Openelec RPi v3.95.3 bug thread
#1
OpenELEC-RPi.arm-3.95.3

Quote:Every time I open System>system>video output, I get the message "Save resolution. Would you like to keep this change"
(2014-03-20, 22:17)606u Wrote: I have the same two problems, except I got the message "Save resolution..." twice: here is a short clip; message also appears when video settings are opened right after a restart (while Pi is restarting TV and receiver are switched on and TV's active input is the receiver).
This issue is still present

OMXPlayer:
FLAC
Variable 3 048 Kbps
4 channels Front: L R, Side: L R
48.0 KHz 24 bits

Play is NOT gapless

PAPlayer:
Nums_7dot1_24_48000.wav
PCM
Little
Signed Constant
9 216 Kbps
8 channels Front: L C R, Side: L R, Back: L R, LFE
48.0 KHz 24 bits

===> LFE (subwoofer) is not mapped
Channels 7 and 8 are mapped only with Back L and Back R instead of Back-Front L and Back-Front R

These issues are the same as MilhouseVH builds in OpenELEC Testbuilds for RaspberryPi Part 2 thread
Config, video/audio player:
3T HDD <USB> Odroid N2+ / CoreElec <HDMI> Denon AVR-2313 <HDMI> LG TV 55UF860V
                                          <nfs wired> Linksys WRT32X router <USB> 4T HDD
Reply
#2
(2014-04-02, 09:06)MrNice Wrote: Every time I open System>system>video output, I get the message "Save resolution. Would you like to keep this change"
Can you upload the guisettings.xml file that is present when you get this error? Also report output of "tvservice -s".

Quote:OpenELEC-RPi.arm-3.95.3
OMXPlayer:
..
Play is NOT gapless

OMXPlayer is not meant to be gapless. That's the main reason we made PAPlayer the default.

Quote:PAPlayer:
Nums_7dot1_24_48000.wav
..
===> LFE (subwoofer) is not mapped

Didn't you say you had speaker setup as 5.0? You won't get LFE mapped unless you have an LFE channel.
That is by design (xbmc does this on all platforms).
Reply
#3
(2014-04-02, 12:21)popcornmix Wrote:
(2014-04-02, 09:06)MrNice Wrote: Every time I open System>system>video output, I get the message "Save resolution. Would you like to keep this change"
Can you upload the guisettings.xml file that is present when you get this error? Also report output of "tvservice -s".
Sorry, it happened the first time but I can't reproduce it now.


Quote:
Quote:PAPlayer:
Nums_7dot1_24_48000.wav
..
===> LFE (subwoofer) is not mapped

Didn't you say you had speaker setup as 5.0? You won't get LFE mapped unless you have an LFE channel.
That is by design (xbmc does this on all platforms).
Yes my config is 5.0 but with OMXPlayer LFE is mapped with FL and FR (large band) and channels 7 and 8 are mapped with FL-SBF and FR-SBR.

If you don't have LFE, downmixing is to use low frequency from FL and FR bass speakers (this config is mainly used by audio fans).
If you don't have surround (side) speakers but only surround back speakers the downmixing is to use Front and surround back at the same time.
This is the way OMXPlayer works but not PAPlayer, IMHO this is a bug.
Config, video/audio player:
3T HDD <USB> Odroid N2+ / CoreElec <HDMI> Denon AVR-2313 <HDMI> LG TV 55UF860V
                                          <nfs wired> Linksys WRT32X router <USB> 4T HDD
Reply
#4
The file you reference has these channels (in this order):
Code:
FL      FR      C      LFE      BL      BR      SL     SR

omxplayer when mapping to 5.0 uses:
Code:
FL      FR      C      FL/FR    BL      BR      FL/BL  FR/BR

paplayer when mapping to 5.0 uses:
Code:
FL      FR      C      -        BL      BR      BL     BR

Now, I believe omxplayer uses the mapping that xbmc used to use.
paplayer uses the mapping that xbmc currently uses.

I don't think either are wrong. They are both deliberately making a decision about how to map channels that don't fit the speaker layout.

However if anything is going to change, it would be making omxplayer match paplayer as that at least has consistency with other xbmc platforms.
Reply
#5
I agree regarding the mapping.

However, with PAPlayer everything depend of the studio mixing people:
For LFE
- If they cut the low frequency for FL-FR at let say 60Hz and send below frequencies to LFE there is a loss
- If they send sound effects only to LFE there is a loss

For surround
- There is a loss in the feeling of the sound position, everything is coming from the back.
We could compare with stereo, if the 2 channels are totally separated. In this case there is no middle feeling ie: the singer is at left or at right, never in the middle. Strange!
In the same way if the record is with FL, C, FR with the singer in the Center only, if you downmix to stereo removing the center, the singer is gone. Bad!

