Something wrong with mp4box (BUT THIS TIME REALY)

[shon3i]

I have mp4box 0.4.1-DEV. but mp4 maded by him is not real mp4 files and rejected by my NeroDigital standalone.

For example:

I encode few tracks with this new nero encoder and everything is fine, but when i convert tracks via Coding Technologies and mux aac's into mp4's, player rejected this files.

but when i rip this tracks directly from winamp aslo is everything fine.

in short:

CD->NDMP4->PLAY
CD->WINAMP(CT)->PLAY
CD->CTMP4(via mp4box)->ERROR

comandline i used:

mp4box -add file.aac -new file.mp4

files are LC @ 128kbps, same thing hapen's when extract nero files to aac and then again mux to mp4.

So then mp4box have some bug, or what?

[Kurtnoise]

Copy & Paste here some infos about your files with this cmdl :

MP4Box -info NDMP4.mp4

and

MP4Box -info CTMP4.mp4

To compare each others.

[shon3i]

OK here is

Nero MP4 file created with this last Nero encoder and plays fine on standalone

* Movie Info *
Timescale 1000 - Duration 00:03:42.280
Fragmented File no - 1 track(s)
File Brand mp42 - version 0
Created: GMT Tue May 16 21:38:06 2006

File has root IOD
Scene PL 0xff - Graphics PL 0xff - OD PL 0xff
Visual PL: No visual capability required (0xff)
Audio PL: No audio capability required (0xff)
No streams included in root OD

Chapters:
Chapter #1 - 00:00:00.059 - ""

Track # 1 Info - TrackID 1 - TimeScale 44100 - Duration 00:03:44.406
Media Info: Language "Undetermined" - Type "soun" - Sub Type "mp4a" - 9665 samples
MPEG-4 Config: Audio Stream - ObjectTypeIndication 0x40
MPEG-4 Audio AAC LC - 2 Channel(s) - SampleRate 44100
Self-synchronized

Nero MP4 remuxed with YAMB, first extracted to AAC then again muxed to new mp4, not playable on standalone

* Movie Info *
Timescale 600 - Duration 00:03:44.420
Fragmented File no - 3 track(s)
File Brand mp42 - version 1
Created: GMT Mon May 15 21:45:18 2006

File has root IOD
Scene PL 0x01 - Graphics PL 0x01 - OD PL 0x01
Visual PL: No visual capability required (0xff)
Audio PL: AAC Profile @ Level 2 (0x29)

Track # 1 Info - TrackID 101 - TimeScale 44100 - Duration 00:03:44.420
Media Info: Language "Undetermined" - Type "soun" - Sub Type "mp4a" - 9665 sampl
es
MPEG-4 Config: Audio Stream - ObjectTypeIndication 0x40
MPEG-4 Audio AAC LC - 2 Channel(s) - SampleRate 44100
Synchronized on stream 1

Track # 2 Info - TrackID 2 - TimeScale 600 - Duration 00:00:01.000
Track is present in Root OD
Media Info: Language "Undetermined" - Type "odsm" - Sub Type "mp4s" - 1 samples
MPEG-4 Config: ObjectDescriptor Stream - ObjectTypeIndication 0x01
Synchronized on stream 1

Track # 3 Info - TrackID 1 - TimeScale 600 - Duration 00:00:01.000
Track is present in Root OD
Media Info: Language "Undetermined" - Type "sdsm" - Sub Type "mp4s" - 1 samples
MPEG-4 Config: SceneDescription Stream - ObjectTypeIndication 0x02
BIFS Scene description - Command stream
Self-synchronized

Coding Technologies encoded via Winamp, and playable on standalone, everything is fine.

