One more thingā¦ Iāve been bugging @jussi_laako for some time now to create a āplug-inā for HQP to allow it to simulate a sound output device and use it as its input, much like it take Roonās stream as input. On the mac, there is Soundflower and now Loopback that do this kind of thing.
IF we had this, we could use it in the following cases:
1- Spotify: Upsample with HQP (and route to an NAA endpoint if desired)
2- A+: Route to an NAA endpoint
3- TIDAL: Upsample MQA software decoded streams with HQP (and route to an NAA endpoint if desired)
4- ClassicsOnline: Upsample/route to NAA
List goes on and onā¦
I will gladly pay for such a plugin if it were availableā¦