Do you understand my opinion? I really thing there is a good and a bad way to downmix.
Config, video/audio player:
3T HDD <USB> Odroid N2+ / CoreElec <HDMI> Denon AVR-2313 <HDMI> LG TV 55UF860V
                                          <nfs wired> Linksys WRT32X router <USB> 4T HDD
Reply
#6
I've checked with Fernetmenta (author of new audio engine) and he responded:
Quote:"The LFE channel must therefore be considered as, at most, an enhancement, and definitely not as an essential part of the mix"
ITU-R BS.775-3 Annex 7 (LFE)

Means that we don't consider LFE with the default downmix matrix.

So, not mixing in the LFE is considered the correct thing to do.
Reply
#7
We intend let users configure the downmix matrix in a future version of XBMC.
Reply
#8
(2014-04-04, 15:03)FernetMenta Wrote: We intend let users configure the downmix matrix in a future version of XBMC.
Many thanks FernetMenta this is the best way, let user the choice.

I'd glad to know what is "future", months, years?
Config, video/audio player:
3T HDD <USB> Odroid N2+ / CoreElec <HDMI> Denon AVR-2313 <HDMI> LG TV 55UF860V
                                          <nfs wired> Linksys WRT32X router <USB> 4T HDD
Reply
#9
The future is near. First iteration of audio DSP has been submitted for review:
https://github.com/xbmc/xbmc/pull/4402

Within this context we think about how to manipulate the downmix matrix too.
Reply
#10
That's very good FernetMenta!

I'd guess this will be in a MilhouseVH build soon. I'll be happy to test it.
Config, video/audio player:
3T HDD <USB> Odroid N2+ / CoreElec <HDMI> Denon AVR-2313 <HDMI> LG TV 55UF860V
                                          <nfs wired> Linksys WRT32X router <USB> 4T HDD
Reply
#11
HI,

I've noticed what i believe is a bug but for all i know, it may be by design..

I've been backfilling a lot of old tv shows and i had to do a full scan on a fresh source.. whilst the scan was going on, i was getting notifications saying other things had finished downloading.. every time i received a notification, the scan would stop...

is this is a bug or is it by design?
Reply
#12
(2014-04-07, 14:11)danmedhurst Wrote: I've noticed what i believe is a bug but for all i know, it may be by design..

I've been backfilling a lot of old tv shows and i had to do a full scan on a fresh source.. whilst the scan was going on, i was getting notifications saying other things had finished downloading.. every time i received a notification, the scan would stop...

is this is a bug or is it by design?

I don't know. It's unlikely to be Pi specific, so may be worth asking in a general xbmc forum.

Might be worth trying (http://wiki.xbmc.org/?title=Advancedsettings.xml)
Code:
<videoscanner>
    <ignoreerrors>true</ignoreerrors> <!-- Set to true to silently ignore errors while scanning videos. This prevents the error dialogue box, so you don't have to keep hitting "yes" to keep scanning.-->
</videoscanner>
Reply
#13
Right, it's not Pi specific. This is by design. The library update trigger stops the current execution in case it's already running.
See https://github.com/xbmc/xbmc/blob/master....cpp#L1394
Always read the online manual (wiki), FAQ (wiki) and search the forum before posting.
Do not PM or e-mail Team-Kodi members directly asking for support. Read/follow the forum rules (wiki).
Please read the pages on troubleshooting (wiki) and bug reporting (wiki) before reporting issues.
Reply
#14
(2014-04-07, 14:50)vdrfan Wrote: Right, it's not Pi specific. This is by design. The library update trigger stops the current execution in case it's already running.
See https://github.com/xbmc/xbmc/blob/master....cpp#L1394

That code seems to stop scanning if a second "updatelibrary video" command is triggered.
But should it stop because an unrelated notification popped up?
Reply
#15
(2014-04-07, 15:02)popcornmix Wrote:
(2014-04-07, 14:50)vdrfan Wrote: Right, it's not Pi specific. This is by design. The library update trigger stops the current execution in case it's already running.
See https://github.com/xbmc/xbmc/blob/master....cpp#L1394

That code seems to stop scanning if a second "updatelibrary video" command is triggered.
But should it stop because an unrelated notification popped up?

This makes sense.. if i think about NZBDrone is not only triggering a notification but a single episode library update.. that explains it!!
Reply

Logout Mark Read Team Forum Stats Members Help
Openelec RPi v3.95.3 bug thread0