Kodi Community Forum
Keep Audio Device Alive - doesn't seem to work? - Printable Version

+- Kodi Community Forum (https://forum.kodi.tv)
+-- Forum: Support (https://forum.kodi.tv/forumdisplay.php?fid=33)
+--- Forum: General Support (https://forum.kodi.tv/forumdisplay.php?fid=111)
+---- Forum: OS independent / Other (https://forum.kodi.tv/forumdisplay.php?fid=228)
+---- Thread: Keep Audio Device Alive - doesn't seem to work? (/showthread.php?tid=205205)



Keep Audio Device Alive - doesn't seem to work? - keiron99 - 2014-09-26

I'm running XBMCbuntu (Gotham 13.2) on an Intel NUC, and connected via HDMI to a Rotel processor.

When playing audio, it misses the first beat of a song.

I always assumed that this is because the Rotel doesn't "wake up" until it receives a signal.

I also assumed that the "Keep Audio Device Alive" in Audio Settings, which i set to "Always", would prevent this from happening. But this is not the case, and I still get that irritating dropping of the first fraction of a second.

Can anyone advise on this? Thanks.


RE: Keep Audio Device Alive - doesn't seem to work? - fritsch - 2014-09-26

Keep Audio alive sends noise - an AVR cannot ignore that. If you use "Best Match" though, the device is reopened whenever the format of the next song changes -> it's the audiophile mode.

To see what I mean:
Keep Audio alive: always
Choose: Fixed
Samplerate: 48000
Channels: Whatever you want

and see the offset disappears.


RE: Keep Audio Device Alive - doesn't seem to work? - keiron99 - 2014-09-26

By jove, fritsch, it works! This has been bugging me for years!

However. I have set it to "fixed" and set the "limit sampling rate" to 192.

Now, everything I play, regardless of whether it's 2 channels or 5.1, shows "multichannel" in my receiver display. Everything also shows 192k.

However, to my ears, it sounds exactly as it does when set to "best match".

What is happening exactly? Is the music being "upsampled"? Is it in theory degrading the sound?


RE: Keep Audio Device Alive - doesn't seem to work? - fritsch - 2014-09-26

Have not much time. Retry with: Optimized.

Most is right of what you said.


RE: Keep Audio Device Alive - doesn't seem to work? - FernetMenta - 2014-09-26

192khz is nonsense, no matter if this is the original source or if you resample. In any case it will decrease quality: http://xiph.org/~xiphmont/demo/neil-young.html


RE: Keep Audio Device Alive - doesn't seem to work? - fritsch - 2014-09-26

@keiron99: I think, you are the perfect candidate for: http://forum.xbmc.org/showthread.php?tid=204019&pid=1793232#pid1793232

Do the above with "Best Match".


RE: Keep Audio Device Alive - doesn't seem to work? - keiron99 - 2014-09-27

(2014-09-26, 22:09)fritsch Wrote: @keiron99: I think, you are the perfect candidate for: http://forum.xbmc.org/showthread.php?tid=204019&pid=1793232#pid1793232

Do the above with "Best Match".

Will do. Do you want me to test files posted in post 13 or 18?


RE: Keep Audio Device Alive - doesn't seem to work? - fritsch - 2014-09-27

Those 2 x 3 samples. Classic and other. Also compare with the original posted.

Edit: Yeah Post 18


RE: Keep Audio Device Alive - doesn't seem to work? - keiron99 - 2014-09-27

(2014-09-26, 20:29)FernetMenta Wrote: 192khz is nonsense, no matter if this is the original source or if you resample. In any case it will decrease quality: http://xiph.org/~xiphmont/demo/neil-young.html

I am minded to agree. It's a marketing gimmick.

With this in mind, and given I have to fix the sample rate to avoid the problem of the beginning of songs being clipped (I tried with "optimized" and that clips the beginning too), than what, in theory, would be the best sample rate to set? 99% of my music is 44.1khz. Does it decrease the quality of this if I set XBMC fixed at a higher rate?


RE: Keep Audio Device Alive - doesn't seem to work? - fritsch - 2014-09-27

Make the test above and see - this was exactly made for testing this.

Those were 192/24 bit samples downsamples to 44.1 khz by three different quality resamplers.

The challenge is:
- Hear the difference between the original and the "best resampler"
- Therefore you need to find out the "best resampler" first :-)