* Movie Info *
Timescale 90000 - Duration 00:03:43.940
Fragmented File no - 3 track(s)
File Brand mp42 - version 0
Created: GMT Tue May 16 21:41:46 2006

File has root IOD
Scene PL 0xff - Graphics PL 0xff - OD PL 0xff
Visual PL: No visual capability required (0xff)
Audio PL: High Quality Audio Profile @ Level 2 (0x0f)

Track # 1 Info - TrackID 1 - TimeScale 44100 - Duration 00:03:44.026
Media Info: Language "Undetermined" - Type "soun" - Sub Type "mp4a" - 9648 sampl
es
MPEG-4 Config: Audio Stream - ObjectTypeIndication 0x40
MPEG-4 Audio AAC LC - 2 Channel(s) - SampleRate 44100
Synchronized on stream 2

Track # 2 Info - TrackID 2 - TimeScale 1000 - Duration 00:03:43.940
Track is present in Root OD
Media Info: Language "Undetermined" - Type "odsm" - Sub Type "mp4s" - 1 samples
MPEG-4 Config: ObjectDescriptor Stream - ObjectTypeIndication 0x01
Self-synchronized

Track # 3 Info - TrackID 3 - TimeScale 1000 - Duration 00:03:43.940
Track is present in Root OD
Media Info: Language "Undetermined" - Type "sdsm" - Sub Type "mp4s" - 1 samples
MPEG-4 Config: SceneDescription Stream - ObjectTypeIndication 0x02
BIFS Scene description - Command stream
Synchronized on stream 2


Coding Technologies encoded by MeGUI and muxed to mp4 via mp4box, aslo can't be played.

* Movie Info *
Timescale 600 - Duration 00:03:44.350
Fragmented File no - 3 track(s)
File Brand mp42 - version 1
Created: GMT Mon May 15 21:40:26 2006

File has root IOD
Scene PL 0x01 - Graphics PL 0x01 - OD PL 0x01
Visual PL: No visual capability required (0xff)
Audio PL: AAC Profile @ Level 2 (0x29)

Track # 1 Info - TrackID 101 - TimeScale 44100 - Duration 00:03:44.351
Media Info: Language "Undetermined" - Type "soun" - Sub Type "mp4a" - 9662 sampl
es
MPEG-4 Config: Audio Stream - ObjectTypeIndication 0x40
MPEG-4 Audio AAC LC - 2 Channel(s) - SampleRate 44100
Synchronized on stream 1

Track # 2 Info - TrackID 2 - TimeScale 600 - Duration 00:00:01.000
Track is present in Root OD
Media Info: Language "Undetermined" - Type "odsm" - Sub Type "mp4s" - 1 samples
MPEG-4 Config: ObjectDescriptor Stream - ObjectTypeIndication 0x01
Synchronized on stream 1

Track # 3 Info - TrackID 1 - TimeScale 600 - Duration 00:00:01.000
Track is present in Root OD
Media Info: Language "Undetermined" - Type "sdsm" - Sub Type "mp4s" - 1 samples
MPEG-4 Config: SceneDescription Stream - ObjectTypeIndication 0x02
BIFS Scene description - Command stream
Self-synchronized

So what is wrong.

[Kurtnoise]

Maybe the brand name version...

[shon3i]

Maybe, but why then player, plays mp4 encoded with winamp. Only mp4's created by mp4box are not played. aslo i found small app ReplayGain for AAC, works perfetly only with winamp's and nero's mp4, but with mp4box mp4 says that is not valid mp4 file.


Something defintly wrong in mp4box or there is some option.

[Drachir]

Compare/post the output from mp4dump(mpeg4ip).

[shon3i]

Here is a dump's (http://www.mytempdir.com/669732), aslo i try to mux aac with some version of m4UI and everything works perfect

[mediator]

tkhd.trackId is illegal for the two files that are reported not to work. This might not be the problem (no idea why a parser should have a problem about that), but it's noticable anyhow.

[Drachir]

tkhd.trackId is illegal for the two files that are reported not to work. This might not be the problem (no idea why a parser should have a problem about that), but it's noticable anyhow.
Why should the ID 101 be illegal? The ID of 101 is allowed.

[mediator]

I meant mvhd.nextTrackId is illegal (it has the same value as one of the tracks), sorry for the confusion

[Drachir]

I meant mvhd.nextTrackId is illegal (it has the same value as one of the tracks), sorry for the confusion
Shouldn't be a problem.


next_track_ID is a non-zero integer that indicates a value to use for the track ID of the next track to be
added to this presentation. Zero is not a valid track ID value. The value of next_track_ID shall be
larger than the largest track-ID in use. If this value is equal to or larger than all 1s (32-bit maxint), and
a new media track is to be added, then a search must be made in the file for a unused track identifier.

[mediator]

You never know if it really is the problem or not until you check it. After all the file in general seems to have no problem, doesn't it ? ;)

[shon3i]

So what i can do?

