Recent changes to feature-requests

crossfade would be so nice

Benjamin Carter — Fri, 12 Sep 2008 03:19:15 -0000

It's been a while since a feature request... Don't know if anyone is working on this project or not, but I use it all the time. Great piece of software. Thanks!

If you do release a new version, I would love to be able to crossfade the joints in an mpgjoin operation, such that the volume remains constant, with an option to specify the fade duration.

Thanks again,
BC

Windows build

Jamalu — Mon, 11 Sep 2006 20:21:50 -0000

Dear MpgTx,

I have written a simple UI for people to use mpgtx on
windows. http://movica.sourceforge.net However the
version of mpgtx that is have very old. I got it from
the discussion board. I am sure you have fixed a great
deal of stuff since then.

A user has written in with the following comments

1) A possible performance issue: On the first MPG I
tried, the output was
off (the cutpoints on the output file were too early
by varying amounts up
to a second or more). But then I tried a different
MPG file, and it was
perfect on that one. I'm sure the problem with the
first one was the
fault of the mpg file, but do you have any opinion on
why, or how to get
around it? For that first "bad" file, I then tried
using tMPGEnc to edit,
and it worked fine, so I'm not sure what the issue was
with Movica on that
particular file. Not sure if this is really a bug,
but I'm just letting
you know about it.

I am a novice programmer and would need your help in
compiling mpgtx for the x86 32-bit.

Regards,
rbhargav

support splits relative to end

Anonymous — Tue, 21 Mar 2006 13:51:57 -0000

hello, hope this isn't an rtfm i missed, but:

it would be good to be able to split files relative to their end. for example, to split from 15 seconds into the file, to 35 seconds before the end (rather than having to figure out absolute timeline values for the end). this would allow better automation of clipping where there is stuff to be removed from the end of the file.

cheers...

support for files with mangled time

Robert Forsman — Wed, 26 Jan 2005 20:22:00 -0000

Some input files (e.g. Hauppauge PVR 250) have
discontinuous time fields, where the PTS jumps, or the
GOP timecodes don't align.

If mpgtx can not be enhanced to robustly handle input
with bad time, its documentation should recommend
command-line tools that CAN repair the time fields so
that a repaired stream can be piped into mpgtx.

I have a 2-hour, 3.9G file which has at least one PTS
discontinuity in it (the PTS increments by 348342 in
one spot). I think this corresponds to a point in time
around which mpgtx cuts result in thoroughly psychotic
PTS fields in the output:

36036
66234
69238
81250
75244
78246
90258
84252
87256
99268

Now if they were DTS fields, I would expect them to
jump around a little due to frame reordering, but PTS
fields should be steadily increasing.

VCD support ---Won't fix.---

Anonymous — Mon, 03 Jan 2005 03:54:36 -0000

I'd like MPEGX to have VCD support.

matrix_coefficients info

Wilbert Dijkhof — Sun, 29 Aug 2004 12:16:59 -0000

It would be nice if

mpgtx -i thefile.mpg

gives the "matrix_coefficients" and the
"display_horizontal_size" (both in the
sequence_display_extension) if present. The former
tells you how to decode your stream.

The thing is that there are multiple choises: FCC's
coefficients for red, green, and blue (0.3, 0.59,
0.11), the ITU's default MPEG-2 coefficients (0.2125,
0.7154, 0.0721), or the more common MPEG-1 coefficients
(0.299, 0.587, 0.114).

Most encoders use the MPEG-1 coefficients, which is
wrong for a dvd rip. It would be nice that have that info.

Wilbert

DLL of project

Alexei Shkolnik — Thu, 22 Apr 2004 10:15:10 -0000

I need to convert this project to DLL API.

Can't process Audio.TS files (core dump)

Martin Johnson — Sun, 23 Nov 2003 17:25:34 -0000

Should mpgtx be able to process Audio TS files?

For example, I tried running mpgtx -i and -d with
audio files recorded as TS files by DVB satellite
receiver "Dreambox". One example of a recordings of
BBC Radio 4 (Astra 28E), is shown below.

I have attached the first 32 KB of the Radio 4 file. I
believe this is a valid TS file because "X081" can
demux it (see below). The .MP2 file after X081
contains about 2 seconds of speech by BBC presenter Sue
Lawley.

