I have a Denafrips Ares II DAC that can be on NonOverSampling or in OverSampling mode.
I can also use the DSP in Roon to oversample.
What is the best way to configure this.
A: Max PCM Rate in Roon and NonOverSampling Mode in DAC
B: Sample Rate Conversion Disabled in Roon and OverSampling enabled in DAC
C: Oversampling enabled on Both.
I am also using Tidal, so there is a lot of MQA music. I have also a Node 2i that supports MQA and currently my Ares II in connected to the coax on the Node 2i that I use as my Roon Bridge.
Try it every way and trust what you hear and like best…no point to ask us when you can try yourself the options at any time you like. Personally I have resorted to Native rates and keep things simple in my setup (not denafrips).
What do you mean with native rates? (English is not my native language)
Yes, I have tried all the options. Now it is OS in the DAC and No Oversampling in the Roon. Sounds nice, but I cannot stop wondering if there a benefit to enable the oversampling in the DAC also. or vise versa…
My engineer mind does not let me trust my ears only. I would like to know if there is a theoretical benefit to do it some other way.
Like, is there a difference in the oversampling algorithms between the DAC and Roon. Roon probably has more CPU power to do that. (10th gen i7)
Native means playing effective playing at the rate they were encoded at … ie no oversampling in roon or DAC. Roon probably has more horse power in the PC to do that than the Ares II but some will like one over another another no matter who says what and how its done. they only one your choice has to please is you - hence trust your own ears. Some also prefer the HQPlayer upsampling too but that can add much horsepower to your system needs and maybe even another PC to do it.
There comes a point where one needs to just get on with enjoying the music
Another thing is you have no control of what the Ares will do when set to OS very NOS where in roon you can do to just up to PCM or DSD etc so many more options to play with…in the end it comes down to what you like best - I prefer not upsampling in roon, maybe OS in Denafrips is a different thing - I dont know. Maybe @VinshineAudio Alvin can time in.
I also own a Denafrips Ares II. I have played around quite extensively over sampling rates in both Roon and through the DAC as well as comparing DSD to PCM. I use both Tidal and locally stored files. For both PCM and DSD, I prefer letting BOTH Roon and the DAC over sample. I set Roon to Max or 784Hz and use the I think “smooth” setting in the Ares OS filter. Like others have stated, Roon gives you the ability to switch sampling rates on the fly so it’s very easy to make your own comparisons.
The NOS mode on the Ares sounds nice but a bit rolled off compared to the OS mode. I originally though the letting Roon do all the upsampling and playing the Ares in NOS would sound best, but I found my initial assumption to be incorrect at least for me.
I’d also recommend trying to convert redbook filles in Roon to DSD512 and DSD256 and try the Ares playing DSD which it does very well. For me, DSD sounds a bit cleaner compared to PCM put PCM engages me with the music more but you may prefer DSD. I did for for a few weeks but ended up going back to PCM.
If you DAC has the ability to play DSD natively like the Ares, make sure to select that option in the last screenshot
I think that’s about it. If you experience ticks or stuttering when playing DSD512, it means your server is being over taxed so you may need to step down to DSD256. Some even prefer the sonic qualities of DSD256 to DSD512.
Let’s us know what you think as well as comparisons between DSD and PCM.
and that’s the digital problem. Digital to analog is not perfect. PCM and DSD try to solve different parts of the “not perfect”. Different filters in PCM try. Upsampling and higher bitrates try to move the “not perfect” out of the audio range.
Now, most people are not bothered by the “not perfect” which is why we play in a niche audiophile space while the most sitting next to us on the bus are perfectly happy with lossy Spotify through lossy bluetooth. But those of us that are bothered by it are all bothered in different ways. This is why there are options. Options change where the “not perfect” arrives in the analog. It’s also why you just need to try every combination and be happy with the one that sounds best to you (which might also change based on what kind of music you’re listening to).
Now, if you’re still curious you can lookup what ringing is and how upsampling moves it around an impulse. Some people highly value linear phase filters over all else. Roon has a decent write-up specific to their DSP engine. But, ultimately, what sounds right to you is what’s right. It’s not only your gear but you’re the only one who has to listen to it (and therefore be bothered by it)
I would like to know if there is a theoretical benefit to do it some other way.
