|Milestone:unknownComponent:mythtv
|Version:headSeverity:medium
|
Resolution:-+--
Comment (by Mark Spieth):added 14 which fixes- ac3 audio
amplitude levels are now invariant (encode/decode producessame
levels)- only allow 2
yes this is a bug which I havent got around to fixing yet.
mark
#1106: SVN8701: Adjust Audio Sync forgets setting after jump
back/forward
--+-
Reporter: [EMAIL PROTECTED] | Owner: ijr
Type: enhancement
ok Ive finally found the channel swap problem in ac3. the encoder and
decoder are not symetric.
dts decoder channels are also not symetric.
what alsa and other 6 channel audio outs require is
L R LS RS C LFE
this is what dts decoder outputs
what AC3/A52 decoder outputs is
(LFE) L C R LS RS
what
When so much of the code within that class is dedicated to ensuring that
the timecode produced is very accurate (with going to the lengths of using
a high resolution timer to estimate the soundcard’s position based upon
elapsed time since the position was previously reported). I don’t under
On Tuesday 17 January 2006 22:52, Andy Foster wrote:
Just wanted to share my experiences and see if anybody was in the same
place as me.
When seeking, video/audio stutter for about 2-3 seconds. After that,
playback is fine. As mentioned in other threads, it seems to be related
to
the osd
distclean and rebuild from scratch. you should know better :)
cheers
mark
I updated to the current SVN yesterday, and very quickly had problems
with the backend not responding to the frontend. It would work for a
while, but after an hour or so, there would be problems.
Looking at the log, I
distclean and rebuild from scratch. you should know better :)
cheers
mark
Except, I did not do a svn update, but a svn co.
I will try a distclean, but I don't expect it to fix the problem.
the other problem I usually have is that I forget to update the frontend
after I did the backend so
-+--
Reporter: Mark Spieth |Owner: ijr
Type: enhancement | Status: reopened
Priority: minor|Milestone: unknown
Component: mythtv | Version: head
Severity: medium | Resolution
fingers.
no mmx support as yet of course.
dont forget to change maxchannels in audio setup too otherwise youll only
get 2
mark
On Sun, 01 Jan 2006 10:15:48 +1100, Mark Spieth wrote:
On Saturday 31 December 2005 14:37, Allan Stirling wrote:
bummer. Maybe then just convert the AC3 to stereo PCM
so your actually encoding *back* to AC3/DTS?
yes
___
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
AM
Subject: Re: [mythtv] [patch] multi channel audio patch 1
On Mon, 02 Jan 2006 23:26:22 +1100, Mark Spieth wrote:
woops spotted a bug already.
attached is another patch.
mark
Compile failure on r8474:
mythcontext.cpp:2276: error: prototype for 'bool
MythContext::SendReceiveStringList
Have you tried using the secondary rewrite code? It's the commented stuff
near the bottom of the file (I just committed some stuff to rearrange it,
it used to be on the very bottom). Uncomment the line that says
mythweb.php?PATH_INFO=/$1 and comment out the one above it.
It's odd, though. I
For kicks I just tried installing the nVidia 8178 driver set. Result
was HORRIBLE stuttering. Went back to the nVidia 7676 driver set, and
the stuttering is 99% gone.
anything after 7676 uses heaps of CPU due to opengl avsync.
I went through the same tests a few days ago.
mark
to change maxchannels in audio setup too otherwise youll only
get 2
mark
On Sun, 01 Jan 2006 10:15:48 +1100, Mark Spieth wrote:
On Saturday 31 December 2005 14:37, Allan Stirling wrote:
bummer. Maybe then just convert the AC3 to stereo PCM for timestretch?
At least you'd get sound...
I dunno, I've
On Saturday 31 December 2005 14:37, Allan Stirling wrote:
Steve Adeff wrote:
So, if ALSA can reencode to AC3, I think it would be a great idea for
the
timestretch code.
I hate to be the bringer of bad news (again), but ALSA can't without a
lot
of glue. JACK can - With an external,
was this stuttering showing face in regular playback? I just updated to
8453
and am noticing 1080i is stuttering on 1080i playback where I never had
issue
before.
I am seeing stuttering at normal playback speeds with the latest svn.
