[vdr] Weird frame rate issue (VDR 1.6 recording, VDR 1.7 playback)

2012-01-23 Thread Tobi
Hi!

Several recordings recorded from JSC Sports (S13.0E-318-500-8440) seem to
have a weird framerate. The recordings were done in PES format with VDR
1.6, so there's no framerate setting in the info file and VDR will use 25
fps by default. This leads to a wrong recording length display (30 minutes
for a recording that was actually 20 minutes long). VDR 1.7.22 is used for
playback and with a small patch I can provide the framerate via the
info.vdr (Which might be a useful feature anyways, patch attached.).

So far so good, but the recording seems to have a weird framerate

This is the recording:

501593483 2012-01-20 13:15:00.0 +0100 001.vdr
   366032 2012-01-20 13:15:00.0 +0100 index.vdr
  149 2012-01-20 12:55:05.0 +0100 info.vdr

The mtime diff between info.vdr and 001.vdr is 19:55 minutes. index.vdr is
366032 bytes in size, which divided by 8 gives a framerate of 38,29 fps.

Can anyone explain this framerate?

ffmpeg and mediainfo both say 25fps (see below) and VLC tells me it's 50fps.

When I convert the recording to TS using project-x, set the framerate in
the info.vdr to 25 and reindex it with VDR 1.7.22, it still shows the
wrong length.

bye,

Tobias


ffmpeg says:

Input #0, mpeg, from '001.vdr':
  Duration: 00:19:55.77, start: 59497.383833, bitrate: 3355 kb/s
Stream #0.0[0x1e0]: Video: mpeg2video (Main), yuv420p, 720x576 [PAR
64:45 DAR 16:9], 3049 kb/s, 25 fps, 25 tbr, 90k tbn, 50 tbc

and mediainfo says:

Complete name: 001.vdr
Format   : MPEG-PS
Duration : 19mn 55s

Video
Format   : MPEG Video
Format version   : Version 2
Duration : 19mn 55s
Width: 720 pixels
Height   : 576 pixels
Display aspect ratio : 16:9
Frame rate   : 25.000 fps
Standard : PAL

#! /bin/sh /usr/share/dpatch/dpatch-run
## xeatre-manualframerate.dpatch by Tobias Grimm 
##
## All lines beginning with `## DP:' are a description of the patch.
## DP: Use frame rate from info file even for PES recordings

@DPATCH@
diff -urNad '--exclude=CVS' '--exclude=.svn' '--exclude=.git' '--exclude=.arch' 
'--exclude=.hg' '--exclude=_darcs' '--exclude=.bzr' vdr~/recording.c 
vdr/recording.c
--- vdr~/recording.c2011-10-25 22:04:26.0 +0200
+++ vdr/recording.c 2012-01-23 23:25:26.0 +0100
@@ -719,8 +719,8 @@
 else if (!isPesRecording) {
priority = info->priority;
lifetime = info->lifetime;
-   framesPerSecond = info->framesPerSecond;
}
+framesPerSecond = info->framesPerSecond;
 fclose(f);
 }
  else if (errno != ENOENT)
___
vdr mailing list
vdr@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/vdr


Re: [vdr] [ANNOUNCE] VDR developer version 1.7.23

2012-01-23 Thread VDR User
On Mon, Jan 23, 2012 at 9:46 AM, Halim Sahin  wrote:
> Hi,
> The media_build repo contains only backported drivers.
> Are you sure that the new api changes were backported?

The media_build repo is a daily snapshot of what you'd get with
media_tree (minus the kernel source itself).

___
vdr mailing list
vdr@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/vdr


Re: [vdr] [ANNOUNCE] VDR developer version 1.7.23

2012-01-23 Thread Halim Sahin
Hi,
The media_build repo contains only backported drivers.
Are you sure that the new api changes were backported?

