rbouqueau / gpac Goto Github PK

Reported by moissinac on 2005-03-18 10:56 UTC
I'm using mp4box with the following command line
mp4box -inctx rect.bt -outctx rectUpdated.bt -mp4
rect-update.bt

The chunk encoding work fine, but the output of the
updated context doesn't work.
It crashes in the ReorderAUContext function called by
M4SM_DumpToText on
assert(au->owner->timeScale);

Thank's in advance for your help.

Jean-Claude Moissinac

Reported by b-o-n-d on 2005-04-06 22:29 UTC
i tried muxing the following xvid .avi into .mp4 with
mp4box:
http://www.8ung.at/bond/bdayvideo\_cut.avi

it signals simple profile at level3 but uses b-frames,
when playing back on dshow with the divx5, xvid and
ffdshow (xvid) decoder i get heavily jerky playback
(only when i play with ffdshow (libav) it works)

the file has been created with xvid 1.1 beta3 by using
the dxn portable profile, but it seems to also be there
with the beta2 i often tested (tough never with the dxn
portable profile set)

when muxing the same file with mp4creator everything
plays fine

i compared the dts/cts of mp4creator and mp4box and
they seem to be exactly the same

i assume that maybe this SP@L3 with b-frames
combination somehow messes mp4box up?

btw it doesnt make a difference whether i use -convert
or -add

Reported by b-o-n-d on 2005-03-13 11:46 UTC
i noticed that when importing raw video streams with
30fps that the final framerate isnt really 30fps, but
something like 30.006523 fps...

i assume thats some problem related to the timescale?

with 25fps it works fine

*Reported by anonymous on 2005-07-13 08:41 UTC
mp4box -isma, which supposedly produces isma 1.0 compatible
files, accepts mp3 audio without deleting the track or giving an error
message. instead, it "isma-izes".
afaik, isma 1.0 only covers aac audio.

Reported by b-o-n-d on 2005-05-16 21:17 UTC
ok i have encoded an adts .aac stream with faac, which
seems to be mpeg-2 aac-lc (this also happened with
another file produced by nero)

1. when importing with mp4box to .mp4 i get the
   following with -info:
   MPEG-4 Audio AAC Main and
   Audio PL: Not part of MPEG-4 audio profiles (0xfe)
   (note too that 4 is displayed, not mpeg-2)

mp4info from mpeg4ip tells me its MPEG-2 AAC LC,
which should be the correct one

faad2 tells me too that its aac main

2. on the other side when importing the same stream
   with mp4creator from mpeg4ip, -info tells me:
   MPEG-4 Audio AAC LC and
   Audio PL: High Quality Audio Profile @ Level 2 (0x0f)
   so it says 4 and not mpeg-2

mp4info says MPEG-2 AAC LC again

faad2 says its lc-aac (doesnt mention if mpeg-2 or mpeg-
4)

i have uploaded the sample here:
http://www.8ung.at/bond/faac.aac

Reported by b-o-n-d on 2005-01-30 14:09 UTC
i tried importing mpeg-1 and mpeg-2 video streams
into .mp4 with mp4box (latest official release)

and the bitrate in both cases gets always signalled as "0
kbps"

both streams where encoded with tmpgenc

Reported by b-o-n-d on 2005-01-12 13:02 UTC
as described here:
http://sourceforge.net/forum/forum.php?
thread_id=1200631&forum_id=287547

Reported by doom9 on 2005-06-18 16:26 UTC
If language assignment on import is used for an audio
track, the language of the video track is also
automatically assigned, and defaults to the language of
the first track.

For instance, the following commandline yields English
video, even though nothing was assigned:

"mp4box.exe" -add "D:\DVDs\DVDVolume\VIDEO_TS\xvid.m4v"
-add
"D:\DVDs\DVDVolume\VIDEO_TS\trailer-audio.mp4";lang=eng
-add "D:\DVDs\DVDVolume\VIDEO_TS\Copy of
english.srt";lang=eng -fps 25 -new
"D:\DVDs\DVDVolume\VIDEO_TS\xvid-muxed.mp4"

Then if I try to override the video language

"mp4box.exe" -add
"D:\DVDs\DVDVolume\VIDEO_TS\xvid.m4v";lang=fre -add
"D:\DVDs\DVDVolume\VIDEO_TS\trailer-audio.mp4";lang=eng
-add "D:\DVDs\DVDVolume\VIDEO_TS\Copy of
english.srt";lang=eng -fps 25 -new
"D:\DVDs\DVDVolume\VIDEO_TS\xvid-muxed.mp4"

the audio remains English as well.

