hi, I did this kind of A/B tests a couple of years before when I owned a pi3+higiberry HAT. I noticed differences comparing Moode, volumio, picore, rune, hifiberryOS.
My own explanation was that pi was such a low end hardware not designed for hifi that way OS interfere with motherboard components(CPU,RAM,USB,Ethernet) could induce some digital differences (jitters?). Look at Moode’s advanced option or cpu affinity plugin in volumio, EVERYTHING Count !
Good news is that you can choose the best option based on your tests
blinded tests look quite difficult, isn’t it ? … and deserve nothing here since biased answer is still a good answer (meaning dolorless)when comparing free solution
Citation
RuneOS also includes a number of kernel profiles ,
which are switchable sets of performance tweaks acting on some system kernel parameters. This does not have anything to do with Digital Signal Processing or other sound effects – the output is kept untouched (bit perfect).
It can happen that these parameters introduce an audible impact on the overall sound quality, acting on kernel latency parameters (and probably on the amount of overall jitter). Sound results may vary depending on each setup, so users can choose a profile, according to their personal tastes.
A beginning of explanation around “bitperfect is not the sole parameter” taken from 2014 rune audio OP
yes rune make adjusting kernel profile easily accessible for user but this is not only kernel profile I guess, volumio and Moode are both based on debian but each OS is cooked over a heavy set of own patches to adjust things here and here and make software more “audiophile”. Picore is build from scratch and very optimized because memory footprint is another key, archphile, based on arch like rune, is also known to be heavily patched version of baseOS. I’ve always found that volumio had a very engaging and shiny sound compared to other OS, but my preference went to Moode or picore for better dinamic and stability in time.Of course this is my own experience based on my old setup, every crowler need to test everything, blind or not
I guess some replies have already pointed at the reasons.
Jitter could be one reason of the differences you are hearing, even though a bit is a bit, when that bit comes in makes all the difference of the world.
The second reason is noise, in particular high frequency noise (in the MHz range, not the 50/60 Hz hum from the power socket) generated by the switching power supplies and the computer boards and chips, that if not properly isolated can leak through the S/PIDIF interface you are using to connect to your DAC. The DAC can be heavily affected by that noise signal, if not properly engineered.
Toslink connection will avoid that noise leak, to some extent, but introduce more Jitter in the process (pick your poison).
Jitter issues on cheap digital tranports like these (Pi based with non-hard-realtime SW), can be mitigated when using an async interface, like USB. The DAC is then in charge of reclocking all the samples that come in through the USB connection. The quality of the DAC will make the difference here. And a good DAC will also be able to reject that high frequency noise mentioned above.
I am using Volumio on a Tinkerboard, connected via USB to a very well made DAC (Marantz SA-KI) and I can attest that the perormances I get out of this combo are of outstanding level. But it is the DAC really doing an outstanding job in my case, with reclocking and noise rejection.
In the A/B testing on your setup, did you like more Moode or Volumio?
Interesting reading thx. Personally moode had my favor on my pi3 + hifiberry OS. Now I m on tinkerboard (primo) like you connected via USB to an excellent nuprime dac10. No more quality problem with volumio and, unfortunately, flashing other OS on emmc is not so easy so I can’t A/B test and investigate anymore
Both Volumio and Moode customise the kernel.
They both configure ALSA and MPD differently.
No doubt there’s a couple other hidden aspects in here somewhere as well.
I’m happy now as I have learnt a lot and have the ability to tweak either software to try get my most preferred sound on my humble system (in my home and to my ears).
I was quite surprised by the difference between Volumio and Daphile. Volumio was more transparent, dynamic and detailed than Daphile. I wasn’t sure I believed my own ears so I sent a friend a file made from each without prompting him and he came back with the exact same thoughts.
Shot my “bits-is-bits” philosophy all to hell.
I have just done a similar test: I loaded Volumio and MoOde onto two Mercury2 streamers. Both with USB storage attached and loaded with the same song files. Both OS disc images from the supplier Pi2Design. After I listened I asked family to have a listen and tell me if there is any difference between the two song sets. Everyone picked out a clear difference. Everyone described it in the same way. They have no idea what I was changing so I’ll call that a “blind test”.
Then I swapped the cards to see if the difference was hardware on the 2 Mercury’s. The sound followed the software.
Volumio was described as “more punchy”, “more sibilant”, “more detail”. MoOde as “smoother”, “softer”.
I was surprised at just how obvious the differences were.
NOTE: I have found out that Moode do not, in fact customise the Kernel.
I am of the understanding that Volumio do.
Eitherway, as far as I know both try to cut out unrequired supporting services in their distribution.
And ALSA and MPD are at least configured differently which is likely the main reason they sound different.
I’m happy for anyone to correct my understanding for these statements.
At the end of the day, it does come down to personal preference which is also influenced by your hardware including amplifier and speakers, the room acoustics and of course your ears!
Tim Curtis sent me this in response to the same assumtions I posted on DIYAudio forum.
"moOde uses the stock Raspberry Pi Linux kernel and config. The only customizations we ship are 2 kernel drivers that are patched to bump the max sample rate they support and 1 WiFi driver we compile for the nice Comcast CF-912AC dual-band adaptor.
You can examine the 3 packages (aloop, pcm1794a, rtl88xxau) in our package build repo.
MPD and ALSA are also stock in our releases as is the rest of RaspiOS. The versions we ship with a given release are generally the latest from the maintainers and are specified in our package and image build repos."
Volumio, Moode Audio, PicorePlayer are all bit perfect sw player! Under standard conditions there is no reason to find even minimal audio differences. It’s just psychoacoustic!
Considering both should be able to deliver bit perfect streams to my DAC, can anyone advise why Volumio ad Moode sound different?
So which one sounds better?
I’m not familiar with Moode - for a same source setup they should be pretty similar I assume, with all other things the same (kind of like comparing software players foobar vs ??). I am assuming there is no loss with WLAN casting type setups (I also have a Rpi4 → USB to DAC, and very happy with the setup)
May be oversimplified but probably comes down to features and subscription cost but good to test both.
In the same gear/equipment they are not “quite similar”, but exactly THE SAME! You can measure the test or there is already someone who has measured! Please, search on the web what means “bit perfect” player like Volumio or MoOde Audio
… and deserve nothing here since biased answer is still a good answer (meaning dolorless)when comparing free solution
, interesting to get the details behind changelog. Thx