Changing that read size did help, but does not eliminate
daniel
I got a 0 byte recording the other day (tuesday).
recent head but not sure what. within a week.
the recording in questing immediately followed on
from another.
it looks like from the log that the timeout system
was circumvented somehow.
the 2 lines
2005-12-27 20:30:06.888
I suspect this could become very annoying to people who bring up the
timestretch osd and just want to view/modify their existing setting. which
is what I do :)
cheers
mark
(In [8408]) Added the ability to quickly switch between 1.0X and the last
time-stretch speed. By default, this action
strike that. it will still be fine as ADJUSTSTRETCH is the main one being
used.
TOGGLESTRETCH is not used normally.
didnt look at the real changes just the patch. apologies.
mark
I suspect this could become very annoying to people who bring up the
timestretch osd and just want to view/modify
Thank you for taking the time to investigate my patch.
I am happy that your testing revealed no noticeable change in the
behaviour of the 2 algorithms, I hope that this indicates that my patch
is
working correctly; video playback was working correctly for me before I
developed the
greg, your patch to utils.c is not quite
correct.
looks like extradata is freed by many modules but
not set to NULL after the free so your catchall free can correctly work. I would
suggest backing it out until this is fixed (lots of places). I played an avi to
make this happen.
cheers
mark
greg, your patch to utils.c is not quite correct.
looks like extradata is freed by many modules but not set to NULL after
the
free so your catchall free can correctly work. I would suggest backing
it
out until this is fixed (lots of places). I played an avi to make this
happen.
woops forgot the soundtouch mods. here it is
again.
cheers
mark
- Original Message -
From:
Mark Spieth
To: mythtv-dev@mythtv.org
Sent: Sunday, December 18, 2005 11:00
PM
Subject: [mythtv] [patch] multi channel
audio patch 1
here is my interim patch
Im trying to fix what I think is a fault that
doesnt allow lirc events to clear the failed to connect to backend ok
popup.
my initial analysis shows that customEvent doesnt
get any lirc events.
what should I be looking for not being that
familiar with lirc?
or should I just open a ticket
On Thursday 15 Dec 2005 23:10, Stuart Auchterlonie wrote:
Hmmm, thinking out loud here, but it's almost like the data rate
drops too low for the decoder at this point. Would make sense since
a pure black screen would have a very low data rate compared to a
much more colourful screen.
I
Im just working on multi channel audio.
(finally)
what is the max number of channels a stream can
have?
AC3 is 6 (5.1)
DTS? 8?
is 3 possible with existing soundcards/streams?
(L+C+R) probably as eventually there should be a config item to indicate max
channels which may be the channels the
-+
--
Reporter: Mark Spieth |Owner: ijr
Type: defect | Status: closed
Priority: minor|Milestone: 0.19
Component: mythtv | Version: head
Severity: medium | Resolution: fixed
also however reported this can I get a bit of such a stream so I can analyse
it.
thanks
mark
Author: danielk
Date: 2005-12-14 20:17:57 + (Wed, 14 Dec 2005)
New Revision: 8265
Changeset: http://cvs.mythtv.org/trac/changeset/8265
Modified:
Adam Greenbaum wrote:
On Tue, 2005-12-13 at 13:36 +, Stuart Morgan wrote:
On Tuesday 13 Dec 2005 13:26, Adam Egger wrote:
It sounds like it's the commercial flagger causing it. It's the only
thing that looks for black screens or scene changes.
Good theory but I'm not flagging any
1. Seek table weirdness. For the third time in two days the seek table for
a
recording appears to be corrupted - that is the 'seek table' not the 'seek
table table' if you follow! I started to watch two of these recordings
whilst
they were still in progress but otherwise I cannot see any
I thought maybe everyone had given up on it because it doesn't work with
time-stretch. I never watch anything below 1.15 these days, and things
like ST:TNG, SG:SG-1 get the 1.5 treatment. I'm a little unhappy with
the non-working multi-channel audio support in MythTV. I might fix that
in 0.20 if
I have3 more playback issues which I am
looking at (if I dont find more)
1. an avi I have plays audio in sync but pauses
frequently (1 per sec)
2. occasionally in head since about 2 weeks ago, If
I play a recording in progress (not live)
I get the seek table update from the
recorder
#781: RebuildSeekTable doesnt
I'm trying to understand what this bug was, since I've been using
mythcommflag
--rebuild quite a bit lately. Did it prevent mythcommflag from rebuilding
the
seek table? is it something I should bother updating from svn and re-run
mythcommflag for?
I believe
The --rebuild option was originally meant to allow people to build the
seektable for a corrupt nuppel-format .nuv file or when they had to
rebuild the recordedmarkup table because of MySQL corruption. That's
probably why I never noticed the bug because those were the cases I
was testing.
