Video: picture stops while sound continues - Nightly/ Slacko

Booting, installing, newbie
Post Reply
Message
Author
User avatar
canbyte
Posts: 264
Joined: Sat 10 Jan 2009, 20:20
Location: Hamilton, Canada

Video: picture stops while sound continues - Nightly/ Slacko

#1 Post by canbyte »

I'm having problems watching youtube clips using Slacko's Nightly. The video stops moving while the sound continues. To get the picture moving again, you have to click on the round thing that scrolls along the bottom that shows the position you are in. It stops moving when this happens. The occurrence is intermittent, usually every few minutes. These are short videos not full length. Don't know what other details are needed except that it never happened on older puppies, so i hope someone can diagnose the problem.

Also i'm hoping someone can drop by on my other question relating to downloading/ installing a 'client' on puppy - the question is here -
http://www.murga-linux.com/puppy/viewtopic.php?t=68016

Thanks in advance
[color=orange]1. Dell Dimension E521, AMD Athln 64, 2 GHz 1.93GB ram,
Puppy 533 on CD, accesses flash drive only,
FFox Nightly12.0
2. Compaq P3 733Hz 375RAM
Printer: Oki C3400 > LAN [/color]

mcewanw
Posts: 3169
Joined: Thu 16 Aug 2007, 10:48
Contact:

Re: Video: picture stops while sound continues - Nightly/ Slacko

#2 Post by mcewanw »

Might be a conflict between adobe flash plugin and alsa sound modules (I had that same freezing effect on my system and kept having to click on the round scrolling thing too). The following fixed it for me: Try stopping and then starting alsa by opening a console and entering the following three lines. Be careful to enter each line below exactly (cut and paste useful); in particular 10alsa is with a zero and the slashes (/) are all required. (You have to wait for each step to complete before entering the following line). Do the above straight after you have booted into Slacko and after that immediately try running the youtube videos in the browser. Let me know if that works please and we can look into a more permanent solution that works between reboots.

Code: Select all

/etc/init.d/10alsa stop
/etc/init.d/10alsa start
retrovol -hide &
github mcewanw

User avatar
canbyte
Posts: 264
Joined: Sat 10 Jan 2009, 20:20
Location: Hamilton, Canada

#3 Post by canbyte »

Hi mcewanw
I tried those commands (but without rebooting) and it seems to work - tried a few clips, no problem. I'll try rebooting now and report back. fwiw, here is the response to the final command

[2] 31282
[1] exit 127
cannot read file: /root/.retrovolrc
using defaults...

(i couldn't remember how to cut/paste into the console - something about using the scroll wheel?)

rebooting & walking the dog.... You're right, the fix doesn't survive the reboot. Well, hopefully those lines can be done during startup. Hmm. How about a small file in the /root/startup directory? Or maybe where I was fooling around with another file executed on statup. I'll report back after I've had a look.

Well, definitely not to be put in xorg.conf! I'll try adding the two lines to the file /root/startup/retrovol_tray, which already has the third line of your script. processing..... Well, it started ok but the fix not implemented. Now I'll try putting that same script into a small file and try to execute it. ... Processing... Seems to work at least for awhile and it gives an easy way to just repeat at any time. I put the file in /root so it's easy to find.
Note to others adopting this fix: the browser sometimes shuts down during this op so be sure to terminate any work in progress first.

So, thanks very much though I won't mark the issue as resolved pending hopefully, a permanent fix.
[color=orange]1. Dell Dimension E521, AMD Athln 64, 2 GHz 1.93GB ram,
Puppy 533 on CD, accesses flash drive only,
FFox Nightly12.0
2. Compaq P3 733Hz 375RAM
Printer: Oki C3400 > LAN [/color]

mcewanw
Posts: 3169
Joined: Thu 16 Aug 2007, 10:48
Contact:

flash video freezing caused by alsa sound module conflict

#4 Post by mcewanw »

canbyte wrote: Note to others adopting this fix: the browser shuts down during this op so be sure to terminate any work in progress first.

So, thanks very much though I won't mark the issue as resolved pending a permanent fix.
My Seamonkey browser actually doesn't shutdown when I manually enter these lines, but definitely safer to close it before such actions. I also get the error message about /root/.retrovolrc not being found, but retrovol seems to start up okay again anyway.