Reported by b-o-n-d on 2004-04-06 20:50 UTC

basically there are two sorts of not coded frames (n-
vops)

1. the ones caused by packed bitstream
2. the ones caused because the encoder decided to not
   code a frame as it was the same/similar to the frame
   before

both types are technically the same (no way to
differentiate them by only looking at the n-vops
themselves)

ad 2) xvid offers a feature called "frame drop ratio",
which creates n-vops if two (or more, depending on the
setting used) frames are the same. it then simply drops
the second frame, puts a n-vop instead (and tells the
decoder to display the same frame again)

now the problem:
if the type2 n-vops get dropped (as mp4box does atm)
the video will get out of sync as the dropped frames are
simply missing (there will be less frames than needed in
the video stream)
the solution would be to

• not drop them or
• create a variable framerate mp4 (as 3ivx tries) or
• simply dont care

of course this causes heavy problems with the type1 n-
vops, caused by packed bitstream, which have to be
removed in any way (without the need for vfr)

ok now my solution suggestion is:
simply continue handling n-vops as you are doing now! i
submitted this report so you know that this could be a
potential problem (if someone reports missing frames)

of course you could try to differentiate between type1
and type2 n-vops by analysing the surrounding frames
(like the 3ivx guys try) but it maybe will never work
100% correctly
the point is: as good as noone uses the xvid frame drop
ratio option anyways, so to say there will hardly ever be
any type2 n-vops anyways!!!
i for myself prefer 99 out of 100 absolutely 100% correct
muxings, than something like 60:40 or 70:30

*Reported by anonymous on 2005-01-19 13:45 UTC
When the option "-convert" and "-3gp" are used together
the output file is a 3GP file as expected.
But when the option "-add" and "-3gp" are used together
the output file is a MP4 file. So it seems that the
option "-3gp" is ignored when using "-add". At least
the file brand is not set correctly.

Reported by b-o-n-d on 2005-01-12 13:05 UTC
when merging multichannel aac .mp4 audio only files with
video .mp4 files mp4box reports "2 channel audio"

still mp4ui tells me that there are 6 channels in the
resulting .mp4 files
also the players seem to detect the 6ch correctly

does mp4box handle multichannel (more than 2
channels) files correctly?

Reported by b-o-n-d on 2005-06-13 20:00 UTC
i actually seem to have found a bug in the new -delay
option:
when the audio stream for example has the id 101,
setting the delay while refering to 101 doesnt work
(meaning no delay gets set)
when setting the same delay on the other hand on the
same input file but refering to id 1 (which is the id of the
bifs stream in that file) it works

when doing the same on a .mp4 file which has the audio
with id 1 it works fine as it should

Reported by r_b_f on 2005-04-25 11:55 UTC
mp4box Incorrectly converts cyrillic Windows-1251 *.srt
files to UTF-8.

When I make import tx3g from 3gp file:
mp4box -add rus.3gp#1 rus.mp4

• all OK.

When I convert from *.srt:
mp4box -add rus.srt rus.mp4

• the text is not readable.

*Reported by anonymous on 2005-04-07 17:56 UTC
Does the latest mp4box.exe (0.2.5)
http://www.aziendeassociate.it/cd.asp?dir=/gpac/dev

require a new MP4BoxGUI ?

mp4boxgui 0.14 freezes with the new build.

nexus3

Reported by b-o-n-d on 2005-08-04 11:11 UTC
ok i now tested the new 0.4.0 and the only thing i
noticed were a bunch of problems with timed text:

1. when importing from .srt mp4box seems to want to
   set a totally wierd boxsize for the sub stream,
   something like 4656345232x143534645754

2. when extracting ttxt from .mp4 to .srt or.ttxt i noted
   that some special characters like the german umlaute
   are not correctly shown
   also when i decode such a stream with ffdshow i also
   get the wierd characters

maybe this is interesting and fixing this issue:
http://forum.doom9.org/showthread.php?
p=677432#post677432
(some comment on this patch:
http://forum.doom9.org/showthread.php?
p=686828#post686828)

also things like "we're" is shown as "we're" when
doing the srt->ttxt conversion

Reported by b-o-n-d on 2005-03-18 12:47 UTC
i noticed that on a compile made from cvs on 11 march
2005 i get a crash when i want to import a .ttxt text file
to .mp4

on older builds, compiled by the same person, this
doesnt happen (newest one tested was from 26 january
2005)

Reported by b-o-n-d on 2005-04-04 10:27 UTC
i have now been able to calculate the cts/dts values of
the .mp4 files created with mp4box
i first calculated the dts/cts from .mp4 files produced by
nero and moonlight and than remuxed the avc stream
to .mp4 with mp4box and mp4creator and checked
again the dts/cts

i found the following problems:

1. independant of the used number of max. cons. b-
   frames, mp4box sometimes uses the same cts value
   twice, which is obviously not correct, and also not there
   in the original source .mp4 and also not in the
   mp4creator .mp4
   i could reproduce this on all three samples i tested.
   apart from that the nero samples didnt show another
   problem

2. i tested one moonlight sample i have and i, also
   additionally to 1), found another issue which was that
   mp4box seemed to place some cts values to the wrong
   dts (i assume mp4box is wrong as neither the original
   moonlight nor the remuxed mp4creator (which are
   identical btw) show these combinations)

for moonlight/mp4creator:
dts 2064000, cts 2064000
dts 2080000, cts 2096000
dts 2096000, cts 2128000
dts 2112000, cts 2112000 <- !
dts 2128000, cts 2160000

for mp4box:
dts 2064000, cts 2064000
dts 2080000, cts 2112000 <- !
dts 2096000, cts 2096000
dts 2112000, cts 2128000
dts 2128000, cts 2160000

this moved lines dont happen regularily

3. in the last line of the moonlight sample i also noticed
   that mp4box used a totally different value for the cts
   than moonlight/mp4creator (which also used the same),
   but not simply a value twice as described under 1):

moonlight/mp4creator:
dts 35264000, cts 35280000

mp4box:
dts 35264000, cts 35296000

i have uploaded some nero samples, showing 1), here:
http://www.8ung.at/bond/avc\_ctts.7z

i havent tested arbitrary frame coding orders, but maybe
this is also somehow related to this issue reported by
algern0n:
http://sourceforge.net/tracker/index.php?
func=detail&aid=1162106&group_id=84101&atid=571741

Reported by amissemer on 2004-06-18 10:59 UTC
GDIPlus 2D Raster doesn't display IndexedLineSet2D when
it contains only 2 points.
The problem is the same when there are 3 aligned points.