The --rebuild option was originally meant to allow people to build the
seektable for a corrupt nuppel-format .nuv file or when they had to
rebuild the recordedmarkup table because of MySQL corruption. That's
probably why I never noticed the bug because those were the cases I
was testing.
mine was only broken for
http://xxx/mythweb
note no trailing /
fix was int mythweb-htaccess.conf
RewriteCond !%{REQUEST_FILENAME} -d
RewriteRule ^(.*)$ mythweb.php/$1
[QSA,L]
Access to http://localhost/mythweb/ is broken by the rewriting rules in
perhaps someone can set up a CVS repo that will store the debian dirs for
the 3 main projects
mythtv mythplugins and myththemes
it would look like (mine :)
mythtv/debian
mythplugins/debian
myththemes/debian
I used the word CVS as what I would like to do is overlay this tree over the
SVN myth
Author: danielk
Date: 2005-10-25 23:45:02 + (Tue, 25 Oct 2005)
New Revision: 7587
Changeset: http://cvs.mythtv.org/trac/changeset/7587
Modified:
trunk/mythtv/libs/libavcodec/h263.c
trunk/mythtv/libs/libmythtv/avformatdecoder.cpp
Log:
Closes #509, by applying
Changes (by danielk):
I verified that the old slowdown code is still working. It is only active
in LiveTV mode, while the FAST_RESTART code was global. Unfortunately,
the
fast restart code
increase CPU load dramatically so it has been shelved for now.
Marking as invalid since the old
for how far we have to go before end is reached. still thinking about this.
maybe isaac you have some ideas about this.
cheers
mark
---
On Oct 19, 2005, at 3:30 PM, Mark Spieth wrote:
1. The time stretch feature doesn't set the value to 1.00 again when I
watch LiveTV delayed and LiveTV
1. The time stretch feature doesn't set the value to 1.00 again when I
watch LiveTV delayed and LiveTV reaches the end. It was setting to
1.00 when the end of the buffer was reached a few weeks ago, now the
picture/sound are stuttering instead of switching back the time
stretch value to 100%.
probably be better to add an extra menu item which does what you want.
this is designed to bring up the control. IMO this should be non negotiable.
other functionality is probably a good thing.
my 2c
mark
#391: Change TOGGLESTRETCH so it turns Time Stretch on and off.
here it is.
I also removed one local var (now only uses err) which may be extraneous so
pick and choose.
cheers
mark
the main problem is probably that av_read_frame doesnt get any frames
and
thus there is no av stream info.
the test could easily be put into avformatdecoder thus avoiding
the main problem is probably that av_read_frame doesnt get any frames
and
thus there is no av stream info.
the test could easily be put into avformatdecoder thus avoiding changes
to
libav/ffmpeg stuff which will make isaacs merging easier.
if you like Ill generate a patch tomorrow but
here is another patch. the last change affected recording mode so this
eliminates that problem.
for those that are having trouble, please try.
mark
Ive also had livetv problems even with HEAD
so I investigated this morning.
in TVRec::SetupSignalMonitor
there is a comment
// TODO reset the
On Sun, 2005-09-11 at 22:52 +1000, Mark Spieth wrote:
here is another patch. the last change affected recording mode so this
eliminates that problem.
for those that are having trouble, please try.
if (TRANSITION(kState_None, kState_WatchingLiveTV))
{
+if (channel
-int timeout = 20;
+int timeout = 5000;
Bug #2 -- LiveTV should tuning not time out, the user can change the
channel or tuner or input or even exit LiveTV at any time.
Why force one particular decision on the user?
fair enough. the number just looked
#316: Adjust Audio Sync doesn't show current/last value
--+---
--
Reporter: [EMAIL PROTECTED] | Owner: ijr
Type: defect| Status: new
Priority: minor | Milestone:
Component:
remember when I made an auto sync adjust, it stopped xvmc from doing things
correctly.
also now there are 2 audio adjusts. the menu is currently linked to my
version. will need to be consolidated but you probably know that :)
cheers
mark
Author: ijr
Date: 2005-08-30 05:32:41 +
I'm not at home to try it now but I don't remember seeing that audiosync
option. You just mean press M during playback right? I use the menu quite
often for time stretch etc but I don't remember seeing the audio sync
option. It's definitely in 0.18.1?
no its newer than 1.18.1 so you wont
This is almost certainly a pts wrap effect.
This assumes you have a mpeg2 recording which hasnt been transcoded.