To make the above happen at every boot I simply modified the script /root/Startup/retrovol_tray to become:

Code: Select all

#!/bin/sh
/etc/init.d/10alsa stop
/etc/init.d/10alsa start
retrovol -hide &
However, a more permanent fix is to run lsmod before the "fix" and check what sound module(s) get loaded. On my own system both snd_intel8x0 and snd_intel8x0m were amongst the modules being loaded but your sound card may well be a different one. Once you find the modules loaded before the fix, run lsmod again after the fix and compare. You will probably find less sound modules loaded this time. On my system, snd_intel8x0m was no longer being loaded now. Having identified whatever sound module(s) is no longer being loaded, simply blacklist it so that it no longer gets loaded at boot time. Then you won't need the stop or stop alsa lines above at all. On my system I therefore had to blacklist snd_intel8x0m.

The utility that you can use to blacklist modules is the Bootmanager configure bootup utility which you will fine in Slacko start menu -> System. After blacklisting the errant/flash-video-conflicting alsa sound module all should be fine thereafter. Please let me know if that is the case for the benefit of other users that no doubt have experienced the same problem with flash videos. Too many people immediately suspect the adobe flash plugin version and waste a lot of time trying different versions (as I did...) when the real problem may often be a loaded module conflict.

I believe there is probably a problem with Slacko's 10alsa script, so the maintainers of that really need to look into that to find a true permanent fix - but the above repair seems to work once it is done anyway.
github mcewanw

User avatar
canbyte
Posts: 264
Joined: Sat 10 Jan 2009, 20:20
Location: Hamilton, Canada

#5 Post by canbyte »

OK here goes. I'll record the output here and hope you can help figure it out.

First I had to look up Karl's instruction to copy in rxvt:
Normally highlighting with the mouse keeping left button held down, releasing the left button and press the middle button to paste

Before the fix:
Module Size Used by
snd_hda_codec_idt 39895 1
snd_hda_intel 16350 0
snd_hda_codec 54223 2 snd_hda_codec_idt,snd_hda_intel
snd_hwdep 3704 1 snd_hda_codec
snd_pcm_oss 27363 0
snd_mixer_oss 9846 1 snd_pcm_oss
snd_pcm 46865 3 snd_hda_intel,snd_hda_codec,snd_pcm_oss
snd_page_alloc 4685 2 snd_hda_intel,snd_pcm
snd_seq_dummy 907 0
snd_seq_oss 19155 0
snd_seq_midi 3248 0
snd_seq_midi_event 3636 2 snd_seq_oss,snd_seq_midi
snd_rawmidi 11838 1 snd_seq_midi
snd_seq 32220 6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_midi_event
snd_seq_device 3541 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi,snd_seq
snd_timer 11743 2 snd_pcm,snd_seq
snd 32906 12 snd_hda_codec_idt,snd_hda_intel,snd_hda_codec,snd_hwdep,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_seq_oss,snd_rawmidi,snd_seq,snd_seq_device,snd_timer
soundcore 3321 1 snd
b44 19474 0
ssb 34121 1 b44
mmc_core 50083 1 ssb
pcspkr 1195 0
k8temp 2351 0
i2c_nforce2 4056 0
i2c_core 12559 1 i2c_nforce2
shpchp 17907 0
evdev 5816 0
button 3275 0
processor 20513 0
thermal_sys 9722 1 processor
hwmon 965 2 k8temp,thermal_sys
fuse 47599 0
aufs 119852 75
squashfs 18852 1
#

After running the fix:
Module Size Used by
snd_hda_codec_idt 39895 1
snd_hda_intel 16350 0
snd_hda_codec 54223 2 snd_hda_codec_idt,snd_hda_intel
snd_hwdep 3704 1 snd_hda_codec
snd_pcm_oss 27363 0
snd_mixer_oss 9846 1 snd_pcm_oss
snd_pcm 46865 3 snd_hda_intel,snd_hda_codec,snd_pcm_oss
snd_page_alloc 4685 2 snd_hda_intel,snd_pcm
snd_seq_dummy 907 0
snd_seq_oss 19155 0
snd_seq_midi 3248 0
snd_seq_midi_event 3636 2 snd_seq_oss,snd_seq_midi
snd_rawmidi 11838 1 snd_seq_midi
snd_seq 32220 6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_midi_event
snd_seq_device 3541 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi,snd_seq
snd_timer 11743 2 snd_pcm,snd_seq
snd 32906 12 snd_hda_codec_idt,snd_hda_intel,snd_hda_codec,snd_hwdep,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_seq_oss,snd_rawmidi,snd_seq,snd_seq_device,snd_timer
soundcore 3321 1 snd
b44 19474 0
ssb 34121 1 b44
mmc_core 50083 1 ssb
pcspkr 1195 0
k8temp 2351 0
i2c_nforce2 4056 0
i2c_core 12559 1 i2c_nforce2
shpchp 17907 0
evdev 5816 0
button 3275 0
processor 20513 0
thermal_sys 9722 1 processor
hwmon 965 2 k8temp,thermal_sys
fuse 47599 0
aufs 119852 75
squashfs 18852 1
#
My eyes are going squirrel/y - i can't see any diff! But a youtube vid works fine.
[color=orange]1. Dell Dimension E521, AMD Athln 64, 2 GHz 1.93GB ram,
Puppy 533 on CD, accesses flash drive only,
FFox Nightly12.0
2. Compaq P3 733Hz 375RAM
Printer: Oki C3400 > LAN [/color]

mcewanw
Posts: 3169
Joined: Thu 16 Aug 2007, 10:48
Contact:

#6 Post by mcewanw »

I copied the before and after lists you supplied into geany and made two texts files out of them. I then compared them with the utility you can find in Slacko menu called "Xdiff". They are indeed identical...

Either your problem is something else (related to alsa it would seem) or you didn't actually get a listing "before the fix". You would have to make sure that your computer was in the state where it was intermittently freezing videos, per your first post, before running lsmod to obtain the "before the fix" lsmod result. i.e. immediately after rebooting your system with no fix yet applied. If you did that and then applied the fix (and checked videos were no longer freezing) before running lsmod again for the second result and the result is identical as you submitted then I am not sure what is going on!
github mcewanw

User avatar
canbyte
Posts: 264
Joined: Sat 10 Jan 2009, 20:20
Location: Hamilton, Canada

#7 Post by canbyte »

That's a nice little comparator. I did as suggested after rebooting and ensuring that videos were stalling. Xfdiff says the two files are still the same - however something must be not working in that program. I do see several differences to the last post fix in some of the second numbers ('used by' column). So for instance, in the second line, the last number is 3 before the fix and 0 after.
This difference didn't show up in xdiff. 7th line the number changes from 4 to 3. Do you think this is the source of the problem?



Here's the full listing pre fix -
Module Size Used by
snd_hda_codec_idt 39895 1
snd_hda_intel 16350 3
snd_hda_codec 54223 2 snd_hda_codec_idt,snd_hda_intel
snd_hwdep 3704 1 snd_hda_codec
snd_pcm_oss 27363 0
snd_mixer_oss 9846 1 snd_pcm_oss
snd_pcm 46865 4 snd_hda_intel,snd_hda_codec,snd_pcm_oss
snd_page_alloc 4685 2 snd_hda_intel,snd_pcm
snd_seq_dummy 907 0
snd_seq_oss 19155 0
snd_seq_midi 3248 0
snd_seq_midi_event 3636 2 snd_seq_oss,snd_seq_midi
snd_rawmidi 11838 1 snd_seq_midi
snd_seq 32220 6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_midi_event
snd_seq_device 3541 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi,snd_seq
snd_timer 11743 3 snd_pcm,snd_seq
snd 32906 16 snd_hda_codec_idt,snd_hda_intel,snd_hda_codec,snd_hwdep,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_seq_oss,snd_rawmidi,snd_seq,snd_seq_device,snd_timer
soundcore 3321 1 snd
b44 19474 0
ssb 34121 1 b44
mmc_core 50083 1 ssb
pcspkr 1195 0
k8temp 2351 0
i2c_nforce2 4056 0
i2c_core 12559 1 i2c_nforce2
shpchp 17907 0
evdev 5816 0
processor 20513 0
button 3275 0
thermal_sys 9722 1 processor
hwmon 965 2 k8temp,thermal_sys
fuse 47599 0
aufs 119852 72
squashfs 18852 1
#
[color=orange]1. Dell Dimension E521, AMD Athln 64, 2 GHz 1.93GB ram,
Puppy 533 on CD, accesses flash drive only,
FFox Nightly12.0
2. Compaq P3 733Hz 375RAM
Printer: Oki C3400 > LAN [/color]

mcewanw
Posts: 3169
Joined: Thu 16 Aug 2007, 10:48
Contact:

#8 Post by mcewanw »

Xdiff on Slacko did show the differences on the two different lists for me. I believe the number you refer to is a list of how often the given module is being used. man lsmod or google will reveal more about that. Unfortunately I don't see what is going on in terms of finding any module to blacklist. I can only suggest you run alsaconf first and see if the problem goes away thereafter or after running the fixes I suggested. Otherwise, I have to leave the matter for someone else to find a more complete fix for. Meanwhile, at least you have it working :-)
github mcewanw

Bruce B

#9 Post by Bruce B »

Video has less priority than sound. If things are too slow the player will drop video frames and try and keep the sound playing.

You can download the youtube video, you can tell when it is finished downloading by the progress bar of the player. Pause it or let it finish, but don't close the page.

Then using another player, play the video from the /tmp directory. This will help you in the diagnostics.



~

Post Reply