The mainstream development happens in media_tree only, which contains a
complete kernel snapshot.
HTH.
Halim


___
vdr mailing list
vdr@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/vdr


Re: [vdr] [ANNOUNCE] VDR developer version 1.7.23

2012-01-23 Thread Lars Hanisch

Hi,

Am 23.01.2012 05:43, schrieb Hawes, Mark:

I have updated my current build taken from the Media Build Tree last Tuesday 
with the contents of the linux media tarball dated 22/01/2012 and rebuilt my 
drivers.
I still get the same results.
As the system initialises the following lines appear in the syslog:
Jan 23 12:16:44 Nutrigrain kernel: [9.338720] tuner-simple 16-0061: 
couldn't set type to 63. Using 78 (Philips FMD1216MEX MK3 Hybrid Tuner) instead
Jan 23 12:16:44 Nutrigrain kernel: [9.346240] DVB: registering adapter 1 
frontend 0 (Conexant CX24116/CX24118)...
Jan 23 12:16:44 Nutrigrain kernel: [9.349110] DVB: registering adapter 1 
frontend 1 (Conexant CX22702 DVB-T)...
Subsequently when starting VDR the following is logged:
Jan 23 13:10:13 Nutrigrain vdr: [2704] registered source parameters for 'A - 
ATSC'
Jan 23 13:10:13 Nutrigrain vdr: [2704] registered source parameters for 'C - 
DVB-C'
Jan 23 13:10:13 Nutrigrain vdr: [2704] registered source parameters for 'S - 
DVB-S'
Jan 23 13:10:13 Nutrigrain vdr: [2704] registered source parameters for 'T - 
DVB-T'
Jan 23 13:10:13 Nutrigrain vdr: [2704] probing /dev/dvb/adapter0/frontend0
Jan 23 13:10:13 Nutrigrain vdr: [2704] new device number 1
Jan 23 13:10:13 Nutrigrain vdr: [2704] frontend 0/0 provides DVB-S with QPSK ("ST 
STV0299 DVB-S")
Jan 23 13:10:13 Nutrigrain vdr: [2708] tuner on frontend 0/0 thread started 
(pid=2704, tid=2708)
Jan 23 13:10:13 Nutrigrain vdr: [2709] section handler thread started 
(pid=2704, tid=2709)
Jan 23 13:10:13 Nutrigrain vdr: [2704] probing /dev/dvb/adapter1/frontend0
Jan 23 13:10:13 Nutrigrain vdr: [2704] new device number 2
Jan 23 13:10:14 Nutrigrain vdr: [2706] video directory scanner thread ended 
(pid=2704, tid=2706)
Jan 23 13:10:14 Nutrigrain vdr: [2705] video directory scanner thread ended 
(pid=2704, tid=2705)
Jan 23 13:10:19 Nutrigrain vdr: [2704] frontend 1/0 provides DVB-S,DVB-S2 with QPSK 
("Conexant CX24116/CX24118")
Jan 23 13:10:19 Nutrigrain vdr: [2712] tuner on frontend 1/0 thread started 
(pid=2704, tid=2712)
Jan 23 13:10:19 Nutrigrain vdr: [2713] section handler thread started 
(pid=2704, tid=2713)
Jan 23 13:10:24 Nutrigrain vdr: [2704] ERROR (dvbdevice.c,1087): 
/dev/dvb/adapter1/frontend1: Device or resource busy
Jan 23 13:10:24 Nutrigrain vdr: [2704] found 2 DVB devices
I'm sure that I have the latest drivers loaded now, but still the same issue.

> The only conclusion that I can come to is that the necessary changes to the 
drivers
> have not (yet) been made.  Has anyone else tried VDR 1.7.23 with a HVR 4000 
hybrid card
> and if so, do you get the same results?

 Yes, I would get to the same conclusion. There should only be one frontend.
 You should point the guys on the linux-media mailing list to this "bug".

Lars.

___
vdr mailing list
vdr@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/vdr