When it gets in this mode,
use menu, select audiosync, press 0 or 1 (forget which) to do a fake resync.
or apply this patch and see. If you have transcoded the patch wont do
anything
On Tue, 2005-08-30 at 00:52 +0100, Mark Weaver wrote:
With current SVN (and revisions going back to at least 7133, possibly
before), when seeking live TV after changing channels, I sometimes see
data from previous channels.
Could some one point me at the appropriate place to start
did you try the audiosync menu 1?
what about mplayer?
dsmyth may use different algo. havent checked the code so not sure.
mark
---
I'm quite certain that's what you thought last time too. However I applied
your patch then and it didn't help.
I'm currently running 0.18.1 RPMs as
There around line 787 there is this statement:
ok = wait_for_dvb(channel, 1000, abortRecordingStart);
Can you try increasing that 1000 ms until it works?
Then send me the value that seems to work for you?
According to the specs the timeout should be about 450 ms, but it
appears
Are the services that take a while encryped?
all FTA for me. no CI slot.
Maybe we need add the access controller's startup time to the timeout?
the PMT/PAT stuff is generally not encrypted anyway. only CA streams would
be specified as part of the services which allow decryption of the video
not sure why but livetv is not working for me at the moment with HEAD.
recording and watching the recording works ok (which is what I
normally use) DVBt of course. seems to crash backend though process is
still there but backend wont respond to frontend requests. wierd.
havent traced it
happened again for F1 record. more detail in backend log.
sometimes (quite often now say 3 out of 5 times failed now to record) due to
PMT NOT SET
attached is the detailed backend log with many backend restarts to retrigger
record start.
cheers
mark
mythbackend.log.badrecord.bz2
Description:
not sure why but livetv is not working for me at
the moment with HEAD.
recording and watching the recording works ok
(which is what I normally use)
DVBt of course. seems to crash backend though
process is still there but backend wont respond to frontend requests. wierd.
havent traced it
On Fri, 2005-08-19 at 00:30 -0400, Isaac Richards wrote:
On Thursday 18 August 2005 11:46 pm, Daniel Kristjansson wrote:
Ok, then this flag is right.
Any idea why avlib seeking doesn't work with LiveTV? I get a bunch of
messages about safe_read failing. Perhaps it is because the
: |
+-
--
Comment (by ijr):
Mark Spieth had a patch for this, but I didn't really like how it was
implemented. Might be someplace to start, though.
--
Ticket URL: http://cvs.mythtv.org/trac/ticket/189
MythTV http://www.mythtv.org/
MythTV
Author: danielk
Date: 2005-08-19 03:11:49 + (Fri, 19 Aug 2005)
New Revision: 7076
Changeset: http://cvs.mythtv.org/trac/changeset/7076
Modified:
trunk/mythtv/libs/libmythtv/avformatdecoder.cpp
trunk/mythtv/libs/libmythtv/decoderbase.cpp
Log:
This fixes the
try adding
QMAKE_STRIP = echo
to the settings file.
this way I can get a release build with debug syms.
cheers
mark
- Original Message -
From: Paul [EMAIL PROTECTED]
To: mythtv-dev@mythtv.org
Sent: Sunday, July 24, 2005 6:25 AM
Subject: [mythtv] How do you stop qmake from trying to strip
On Sunday 17 July 2005 09:20 pm, Mark Spieth wrote:
your choice. In that case you better comment out the lsb3full call bit
in
the AvFormatDecoder::DoFastForward function. This will break things as
it
assumes the timestamps have been adjusted to 0 at start.
I think to fix
see my last ticket and patch (today's). it explains everything.
cheers
mark
99% of the time my system works fine. However everynow and then I get a
recording (in this case an MPEG4 transcoded from DVB MPEG2) which seems to
have a glitch or something not quite right. From then on in the
the
transcoder failed (presumably from the same problem)
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Mark Spieth
Sent: Sunday, 17 July 2005 8:18 PM
To: Development of mythtv
Subject: Re: [mythtv] audio waiting for buffer to fill
see my last ticket
Severity: medium | Milestone:
Priority: minor | Version:
Owner: ijr |Reporter: Mark Spieth [EMAIL PROTECTED]
+-
--
Changes (by ijr):
* resolution: = invalid
* status: new = closed
Comment
hi donavan,
should I resubmit the mythvideoupdate patch or is it a bit on the grey side
of legal/acceptable?