Using GPAC 2D Raster solves the problem.

The attached file contains a sample xmt producing a
blank presentation with GDIPlus 2D Raster (with windows
XP), whereas with GPAC Raster there are some black
lines.

Reported by b-o-n-d on 2005-03-13 11:50 UTC
i tried to play muxed avc .mp4 files with the 100%
different nero and haali .mp4 directshow parsers and
both show the same: a crash in libavcodec.dll when
trying to use ffdshow as decoder

the same avc streams muxed into .mp4 with mp4creator
play fine

also i have to say that the resulting mp4box .mp4 files
play fine with mplayer, still as both nero and haali's
parser show the same i think its a mp4box related
problem!?

get haali's parser here:
http://haali.cs.msu.ru/mkv/
ffdshow:
http://ffdshow.sourceforge.net/tikiwiki/tiki-index.php?
page=Getting+ffdshow
nero:
http://www.nero.com/de/nero-up.php

*Reported by anonymous on 2005-04-10 05:58 UTC
Salut jeanlf,

for some reason mp4box limits the resolution of
mp4/3gp outputs - for raw H263 - to 128x96.
In my opinion this should be loosened.
Afaik the most common (?allowed?) resolutions for H263
are:
SQCIF (128x96), QCIF (176x144), CIF (352x288), 4CIF
(704x576)

for mp4: all res and fps should be allowed

for 3gp: at least 128x96 & 176x144 (maybe CIF in the
future) should be permitted with fps max. = 30

I created my H263 avis with ffdshow vfw, extracted to
raw with avi2raw.exe.
(Maybe you could add h263-avi support in the future)
The Audio (AMR) was created with ffmpeg.
MP4Box - GPAC version 0.2.5-DEV (05.04.2005)

DShow fiters:

Haalis MP4/3gp splitter:
http://haali.cs.msu.ru/mkv/

ffdshow decoders for H263 and AMR-NarrowBand:
http://ffdshow.sourceforge.net/tikiwiki/tiki-
view_articles.php

nero's audio decoder filter for AMR-WideBand

Another observation:

muxing two amr's (nb+wb) into 3gp:
"Setting major brand to 3GPP Generic file"

before that (or 1 audio only) its:
"Setting major brand to 3GPP V5 file"

Is this intended ?

regards
nexus3

Reported by superdump on 2005-07-28 23:08 UTC
For over a month now (as an estimate, I'm not sure
precisely when) I have been unable to compile the gpac
CVS. I have also just tried to compile the 0.4.0
release and I get the same error.

I have MinGW, MSYS and MSYS DTK (and some other
additional binaries which were necessary for other
things) installed. The error given is as follows:

fileimport.o(.text+0x21ec):fileimport.c: undefined
reference to `gf_isom_get_sample_duration'
collect2: ld returned 1 exit status
make[2]: *** [MP4Box.exe] Error 1

The configuration is as follows:

$ ./configure
./configure: sdl-config: command not found
./configure: sdl-config: command not found
./configure: wx-config: command not found
./configure: wx-config: command not found
./configure: xml2-config: command not found
./configure: xml2-config: command not found
Install prefix: /usr/local
Source path: /home/Robert/gpac
GPAC Version: 0.4.0-DEV
C compiler: gcc
make: make
CPU: x86
Big Endian: no
debug version: no
read-only version: no
fixed-point version: no
IsoMedia movie fragments: yes
SVG Support disabled: no

Detected libraries:
zlib: system
SpiderMonkey: no
FreeType: local
JPEG: system
PNG: system
MAD: no
FAAD: no
XVID: system
FFMPEG: no
Xiph OGG: system
Xiph Vorbis: system
Xiph Theora: no
Linux-Audio OSS: no
SDL Support: no
libXML2: no
OpenGL support: yes
OpenSSL support: no
DirectX Support: yes
wxWidgets support: no

Any ideas why I'm getting this?

Also, not a bug but I'd like to know, I'm using gpac in
conjunction with x264 for mp4 output, the header file
required was m4_author.h, this is no longer present.
Which header do we now require?

Regards

Reported by b-o-n-d on 2005-03-18 13:32 UTC
when opening mp4/3gp files in osmo4, i noticed that the
default option "all known files" doesnt lead to that .mp4
and .3gp files get displayed!

