[sizzlechest]

the cause for the hiccup and gap might be related. solving one might also fix the other.

[jmarshall]

i added debug stuff yesterday + increased the caches a little.

there has to be a bunch of reorganisation done before i can open up the caches for modification.

i'm interested in your debug output when such hiccups occur.

just for fun yesterday i ran my network at 10mbps - no issues whatsoever.

cheers,
jonathan

[axavier]

i added debug stuff yesterday + increased the caches a little.

there has to be a bunch of reorganisation done before i can open up the caches for modification.

i'm interested in your debug output when such hiccups occur.

just for fun yesterday i ran my network at 10mbps - no issues whatsoever.

cheers,
jonathan

i'l see if i can do a build and test today to see if we can find it. do we need to set the debug to level 0 in the xboxmediacenter.xml file to get that debug info? mine is at one right now at the moment.

thansk for your help on this marshal!!

[jmarshall]

0 is what it should be at iirc.

[axavier]

0 is what it should be at iirc.

can you please explain marshal? also are yu guys thingk of making those settings (incresing / decreasing log levels) available via the xbmc interface instead of changing the xml file directly?

[jmarshall]

<loglevel>0</loglevel>

press x+y on boot to set this up.

and i don't know if it's ever gonna be available from the gui.

[Aron Parsons]

edit: files play fine off the hdd and on a smb share. xbms fails.

i'm getting this with builds after the buffers were increased and debugging was added. playback is fine with a build prior to the buffer/debug changes in paplayer.

04-05-2005 18:03:28 info activating window id: 10501
04-05-2005 18:03:28 info attempting to default to: xbms://192.168.0.7:1400/music
04-05-2005 18:03:28 info success! opened destination path: xbms://192.168.0.7:1400/music
04-05-2005 18:03:28 debug initializing fs_seg..
04-05-2005 18:03:28 debug fs segment @ 0x866390
04-05-2005 18:03:33 debug udpclient tx: 16332 -> 'kai_client_caps;013;'
04-05-2005 18:03:37 debug paplayer: starting next track whilst already playing - killing our output buffers
04-05-2005 18:03:37 info mp3codec: loaded decoder at 008a1010
04-05-2005 18:03:37 info pap player: playing xbms://192.168.0.7:1400/music/3 doors down/2000 - the better life/3 doors down - 01 - kryptonite.mp3
04-05-2005 18:03:37 info pap player: new audiodevice created. chunklen 4120
04-05-2005 18:03:37 debug paplayer: thread started
04-05-2005 18:03:37 debug playback has started
04-05-2005 18:03:37 notice start led control
04-05-2005 18:03:37 notice dim lcd on playback
04-05-2005 18:03:37 error mp3codec: error reading file
04-05-2005 18:03:37 info pap player: error while decoding 1
04-05-2005 18:03:37 debug playback has finished
04-05-2005 18:03:37 notice start led control
04-05-2005 18:03:37 notice dim lcd on playback
04-05-2005 18:03:38 debug paplayer: starting next track whilst already playing - killing our output buffers
04-05-2005 18:03:38 info mp3codec: loaded decoder at 0089e010
04-05-2005 18:03:38 info pap player: playing xbms://192.168.0.7:1400/music/3 doors down/2000 - the better life/3 doors down - 02 - loser.mp3
04-05-2005 18:03:38 info pap player: new audiodevice created. chunklen 4120
04-05-2005 18:03:38 debug paplayer: thread started
04-05-2005 18:03:38 debug playback has started
04-05-2005 18:03:39 notice start led control
04-05-2005 18:03:39 notice dim lcd on playback
04-05-2005 18:03:39 error mp3codec: error reading file
04-05-2005 18:03:39 info pap player: error while decoding 1

[axavier]

marshal mine works but i'm still getting the hiccups even after incresing the buffers:

nothing "funky" on the logs

04-05-2005 20:10:09 info pap player: loaded decoder at 00a7c850
04-05-2005 20:10:09 info pap player: playing smb://192.168.1.101/music/rush/a show of hands (the rush remasters)/rush - 03 - subdivisions.mp3
04-05-2005 20:10:09 info pap player: channels 2 : sample rate 44100
04-05-2005 20:10:09 info pap player: chunklen 4120
04-05-2005 20:10:09 notice start led control
04-05-2005 20:10:09 notice dim lcd on playback
04-05-2005 20:10:12 info get url: http://post.audioscrobbler.com/?hs=true&p=1.1&c=xbm&v=0.1&u=xxxxxxx
04-05-2005 20:10:12 error unable to connect to post.audioscrobbler.com: 10061
04-05-2005 20:10:12 notice audioscrobbler: unable to handshake: sleeping...

i have checked my mp3's and they display the right info in foobar2000.

also, has anybody noticed some problems with audioscrobbler?