I was hoping that "mpgtx" would work better than
"X081", which is a larger Java program. In particular
there are some streams which X081 cannot demux, where I
was hoping for better luck with mpgtx.

FWIW the PC is running FreeBSD 4.9 using the port in
/usr/ports/multimedia/mpgtx . Tried with gcc2.95 and
gcc3.3.

Many thanks!

Best regards,

- Martin Johnson.

-------------------------------------

toybox# mpgtx -v
mpgtx Version 1.3
Copyleft 2001 Laurent Alacoque
<laureck@users.sourceforge.net>
Partial work 2002 by Philipp Biermann
<mpegcut@biermann.org>
updates, bugs, patch, money :
http://mpgtx.sourceforge.net

toybox# mpginfo BBC_R4_Archers.ts
Skipped -2 zeroes at start of file
mpgtx: AT EOF - please stop me!
mmm, this file does not start with a pack, offset: -2
use the desperate_mode switch as the first option -X to
search for a header in the whole file!
if you want to force the operation. May yield to an
endless loop if no valid header is found!
Does not even begin with a 00 00 01 xx sequence!
Warning, prog num 6854 redefined !
Warning, prog num 6854 redefined !
Warning, prog num 6854 redefined !
....
Segmentation fault (core dumped)

--------------------------------

Output of X081;

=> File 0: /usr/root/BBC_R4_Archers_TRIMMED.ts (30720
bytes)
=> File is DVB/MPEG-2 TS (Video/Audio/TTX PES)
=> demux
-> no PMT found (at a short scan)
--> PID 0x1F (SIT) (0x0 #1) -> ignored
ok> PID 0x940 has PES-ID 0xC0 (MPEG Audio) (0x11A0 #25)
packs: 163 100% 30832
--> MPEG Audio (0xC0) on PID 0x940
Audio PTS: first packet 00:35:36.264, last packet
00:35:37.032
-> adjusting audio at its own timeline
=> src_audio: MPEG-1,Layer2,48000Hz,stereo,192kbps,CRC
@ 00:00:00.000
audio frames: wri/pre/skip/ins/add 40/0/0/0/0 @
00:00:00.960 done..
-> 5 stereo/jointstereo change(s) detected...
===> new File:
/mnt/dreambox/movie/MP2/BBC_R4_Archers_TRIMMED.mp2
?
summary of created media files:
Audio 0: 40 Frames 00:00:00.960 0/0/0/0
/mnt/dreambox/movie/MP2/BBC_R4_Archers_TRIMMED.mp2
=> 23040 bytes written...

Demux of an open file

Anonymous — Mon, 10 Nov 2003 16:52:17 -0000

Hello Laureck,

I've tried to demux an open file, that is to say, to demux
a file in progress of writing by an encoder, encoding an
video signal and storing the result of that encoding in
that given file.
The result is that mpgtx stops its demux when it
reaches the end of data in the open file. As mpgtx is
much more rapid than the writing of data by the
encoder in the file (demux on a P4), it reaches the end
of data while the encoder is still writing in the file. The
files resulting from the demux do not then contain the
entire content of the original file.
It would the be great if, reaching the end of data in an
open file, mpgtx could "follow" the writing of data in that
file, processing the data by blocks, as soon they are
available, ensuring that the resulting files would contain
the full cotent of the original file.
Furthermore, it would also be great if it would be
possible to discart one stream (video or audio) in the
demux commands, to avoid the creation of unnecessary
files.

Thank you for that smart software, thank you for your
follwup,

best regards

Thierry PIETTE
TVI s.a.
1, Avenue Ariane
B-1201 Brussels
tel: +32 (0)2 778 68 73
fax: +32 (0)2 778 68 09
Mobile: +32 (0)475 64 60 61
Email: tpiette@rtl.be

Need to generate valid MPEG header

Anonymous — Sat, 27 Sep 2003 07:55:34 -0000

I think this is why windows programs can't read the
joined mpegs.

It appears windows media player and burning software
only look at the first header. If several mpegs are
combined, then all of the mpeg headers need to be
removed in the final output and a single header put at
the first of the file indicating the correct length of
the mpeg.