I ask this off list since you made the statement and have had no feedback.
cheers
mark
Mark Spieth, PhD
DC Labs Pty Ltd
2 Mavron Street
Ashwood 3147
Australia
ph: +61-3-9807 8600
Mobile
I did an update this morning (10 hours ago) 6752
compiled fine after a distclean.
note that distclean doesnt do a real distclean as the config.mak and
config.h files are not deleted.
I posted a patch a while ago that does this if you are interested.
cheers
mark
Dear Issac, and all,
Did a fresh
just give isaac some time. I know its on his list. he is very busy.
unless you want this isaac of course.
cheers
mark
Bryan Mayland wrote:
Well this works great! Thanks for coding support for this, it has
been on my wishlist for a long time now.
Mark, would you mind creating a
what is this option doing there?
-fno-exceptions
of course it wont work.
doesnt get inserted for me. but then again I use svn HEAD.
cheers
mark
- Original Message -
From: Sage [EMAIL PROTECTED]
To: mythtv-dev@mythtv.org
Sent: Monday, June 27, 2005 7:09 AM
Subject: [mythtv]
I'm pretty sure -fno-exceptions has never been a default
compile option on any version of MythTV. -fno-rtti will
cause problems too.
But it looks like these are Embedded Qt flags.
Do we support embedded qt?
perhaps a
QMAKE_CXXFLAGS -= -fno-exceptions
in settings.pro will do the
What would people think about filing bugs ( via
http://cvs.mythtv.org/trac/newticket ) for patches, instead of mailing the
dev list? It'd be nice to use a better tracking system for patches than
just
the mailing list, so things can get resolved quicker. It might lose a
little
bit of current
not the correct fix. still compare selectedTrack but use ==-1 to check
whether a track has been found not the maxTracks.
I missed the 2nd comparison though so that was a good find.
patch attached for 2nd one otherwise would fail badly if no track is found.
cheers
mark
--- avformatdecoder.cpp 3
I applied this patch and am having mixed results. The duration shows
accurately, but I can rarely seek forward (30s) more than a few times
before I'm kicked back to the recordings screen. Pressing back (5s)
skips forward or back some amount (I can't really tell how much, it's
pretty
hi daniel,
here are some extra patches for compile support of the older dvb driver
headers.
cheers
mark
mythtv_olddvb.patch
Description: Binary data
___
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
inc track.
cheers
mark
On Monday 13 June 2005 09:01 am, Mark Spieth wrote:
I had the problem that one of my audio tracks dropped out during a
program
and selecting the alternate fixed things. however I couldnt do it from
the
menu (xbox).
so here is a patch that lets you do it.
Doesn't
On Monday 13 June 2005 03:42 am, Mark Spieth wrote:
I dont know about anyone else, but sometimes I get a 10 sec read timeout
on
play start.
attached is a patch which fixes this.
It seems to be a timing issue with readaheadthread and the reader and it
gets stuck in the loop
since
sorry you are probably correct.
I will fix this properly tonight or tomorrow.
mark
No, I meant I think that the patch won't work, since there's no handler
for
NEXTAUDIO in TreeMenuSelected().
___
mythtv-dev mailing list
mythtv-dev@mythtv.org
I have just got an avi with a single mono audio
track.
avformatdecoder::autoSelectAudioTrack while loop
never exits so have fixed this.
patch attached
cheers
mark
mythtv_autotrackselect1.patch
Description: Binary data
___
mythtv-dev mailing list
I dont know about anyone else, but sometimes I get a 10 sec read timeout on
play start.
attached is a patch which fixes this.
It seems to be a timing issue with readaheadthread and the reader and it
gets stuck in the loop
since the buffer is too full to continue but the reader isnt reading.
I had the problem that one of my audio tracks dropped out during a program
and selecting the alternate fixed things. however I couldnt do it from the
menu (xbox).
so here is a patch that lets you do it.
cheers
mark
mythtv_menuaudiotracknext.patch
Description: Binary data
attached is a patch which fixes this.
It seems to be a timing issue with readaheadthread and the reader and it
gets stuck in the loop
since the buffer is too full to continue but the reader isnt reading.