[Aron Parsons]

okay, switched to smb shares for now. i guess for the audio it doesn't really matter, but i'll have to keep the xbms share for video (smb causes much stuttering on my wifi for videos).

same as axavier for the hiccups; they occur, nothing in the log related to it though. looks just like his.

also, replaygain from ape tags works great. i have a question though; how is it handling the tags, and will the ape tags eventually be cached in the database to avoid reading them every time? (a bit premature to be asking, but it seems logical to do).

[jmarshall]

aron parsons: i think xbms doesn't like 128k buffers - i see a note about it in the file copying procedure as well.

in either case, it doesn't seem to matter much as you get stuttering over smb as well.

the issue is i'm not sure why the stutter is occuring. it must be very minor for the logging to not pick it up. i'll add a callback from the actual audiostream and log in that.

in any case, i'm wanting to get the output stage more integrated with the player so that we can have 2 streams playing at once (for instance) without issues. this should allow better control over the way things buffer.

please try some other formats (eg ape or ogg - bobbin007 just did up a nice codec for ogg ) and see if they produce the same issue(s).

cheers,
jonathan

[axavier]

aron parsons: i think xbms doesn't like 128k buffers - i see a note about it in the file copying procedure as well.

in either case, it doesn't seem to matter much as you get stuttering over smb as well.

the issue is i'm not sure why the stutter is occuring. it must be very minor for the logging to not pick it up. i'll add a callback from the actual audiostream and log in that.

in any case, i'm wanting to get the output stage more integrated with the player so that we can have 2 streams playing at once (for instance) without issues. this should allow better control over the way things buffer.

please try some other formats (eg ape or ogg - bobbin007 just did up a nice codec for ogg ) and see if they produce the same issue(s).

cheers,
jonathan

i'll try to encoge some ogg files and test tonight....also. do i need a dll (like for mp3's) in order to play ogg files?

[jmarshall]

yes - it's in cvs.

[petergillett]

just an update on some 54g tests.

been running a packet sniffer and there seems to just be a pause every now and then, even a second or so right in the middle of a directory browse. was watching this over remote desktop - totally smooth, no rd packet more than 0.1sec apart. however xbox packets had occasional pauses, sometimes over a second.

then ran the standard network log. tried doing some file copies and stuff. graph (bytes sent) was just up/down up/down.

strangely though, xbox from laptop is fine, graph 95% at full speed (though laptop is definitely closer/less obstructed to access point).

server to laptop is fine. graph again 95% at full.

xbox even has advantage as it is wired to the access point, where all others are straight wireless through the ap. all links are 48/54ish - nothing silly like dropping down to 5m or anything.

tried all sorts of settings on the wireless, forced lower rate etc, but no difference. just get full speed for that rate with just as many pauses. will try wiring the problem machine and see what differences it make, just logistics won't let me for a few days.

both machines xp pro sp2, only other thing is laptop is much higher spec 2.8p4/512 against 1m duron/256

[axavier]

tested today after updating form cvs. the hiccups are still happening. cache memory set at default levels

[jmarshall]

axavier: see if you have the new threaded file reader in your version - you can check by seeing if there's a filereader.cpp present in the xbmc/cores/paplayer folder.

if so, open up mp3codec.cpp in the same folder and look at the constructer - you will see it initializes the filereader class (m_filepap) with some default settings. depending on what version you have there'll either be 2 or 3 parameters.

in the case of 2 parameters, increase the second parameter to say 16 or so.

in the case of 3 parameters, increase the first parameter to 65536*16 or so.

see how it goes.

cheers,
jonathan