Posts: 3,909
Joined: Dec 2004
Reputation:
20
Nuka1195
Skilled Python Coder
Posts: 3,909
while streaming using daap, i noticed after the first song gets near the end, the music overlay just stays on the first song. i push the right thumstick down to call up the button menu and do a reset. the gui freezes, ftp still works and the following is in the log.
14-05-2005 16:51:21 debug paplayer::output buffer size only 0 bytes. pcm buffer size 179712
14-05-2005 16:55:44 notice stop all
14-05-2005 16:55:44 notice stop time server
14-05-2005 16:55:44 notice stop dvd detect media
14-05-2005 16:55:44 notice stop fancontroller
14-05-2005 16:55:44 notice stop mplayer
all my songs in the itunes playlist are mp3's and i do have in_mp3.dll, and it is using paplayer. the stop mplayer does not show up if i go straight to reset without playing any songs.
edit: the may 6th build has the mplayer message also and daap: works fine for that build. oops.
edit: final post on this and i'm done bothering you. disabling id tags makes no differenc.
Posts: 152
Joined: Oct 2003
Reputation:
0
the hiccups are definetly gone. i've been playing music while maxing out the wireless connection (transferring games to the xbox), and playback was flawless. i also switched back to xbms and the delays between songs is less, but still not gapless. the initial delay is selecting a song is good once again too (i think that was just smb being slow). once jmarshall makes the changes he mentioned about pre-buffering, i'm going to be incredibly happy (more so than now).
Posts: 152
Joined: Oct 2003
Reputation:
0
found a new issue (or maybe it's been there since after i tried smb). pap doesn't read ape tags from a xbms share, but does from a smb share. playback from the xbms share is fine (and in fact browsing is faster), it's just the ape tags aren't read. could you take a look at this jmarshall?
Posts: 12
Joined: May 2005
Reputation:
0
update: i tried a build from 5/17 and gapless works fine over 54g for me. only hiccup during playback was when streaming audio and viewing pictures simultaneously.
Posts: 30
Joined: May 2005
Reputation:
0
still on a build from last week (about the 11th at a guess without looking) so apologies if this has been fixed. i just haven't seen it in the log. let me know if it has and i'll test on newest build.
playback of the 54g has been great. been on fairly solid with only minor problems. however, as i said before, i fixed this without a change of build, settings, software etc. just a slight improvement in connectivity through moving the aerial on my server. not a visible change like a jump in link speed from 5m to 54m or anything, just a feeling of improvement because the pauses reduced considerably. for anyone who hasn't read my earlier posts, wireless from win xp machines to this same server has always been fine.
although almost perfect, i'm still getting the crashes. it's basically if the xbox pauses whilst doing any network activity, presumably due to dropped packet, or latency, it is almost 99% going to hang if you try to do something else.
meanwhile the machine providing the data is running. i'm not saying flawlessly, because i can't check that latency or how many packets are lost/dropped etc. but as far as i can see it is as i have a rd connection to it (again over the same wireless adapter that the xbox has to communicate with) and i see no drop in connectivity, no pause in the network monitors etc. rd is showing no visible sign that the machine has paused, blipped, lost wireless connection or anything.
i'm guessing that the problems are along the lines of:
1. the xbox is making some assumptions about the quality of the connection.
2. after a loss in the above quality, the xbox is not "timing out" and retrying the lost packet again. it seems to be waiting indefinitely for the reply as it thinks it should be there.
i'll do some further tests on cables, but has anyone tried pulling the cable from the serving machine during playback and seeing what happens? i find the easiest way to test is to just select another track for playback. over my 54g it will 99% likely hang. i will post what i find as soon as i can test.
Posts: 30
Joined: May 2005
Reputation:
0
ok, done this test with wires. totally reproducable. will log as a new issue.
Posts: 152
Joined: Oct 2003
Reputation:
0
i don't know if this is directly related, but it does cause a crash. when fast forwarding on a song from a smb share, it will lock up after about 20 seconds of the song. it actually doesn't lock up, playback stops, the gui is still running (scrolling text), but as soon as you stop the song or try to change songs, it locks up. it doesn't do this from an xbms share. i'm thinking this may have to do with it missing it's expected data, as the ff from smb is horrible compared to from xbms.
Posts: 3,909
Joined: Dec 2004
Reputation:
20
Nuka1195
Skilled Python Coder
Posts: 3,909
daap: may have similar issues then.
Posts: 26,215
Joined: Oct 2003
Reputation:
187
daap issues are fixed in paplayer v2.
you may be able to fix them in paplayer v1 by changing from a cfilereader to a cfile in mp3codec.h (m_filepap). not sure - so many changes have been made.
hopefully it won't be too much longer now for paplayer v2.
cheers,
jonathan
Posts: 26,215
Joined: Oct 2003
Reputation:
187
replaygain should work just fine with xbms, assuming you have the new ape tag loader. if it reads the tags, it will read the replaygain data.
i'll setup an xbms share myself when i have a chance and check it's all working in v2.
cheers,
jonathan
)