You can uncomment the section of code that begins with // HACK
Sometimes the readhead
I found another bug since the ffmpeg update due to
a minor functioanlity change on how seek works.
the call to av_seek_frame should set
AVSEEK_BACKWARDS to find the previous, not next keyframe.
thus patch 8 is required. otherwise its the same as
7.
backwards seek by 5 secs now works ok for
sorry forgot to attach again.
here it is
- Original Message -
From: Mark Spieth
To: mythtv-dev@mythtv.org
Sent: Monday, June 13, 2005 2:14 PM
Subject: [mythtv] [PATCH] generic libavformat seek support (8)
I found another bug since the ffmpeg update due to a minor functioanlity
change
hi all
given the issues with the auto audio sync timecode
adjust with XMVC and with ffrew, this patch will remove it.
also included in this patch is 2 extra controls in
audiosync adjust mode
0 - reset audiotimecodeoffset to 0
1 - set audiotimecodeoffset to the same as the
current video
the uint pre cast needs to change to uint32 or whatever is a 32 bit int.
this then will get cast to a pointer (32 bit)
Im not sure what that will be. perhaps int32_t or Int32 or Uint32 or
something like it.
cheers
mark
___
mythtv-dev mailing list
I have a rather clever audio
processor which adds some delay to the audio and need to usually hack
the code a little to account for the offset (I didnt do it as cleanly as
you did)
to see how to handle algorithmic delay for audio, see audiooutputbase.cpp
AudioOutputBase::SetAudiotime(void)
your version is great.
I suggest people use this one with CVS.
cheers
mark
Here's my version of the patch
You'll probably notice that this isn't a cvs diff
it's a patch against a cvs export -D 2005-05-30, but from the commits
list there aren't any changes to that file since.
much appreciated.
Thanks
Neale.
Mark Spieth wrote:
Hi Mark,
Did you edit this patch? it was coming up as malformed when I tried to
apply it. I added a blank line at line 54 to 'fix' it, so that it'd
apply.
Neale.
yes I did. sorry about this. should have checked its integrity
Hi Mark,
Did you edit this patch? it was coming up as malformed when I tried to
apply it. I added a blank line at line 54 to 'fix' it, so that it'd apply.
Neale.
yes I did. sorry about this. should have checked its integrity. have now.
had to remove the irrelevant bits from other things Im
Im in favour of this patch. should have done it myself.
mark
The attached patch adds support for incrementing/decrementing
timestretch by 5 units (i.e. 0.25x) using the UP/DOWN keys,
respectively, while in the adjust timestretch mode. Before the patch,
the UP/DOWN keys caused MythTV to
heres my updated patch for the new ffmpeg.
found an extra bug in the url_fsize function. basically it couldnt work
before. had to change the interface to url_seek so that it returns an
int64_t as current pos. before it returned an int and 0 always which is a
bit useless =) as a result a couple
this is a small patch which allows the adjustment of audio sync to match the
video.
only menu access at this time. I didnt think a key would be necessary as its
not needed in most cases.
it stays up for 10 seconds so you have a chance to adjust it while watching
the lip movements.
the display
:
--
--
Changes committed by bjm on Sat May 28 00:50:19 2005
Modified Files:
in mythtv/libs/libmyth:
uitypes.cpp
in mythtv/libs/libmythtv:
nuppeldecoder.cpp
Log Message:
Some small patches from Mark Spieth
attached is a better hack.
see what you think
mark
On Sun, 2005-05-22 at 08:29 +1000, Mark Spieth wrote:
this fixes the distclean problem removing mythconfig.h and
mythconfig.mak
no but cant figure out a better way of doing it for a distclean.
these files do need to be cleaned
here is what I hope is the last cut of this
patch.
works seek exact for both avi and mpeg files (the
ones I have anyway)
to overcome the keyframe seek issue with mpeg which
doesnt support it inside avf I have set it to seek behind by maxkeyframedist and
then use myth getframe gobble to
Actually, uhm. I just started doing a resync with ffmpeg CVS a bit
earlier
tonight. Not sure exactly when I'll have it in our CVS tree, but, once
I've
got that finished, would you mind verifying that your patch still works
before I apply it?
of course.
those fixes should be sent to them as
I'm not to merging libavformat yet, but it looks like that code may have
changed already..
Ive just checked ffmpeg cvs. looks like avidec.c has changed quite a bit.
wont be a simple merge for me.
utils.c will be the same though.
Ill leave it until you get these in. it will give me a rest on
this fixes the distclean problem removing
mythconfig.h and mythconfig.mak
otherwise a fresh configure doesnt work
correctly.
cheers
mark
mythtv-config_clean.patch
Description: Binary data
___
mythtv-dev mailing list
mythtv-dev@mythtv.org
I just tried Mark Spieth's generic libavformat seek patch (5) but it
didn't make any difference.
Oh well, the patch shows me where to start looking for the problem. I
can post a small sample .nuv if anyone wants to take a look :-)
still looking at this. patience ;-)
The blockiness
1 - 100 of 148 matches
Mail list logo