Focus on Sound Quality (SQ), please !

Beside the current big problem with OTA update with Volumio2-x86 (update from 2.201 to 2.285 -on 7-Oct-2017, fully freeze up computer and get PC not working at all, need to reflash USB key with full image 2.285 to recover Volumio), i would like to highlight key point:

  • Sound Quality (SQ) of Volumio2-x86 is today lower than Daphile (LMS+Squeezelite) with same hardware and configuration setup. Volumio2-x86 SQ is also lower than JRiver setup with same hardware comparison.
  • I test deeply face to face whole configurations with 2x hardware configurations dedicated for audio powered by linear power supplies (MB Atom + ESI Juli@, and MB with AMD Phenom + high end USB output adapter for audio - clock and linear power separated).
  • It is a real pity for me as my key target was to use Volumio2-x86 as a pure high end DLNA renderer to get the best from powerful media server solutions (JRiver or Audivarna).
  • Main configuration target is to get the DLNA renderer running Volumio2-x86 (then, the ‘‘audio PC’’ with whole optimisations needed to get best sound and attached to music system) controlled remotely by ‘‘the control PC’’ (desktop powerful PC for lot of usage, but not optimised for audio) running JRiver or Audivarna.
    There is a huge market to get high end audiophile DLNA renderer capabilities from low cost devices (low power devices or fanless x86 solutions).
    GUI and OTA updates are lower priorities than audio performance, IMHO.
    Could you please stick with this key objective in your roadmap of development
    Thks

Hi,
as you might know, we have 3 main goals with Volumio:

  • Sound Quality
  • Ease of use
  • Universal audio format support

So, yes, we are always keen to improve sound quality. What you are referring IMHO has one main big oversight:
You want to achieve the best audio quality, but using DLNA renderer, which IMHO does not make much sense for one simple reason. It’s not possible to have bit accurate playback in DLNA renderer 100% of the times. Renderer and server always negotiate on file playback support by the renderer, and if for some reason the server thinks that the renderer does not support that kind of playback, it will resample it.
IMHO this is what’s happening in your case.

For this reason we never really optimized DLNA playback: we see it just a nice to have, mainly for convenience reasons. If you want to get the most of your files, you should instead play them directly with Volumio (on a NAS or USB drive). This is what we’ve optimized and you’ll see the results. Simply because we optimize what can really achieve the best quality possible.

I’ve just received a mail this week stating that Volumio (x86) sounds much better than Daphile…
Another tip: to make sure you get the most out of your system also check that Mixer is set to None (or Hardware if your DAC supports it).

I am curious to hear your feedbacks if you want to try what I suggested

Thank you for your explanation Michelangelo. I had the same conclusion last weekend and play my Hi Res music now with a direct Lan connection from an Android tablet. MPD rules, no more lame minidlna, WiFi, Dlna & bubbleupnp.

But what is the best way to shutdown volumio if you are using a usb harddisk and don’t want to damage it?

Using an external HDD or a NAS is better than placing the files on the sdcard?

@Hf4d2skwk8 there is no risk of damaging your hard drive, but if you want to be 100% sure turn off Volumio from its menu entry

@there is no difference in playing from USB or SD Card for files up to 24bits/96k, but if you have higher res files SD Card will give better results just because it avoids stressing the USB BUS of Raspberry PI, which can cause dropouts in audio with higher transfer rates

Hi, today I’ve implemented the first of a series of SQ optimizations. This one aims mainly for those using a USB DAC and hi res files.
Could you guys please test it and report back your findings?
updates.volumio.org/pi/volumio/2 … pi.img.zip

Hello, i have an i2s dac. I would like to know what is better in terms of sound quality for storing the music :
1- usb stick on rpi
2- sd card on rpi
3- minimserver on pc
4- using a nas like synology or other.

Except minimserver, they are all equal and suggested

Why would the sound quality of MinimServer be inferior to the others?

Not sure. Possibly because of transcoding issues? I find minimserver very useful in conjunction with volumio to access BBC radio stations, but wouldn’t use it to access any serious music.

Transcoding is an option (with MinimStreamer installed). But, assuming you’re not availing yourself of that option (which would require some not inconsiderable effort on your part to set up), I’m not sure what the issue is.

Umh. Why?