Reported by b-o-n-d on 2005-04-08 16:34 UTC
i noticed on a few samples that new mp4box places two
tx3g atoms in the stsd atom

mp4dump gives me:
type stbl
type stsd
version = 0 (0x00)
flags = 0 (0x000000)
entryCount = 2 (0x00000002)
type tx3g
data = <56 bytes>
00 00 00 00 00 00 00 01 00 00 00 00 01 ff 00 00
00 00 00 00 00 00 00 e0 02 40 00 00 00 00 00 01
00 12 ff ff ff ff 00 00 00 12 66 74 61 62 00 01
00 01 05 53 65 72 69 66
type tx3g
data = <56 bytes>
00 00 00 00 00 00 00 01 00 00 00 00 01 ff 00 00
00 00 00 00 00 00 00 e0 02 40 00 00 00 00 00 01
02 12 ff ff ff ff 00 00 00 12 66 74 61 62 00 01
00 01 05 53 65 72 69 66
type stts
...

the old mp4box compile i have (some pre 0.2.4final cvs
version) doesnt do this and only create one tx3g

also latest cvs doesnt remove the bifs/od streams
anymore when importing ttxt

Reported by freebee on 2004-09-21 13:44 UTC
make install-lib is generating a "error 127" message
because the install-headers command isn't typed directly
behind the install-lib (on the same line).

To fix this:

Old Code starts at line 111:

install-lib:
install-headers
mkdir -p "$(prefix)/lib"
install -m ...........

New code starts at line 111:

install-lib: install-headers
mkdir -p "$(prefix)/lib"
install -m ...........

In CVS version 0.2.1_dev (cvs update 14:25).

Reported by b-o-n-d on 2005-04-08 18:06 UTC
haali pointed me to the thing that mp4box's handling of
the fromchar/tochar thingie in ttxt might be wrong

when i convert your .srt sample from the regression
suite to .mp4 and than export to .ttxt it tells me that
values like fromChar="0" toChar="19" are used

first of all i wondered of course why the import from .srt
leads to the addition of these entries?

and than haali told me that the 'styl' boxes in the
samples have startChar=1 and endChar=0, he looked at
it via a hexeditor
he meant that it looks like mp4box writes ooks endChar
as int(32), not int(16) altough the spec says it's int(16)

i dunno if this is related to this, but i think that i saw
once that a ttxt player (realplayer) dropped the first letter
of a ttxt entry, dunno exactly anymore.
now also ffdshow gave me this, but only on the samples
where fromchar/tochar is used!?

Reported by doom9 on 2005-06-24 17:10 UTC
I've noted, that when importing large raw .264 streams,
the progress indicator resets back to 0 multiple times
and in the end does doesn't even come close to reaching
100% in the importing phase.

I'll attach the redirected stdout output from running
the release v0.3.0 with the following commandline

mp4box -add movie.264 -fps 25 -new x264.mp4

Reported by kagana on 2004-12-05 15:51 UTC
Hello.
There is a small problem to compile VS project files.
In gpac-0.2.2.tar.gz all *.dsp and *.dsw files are Unix
text files (enter is "\x0A" and not "\x0D\x0A").
When Microsoft VS(6.0 or .NET) see such a file, it get
crazy.
Thanks.
Arkadi.

Reported by jklaue on 2005-07-08 11:13 UTC

The hint track of my mp4-files contains
ImmediateDataEntries. However, if I try to get the actual
data of these entries they are always reported to be zero,
which is definitely wrong. I attached a short C-program
and a sample mp4-file. The error shows in line 76 of test.
c.

regards,
Jirka

*Reported by anonymous on 2005-08-26 14:00 UTC
You can view detailed and fine formated description here:

http://forum.doom9.org/showthread.php?p=704046\#post704046

Reported by b-o-n-d on 2004-03-15 08:47 UTC
as i tested lately the avi import function of mp4box i
found out that when playing the output in a directshow
player (no matter if with b-frames or not) it is resized to
320x240 always
now i found out that in early spec, the size put in the
video track was hardcoded to 320x240 in mp4 files
so normally the video decoder would have to parse the
decoderSpecificDescriptor before outputting a size, but
most popular mpeg-4 decoders like ffdshow, xvid, divx5
aso... dont do that (only 3ivx and nero seems to do this
in dshow)

anyways the early specs hardcoded this resolution but
the final one doesnt!
so i hope you can change this behaviour of mp4box and
allow setting the input resolution automatically :)

Reported by b-o-n-d on 2004-03-15 08:59 UTC
its not a big thing but it seems that mp4box isnt able to
read out the profile level set by xvid as it always reports
undefined or so
xvid uses an own, but spec compliant way of setting
their info as reported here:
http://sourceforge.net/tracker/index.php?
func=detail&aid=916429&group_id=84101&atid=571738
(edgomez is a xvid dev)

hope this helps :)

*Reported by anonymous on 2004-06-19 20:50 UTC
When creating an mp4 from an xvid 1.01 avi AS@L5, no
packed bitstream, mpeg Q, adaptive Q, QPel, VHQ 4,
trellis, and defaults using mp4box, I noticed that the
bitrate displayed for the video in mp4ui is way off.
The video is somewhere around 1400-1500 kbps and mp4ui
shows it as 300 kbps.

e-mail: [email protected]

Reported by stacey_melissa on 2005-04-19 00:04 UTC
Sometime between April 6 and April 10, the -fps
functionality was broken in mp4box. The result is that
when raw .264 video is added to a new .mp4, it is
stored as 23.975 fps video instead of 23.976. On long
clips, such as two-hour movies, this causes the video
to gradually lose sync with the video, and by the end,
there is about 1/3 second difference.

Adding an already-working video track from .mp4 works
fine. The problem is only with raw video.

I haven't tried it with different raw video types or
different framerate material, so I don't know about those.

There is a short discussion about the issue at
http://forum.doom9.org/showthread.php?s=&threadid=93086

Reported by moissinac on 2005-09-22 15:47 UTC
When opening
http://shadok.enst.fr:8080/sudoku/d8.svg

I get a crash in Osmo4
In downloader.c, in the http_do_requests function, in
the section:
exit:
if (e) {
gf_dm_disconnect(sess);
sess->status = GF_DOWNLOAD_STATE_UNAVAILABLE;
sess->last_error = e;
sess->OnDataRcv(sess, NULL, 0, sess->status, e);
}

The sess parameter for the OnDataRcv call seems to be
wrong.

A+

Jean-Claude

Reported by doom9 on 2005-04-23 11:59 UTC
I'm using the 0.2.5 DEV build from CVS, dated April 19th.

Trying to add more than one audio to an mp4, or create
a new mp4 with more than one audio track always ends up
throwing this error:

More than one audio track found, cannot ISMA'ized file

• remove extra track(s)

Error: Feature Not Supported

Here are a few sample commandlines:

mp4box -add audio1.mp4 -add audio2.mp4 -new output.mp4
mp4box -add audio1.mp4 -add audio2.aac -new output.mp4

mp4box -add video.m4v -add audio1.mp4 -add audio2.mp4
-new output.mp4

mp4box -add audio.mp4 output.mp4 (where output.mp4
already contains an audio track).

*Reported by anonymous on 2004-10-22 07:11 UTC
atoms_code.c(6649):*read ++;
=> (*read)++;
Was it already submitted?

-- qiushuiwuhen from China

*Reported by anonymous on 2005-08-01 07:37 UTC
GPAC fails to compile on Debian unstable with GCC 4.0 .
As it seems the problem is in
gpac/src/utils/downloader.c where declaration of
static void gf_dm_connect(GF_DownloadSession *sess);
is put inside the body of gf_dm_session_thread and gcc
complains about invalid storage class. The patch
applied fixes the problem.

*** /home/me/pool/gpac/gpac/src/utils/downloader.c~
2005-07-28 20:57:30.000000000 +0300
--- /home/me/pool/gpac/gpac/src/utils/downloader.c
2005-08-01 10:26:14.046203176 +0300

*** 396,406 ****
return GF_OK;
}

#define GF_WAIT_REPLY_SLEEP 20
static u32 gf_dm_session_thread(void *par)
{
! static void gf_dm_connect(GF_DownloadSession *sess);
GF_DownloadSession *sess = par;

sess->flags &= ~GF_DOWNLOAD_SESSION_THREAD_DEAD;
--- 396,408 ----
return GF_OK;
}

+ static void gf_dm_connect(GF_DownloadSession *sess);

#define GF_WAIT_REPLY_SLEEP 20
static u32 gf_dm_session_thread(void *par)
{
! // Zarro
! //static void gf_dm_connect(GF_DownloadSession *sess);
GF_DownloadSession *sess = par;

sess->flags &= ~GF_DOWNLOAD_SESSION_THREAD_DEAD;

*Reported by anonymous on 2005-04-25 17:53 UTC
http://www.fastvdo.com/H.264.html

muxing those raw streams with mp4box produces weird
results ;-)

->odd resolution (nero parser "corrects" this - haali's
splitter does not)

fastvdo or nero filter should be able to decode H.264
High Profile

nexus3

Reported by b-o-n-d on 2005-03-25 10:47 UTC
i noticed that you get an "import failed end of stream /
file" error and no .mp4 output when importing raw mpeg-
4 part2 video streams, which still contain n-vops, caused
by packed bitstream (!). this error comes right after
the "removed X n-vops" message
i created such streams, by simply dumping from a
packed bitstream .avi to raw. of course the packed
frames get unpacked with that, but the dummy n-vops
still stay in the stream
this error comes independtly of whether i used divx5 or
xvid to encode

the funny thing is that importing a raw stream which
contains "real" n-vops works correctly -> vfr stream gets
created

actually this issue isnt really important, but i thought it
doesnt hurt when reporting

*Reported by anonymous on 2005-07-31 15:03 UTC
please have a look at:
http://forum.doom9.org/showthread.php?p=693337\#post693337

Reported by b-o-n-d on 2005-02-11 17:46 UTC
when importing an mpeg-2 adts .aac file using SBR (SBR
aka HE-AAC is standardised for MPEG-2 AAC) mp4box
seems to force mpeg-4 aac and seems to set the profile
to mpeg-4 main@level1 (aac main) which isnt correct
(as it has to be level2 low complexity)

of course when exporting to raw with mp4box its still
signals main and not LC
when exporting the same stream to raw with mp4creator
it is borked

when importing the original adts stream without sbr
switch with mp4box it signals correctly mpeg-2 lc aac
aso...

the original stream was encoded with the nero aac
encoder by using the "extract iso 13818-7 track" option

Reported by doom9 on 2005-06-24 17:20 UTC
I have this two subtitle files, that can be muxed into
an mp4 just fine. However, if I import them and at the
same time split the output, after splitting 99% mp4box
freezes and starts using all CPU resources. If I leave
out the subtitle track and import just the video (be it
a .264, .m4v or an .mp4), things work fine, with and
without splitting.

Here is a sample commandline:

mp4box -add x264.mp4 -add english.srt -splits 15000
muxtest.mp4

This happens irrespective of the split size. If I leave
out -splits or the subtitle file, things work fine.

Reported by b-o-n-d on 2005-06-28 11:15 UTC
two different people reported that the x264 commandline
encoder (which uses gpac's lib to store the avc stream
in .mp4 while encoding) simply stops after some time

both used the ntsc framerate 23.976 and both got wierd
initial delay values (which i could reproduce. with 25fps i
get normal delays). the used b-frames and b-references
(aka arbitrary frame orders, b-pyramid)
i assume that maybe this wierd cts delay leads after
some time to the crash?

as the problem doesnt seem to happen when exporting
to raw .264, it seems to be caused by the gpac lib

you can read the whole story here
http://forum.doom9.org/showthread.php?t=95653

*Reported by anonymous on 2005-08-29 13:16 UTC
read more here:
doom9.org/showthread.php?p=705167#post705167

Reported by b-o-n-d on 2005-04-16 11:40 UTC
i noticed two problems at the end of ttxt streams when
importing from srt:

1. the total length of the ttxt stream seems to be
   ~1.5secs longer than the time of the last sample on all
   streams (compared .ttxt dump times and -info results)

2. i noticed on one sample that the last sample simply
   gets dropped when doing srt import. i couldnt reproduce
   this on all streams, but on the following one for example:

http://www.8ung.at/bond/ttxt\_missingsample.zip

this sample also shows the behaviour desribed under 1)

Reported by puntloos on 2004-10-31 11:19 UTC
Well, see subject..

My source file is a .3gp file:
http://130.89.22.201/lamb.3gp

If I use MP4Box -hint lamb.3gp it gets stuck and eats all
my system memory, then it dies because it gets killed by
linux's out of memory thingy.

The source was created with a nokia 3650 phone and
might be faulty, (not that I have a reason to assume it
is, I'm just saying I also don't have a way to VERYIFY the
integrity of the file) anyway, the resulting crash was a bit
drastic IMHO.

MP4Box comes from gpac-0.2.1.tar.gz - compiled with
default options, gcc-3.3.4, on linux Debian Testing.

I did notice that compiling did not go flawlessly, I got the
following error with make install: (yes, the make install,
not just 'make' failed for MP4Client which I dont use for
MP4Box but perhaps it points in the direction of the bug?

gcc -Wl,--warn-common -lpthread -ldl -o ../..
/bin/gcc/MP4Client main.o -L../../bin/gcc -lm4systems
-lm
main.o(.text+0x97): In function create_default_config': : undefined reference to NewIniFile'
main.o(.text+0x118): In function create_default_config': : undefined reference to IF_SetKey'
main.o(.text+0x13b): In function create_default_config': : undefined reference to IF_SetKey'
main.o(.text+0x15e): In function create_default_config': : undefined reference to IF_SetKey'
main.o(.text+0x181): In function create_default_config': : undefined reference to IF_SetKey'
main.o(.text+0x1a4): In function create_default_config': : undefined reference to IF_SetKey'
main.o(.text+0x1c7): more undefined references to
IF_SetKey' follow main.o(.text+0x3fc): In function create_default_config':
: undefined reference to IF_Delete' main.o(.text+0x40b): In function create_default_config':
: undefined reference to NewIniFile' main.o(.text+0x4ed): In function GPAC_EventProc':
: undefined reference to M4T_SetOption' main.o(.text+0x51c): In function GPAC_EventProc':
: undefined reference to M4T_GetCurrentTimeInMS' main.o(.text+0x655): In function GPAC_EventProc':
: undefined reference to M4T_PlayFromTime' main.o(.text+0x66e): In function GPAC_EventProc':
: undefined reference to M4T_Pause' main.o(.text+0x68e): In function GPAC_EventProc':
: undefined reference to M4T_GetCurrentTimeInMS' main.o(.text+0x7bd): In function GPAC_EventProc':
: undefined reference to M4T_GetCurrentFPS' main.o(.text+0x83e): In function GPAC_EventProc':
: undefined reference to M4T_Pause' main.o(.text+0x87d): In function GPAC_EventProc':
: undefined reference to M4T_Pause' main.o(.text+0xb84): In function GPAC_EventProc':
: undefined reference to M4ErrToString' main.o(.text+0xbde): In function set_navigation':
: undefined reference to M4T_GetOption' main.o(.text+0xcb1): In function set_navigation':
: undefined reference to M4ErrToString' main.o(.text+0xcf3): In function set_navigation':
: undefined reference to M4T_SetOption' main.o(.text+0xe2f): In function main':
: undefined reference to NewIniFile' main.o(.text+0xf1c): In function main':
: undefined reference to IF_GetKey' main.o(.text+0xf30): In function main':
: undefined reference to NewPluginManager' main.o(.text+0xf3e): In function main':
: undefined reference to PM_GetPluginsCount' main.o(.text+0xf6b): In function main':
: undefined reference to PM_Delete' main.o(.text+0xf79): In function main':
: undefined reference to IF_Delete' main.o(.text+0xffe): In function main':
: undefined reference to NewMPEG4Term' main.o(.text+0x107a): In function main':
: undefined reference to PM_GetPluginsCount' main.o(.text+0x1090): In function main':
: undefined reference to PM_GetFileName' main.o(.text+0x10d6): In function main':
: undefined reference to PM_Delete' main.o(.text+0x10e4): In function main':
: undefined reference to IF_Delete' main.o(.text+0x113b): In function main':
: undefined reference to IF_GetKey' main.o(.text+0x1170): In function main':
: undefined reference to IF_GetKey' main.o(.text+0x11a0): In function main':
: undefined reference to IF_GetKey' main.o(.text+0x121c): In function main':
: undefined reference to M4T_ConnectURL' main.o(.text+0x124d): In function main':
: undefined reference to linux_sleep' main.o(.text+0x128d): In function main':
: undefined reference to M4T_Delete' main.o(.text+0x12c3): In function main':
: undefined reference to PM_Delete' main.o(.text+0x12d1): In function main':
: undefined reference to IF_Delete' main.o(.text+0x1358): In function main':
: undefined reference to IF_GetKey' main.o(.text+0x13ac): In function main':
: undefined reference to M4T_Delete' main.o(.text+0x13e3): In function main':
: undefined reference to IF_SetKey' main.o(.text+0x13ee): In function main':
: undefined reference to NewMPEG4Term' main.o(.text+0x1417): In function main':
: undefined reference to IF_GetKey' main.o(.text+0x1451): In function main':
: undefined reference to M4T_ConnectURL' main.o(.text+0x14fc): In function main':
: undefined reference to M4T_SetOption' main.o(.text+0x1631): In function main':
: undefined reference to M4T_DumpSceneGraph' main.o(.text+0x164a): In function main':
: undefined reference to M4ErrToString' main.o(.text+0x1733): In function main':
: undefined reference to M4T_GetOption' main.o(.text+0x17ae): In function main':
: undefined reference to PM_GetPluginsCount' main.o(.text+0x17c4): In function main':
: undefined reference to PM_GetFileName' main.o(.text+0x185e): In function main':
: undefined reference to M4T_CloseURL' main.o(.text+0x18a8): In function main':
: undefined reference to M4T_ConnectURL' main.o(.text+0x18c9): In function main':
: undefined reference to M4T_Pause' main.o(.text+0x192f): In function main':
: undefined reference to M4T_Pause' main.o(.text+0x1962): In function main':
: undefined reference to M4T_CloseURL' main.o(.text+0x197b): In function main':
: undefined reference to M4T_Pause' main.o(.text+0x19ac): In function main':
: undefined reference to M4T_Pause' main.o(.text+0x19e7): In function main':
: undefined reference to M4T_GetCurrentTimeInMS' main.o(.text+0x1ad3): In function main':
: undefined reference to M4T_GetCurrentTimeInMS' main.o(.text+0x1d84): In function main':
: undefined reference to M4T_GetCurrentTimeInMS' main.o(.text+0x1e3f): In function main':
: undefined reference to M4T_PlayFromTime' main.o(.text+0x1e55): In function main':
: undefined reference to M4T_Pause' main.o(.text+0x1e77): In function main':
: undefined reference to M4T_CloseURL' main.o(.text+0x1e9e): In function main':
: undefined reference to M4T_ConnectURL' main.o(.text+0x205d): In function PrintWorldInfo':
: undefined reference to NewChain' main.o(.text+0x2071): In function PrintWorldInfo':
: undefined reference to M4T_GetWorldInfo' main.o(.text+0x20a4): In function PrintWorldInfo':
: undefined reference to ChainGetCount' main.o(.text+0x20b5): In function PrintWorldInfo':
: undefined reference to ChainGetEntry' main.o(.text+0x2139): In function PrintODList':
: undefined reference to M4T_GetRootOD' main.o(.text+0x2158): In function PrintODList':
: undefined reference to M4T_GetODInfo' main.o(.text+0x220a): In function PrintODList':
: undefined reference to M4T_GetODCount' main.o(.text+0x222f): In function PrintODList':
: undefined reference to M4T_GetODManager' main.o(.text+0x2253): In function PrintODList':
: undefined reference to M4T_GetODInfo' main.o(.text+0x22d3): In function ViewOD':
: undefined reference to M4T_GetRootOD' main.o(.text+0x22fe): In function ViewOD':
: undefined reference to M4T_GetODInfo' main.o(.text+0x2514): In function ViewOD':
: undefined reference to ChainGetCount' main.o(.text+0x256f): In function ViewOD':
: undefined reference to ChainGetCount' main.o(.text+0x279f): In function ViewOD':
: undefined reference to ChainGetEntry' main.o(.text+0x2857): In function ViewOD':
: undefined reference to ChainGetCount' main.o(.text+0x286e): In function ViewOD':
: undefined reference to ChainGetEntry' main.o(.text+0x28bc): In function ViewOD':
: undefined reference to ChainGetEntry' main.o(.text+0x298b): In function ViewOD':
: undefined reference to ChainGetCount' main.o(.text+0x29bd): In function ViewOD':
: undefined reference to ChainGetEntry' main.o(.text+0x2ece): In function ViewOD':
: undefined reference to M4T_GetRemoteOD' main.o(.text+0x2eee): In function ViewOD':
: undefined reference to M4T_GetODInfo' main.o(.text+0x2f38): In function ViewOD':
: undefined reference to M4T_GetODCount' main.o(.text+0x2f72): In function ViewOD':
: undefined reference to M4T_GetODManager' main.o(.text+0x2f92): In function ViewOD':
: undefined reference to M4T_GetODInfo' main.o(.text+0x2ffe): In function PrintODTiming':
: undefined reference to M4T_GetODInfo' main.o(.text+0x3278): In function PrintODTiming':
: undefined reference to M4T_GetRemoteOD' main.o(.text+0x3292): In function PrintODTiming':
: undefined reference to M4T_GetODInfo' main.o(.text+0x32ee): In function PrintODBuffer':
: undefined reference to M4T_GetODInfo' main.o(.text+0x3498): In function PrintODBuffer':
: undefined reference to M4T_GetRemoteOD' main.o(.text+0x34b2): In function PrintODBuffer':
: undefined reference to M4T_GetODInfo' main.o(.text+0x34f6): In function ViewODs':
: undefined reference to M4T_GetRootOD' main.o(.text+0x3521): In function ViewODs':
: undefined reference to M4T_GetODCount' main.o(.text+0x3554): In function ViewODs':
: undefined reference to M4T_GetODManager' main.o(.text+0x35e9): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x361e): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x365b): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x3698): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x36cf): In function PrintGPACConfig':
: undefined reference to M4T_GetOption' main.o(.text+0x3712): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x374f): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x37b3): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x37e3): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x383b): In function PrintGPACConfig':
: undefined reference to IF_GetKey' main.o(.text+0x3878): more undefined references to IF_GetKey' follow
main.o(.text+0x3d90): In function CheckResume': : undefined reference to M4T_Pause'
main.o(.text+0x3de2): In function loadconfigfile': : undefined reference to NewIniFile'
main.o(.text+0x3ef4): In function list_plugins': : undefined reference to PM_GetPluginsCount'
main.o(.text+0x3f04): In function list_plugins': : undefined reference to PM_GetFileName'
main.o(.text+0x20eb): In function PrintWorldInfo': : undefined reference to DeleteChain'
collect2: ld returned 1 exit status
make[3]: *** [MP4Client] Error 1
make[3]: Leaving directory
/usr/local/src/gpac/Applications/MP4Client' make[2]: *** [install] Error 2 make[2]: Leaving directory /usr/local/src/gpac/Applications/MP4Client'
make[1]: *** [install] Error 2
make[1]: Leaving directory
`/usr/local/src/gpac/Applications'
make: *** [install] Error 2

Reported by yurivkhan on 2005-09-18 14:33 UTC
OS version:
Windows XP Pro, English, with SP2

mp4box.exe version:
MP4Box - GPAC version 0.4.0-DEV
GPAC Copyright: (c) Jean Le Feuvre 2000-2005
(c) ENST 2005-200X

Reproducible: Always

To reproduce:

• Log on to Windows using a limited account
• Use a drive that is formatted using NTFS
• mp4box.exe -new -add in.mp4 -add in.aac out.mp4

Expected behavior:

• mp4box creates a new file out.mp4 that contains
  tracks from in.mp4 and in.aac

Observed behavior:

• mp4box fails with this error message:
  Cannot open destination file out.mp4: I/O Error

Cause:

• mp4box attempts to create a file with a random name
  in the root directory of the current drive, e.g. L:\t3b4
• root directory is by default not writable by limited
  user accounts

Workarounds:

• use mp4box on a drive formatted with FAT16 or FAT32
• grant user write access to the root directory
• use the SUBST command to emulate a drive whose root
  will be in a writable directory

Observations:

• no writes are performed on the file in question
• the file is deleted immediately after creation

Proposed solution:

• remove unnecessary file creation

Reported by zwommie on 2005-05-23 01:05 UTC
Hi,

i created some QuickTime TeXML samples for a little demo.
I also wanted a ttxt version of them, so I added them
to an MP4-file (mp4box -add sample.xml sample.mp4) and
dumped the texttrack in a ttxt file(mp4box -ttxt 1
sample.mp4).

There are some problems u might considere to fix in
that conversion to an MP4-file, take a look at the
uploaded examples.

I notices Osmo shows the correct fonttypes, QT doesnt
at all. The xml-samples should play correct in QT
though. But this isnt about the fonts here. I tried
Times and Serif : the fonttypes are an QT problem I
guess, but there are some conversion problems:

- scrolling doesnt work anymore

• some textsamples are duplicated
• blink conversion was 1 char too short
• selection conversion 1 char too short
• the tab/crlf arent recognized
• the link ist working

I uploaded the original XMT-files, the created
MP4-files from xmt and the dumped ttxt tracks.

Reported by b-o-n-d on 2005-01-10 14:42 UTC
when trying to export an aac stream from .mp4 with
the "-raw" option mp4box always says "extracting mpeg
audio - compressor unknown" and creates a file with
the .mpeg extension (maybe in the .mpeg container?),
which than isnt playable by any tool (eg foobar, winamp)
that is supposed to handle adts .aac streams
(this is the case with files encoded with qt, nero and
sorenson)

the option "-single" seems to work tough

with a compile of version 0.2.2 the extraction works fine

Reported by b-o-n-d on 2005-03-09 08:13 UTC
when analysing a mp3-in-mp4 stream with mp4box' -info
option it always displays "layer 1" which is obviously not
correct

this happens with mp3s muxed into .mp4 by mp4box,
mp4creator and 3ivx so i assume its a bug in the -info
module of mp4box

on .mp1 streams muxed into .mp4 with mp4creator,
mp4box displays "layer3", so i assume -info somehow
mixes up mp1 and mp3 internally?

mp2 is displayed correctly