[mediator]

provide me with a test file (small,please !!) and I will patch it for you.

[shon3i]

small sample (http://www.mytempdir.com/695712)

[mediator]

I had a closer look on your file.
I guess the reason for the problem is probably not the trackId. More likely is that some parsers are confused by the fact that the two problematic files store decSpecificInfo explicitly for the OD track (using tag + length syntax) but actually with zero lenght data. According to my understanding it would be better not to save decSpecificInfo at all in this case. You should be able to verify yourself if this is the problem, if you just strip any BIFS and OD tracks, which will also strip the stuff which I consider as problematic. If this doesn't help, let me know.

-mediator

PS: why not kick BIFS and OD tracks in general? In current real-life scenarios they are not needed. And I also see no tendence for the future...

[Drachir]

PS: why not kick BIFS and OD tracks in general? In current real-life scenarios they are not needed. And I also see no tendence for the future...
If you don't want BIFS/OD use 3GP or maybe MOV. (just my opinion)

[shon3i]

I had a closer look on your file.
I guess the reason for the problem is probably not the trackId. More likely is that some parsers are confused by the fact that the two problematic files store decSpecificInfo explicitly for the OD track (using tag + length syntax) but actually with zero lenght data. According to my understanding it would be better not to save decSpecificInfo at all in this case. You should be able to verify yourself if this is the problem, if you just strip any BIFS and OD tracks, which will also strip the stuff which I consider as problematic. If this doesn't help, let me know.

-mediator

PS: why not kick BIFS and OD tracks in general? In current real-life scenarios they are not needed. And I also see no tendence for the future...
Can you write mp4box comandline becouse i am confused with mp4box switches, Thanks

[Drachir]

Use "MP4Box -no-sys ..."

[shon3i]

Use "MP4Box -no-sys ..."
Same problem.

[bond]

did you contact the standalone manufacturer cause its propably a bug on their side

[shon3i]

did you contact the standalone manufacturer cause its propably a bug on their side
No but i will try. I have one more trouble with this mp4box mp4 files. I use small tool caled AAC Gain (http://www.rarewares.org/files/aac/aacgain.zip) to incrase volume with replygain for AAC, but aslo can't use it for mp4 maded by mp4box, only for nero or winamp's mp4. I recivie error mp4 is not valid mpeg file or something like that

[Kurtnoise]

I think aacgain doesn't support mp4 files with multiple trak atoms. By default, MP4Box introduces this during the mux (at least with mp4 audio streams only)...

[shon3i]

I dont know Kurt, but mp4box's mp4 is almost like winamp's mp4 but both aacgain and standalone reject these files, aslo mp4's created by mp4 UI works fine on both situations.

[shon3i]

Since now Winamp 5.25 build 834 is out, i downloaded it i try to play my mp4 collection of files which encoded by CT encoder to aac and than muxed to mp4 via mp4box (via BeHappy) files are played horrible like fast forward is on. I think that is Winamp problem because files plays fine on lastets winamp and i contact winamp devs, but he says this

http://forums.winamp.com/showthread.php?postid=2012745#post2012745

Same problem i haved with my standalone which i mentioned in my first post

[SeeMoreDigital]

It sounds as though you have a sample-rate signalling issue somewhere!

Are your encodes AAC-LC or AAC-HE. If they are AAC-HE, did you remember to apply the "SBR" flag in MP4Box?

Also, you might have more consistent playback success if you changed the ".MP4" container to ".M4A" for your audio-only muxes.


Cheers

[shon3i]

Are your encodes AAC-LC or AAC-HE. If they are AAC-HE, did you remember to apply the "SBR" flag in MP4Box?File is HE-AAC and is backward compatible (i use -sbr switch), Like winamp dev says

The file is messed up. It's specifying each MP4 frame's duration as being 1024, when it's actually 2048. This causes Winamp to only play back half of each frame. I've fixed for now (by only checking the duration on the last frame of the file), but you need to report the bug back to the author of the tool you used to encode this file.

ME: Why then worked in earler versions?

It worked in prior versions because they were never checking the durations, but it seems to be an mp4box bug (the problem is in the MP4 container, not in the AAC data).

Also, you might have more consistent playback success if you changed the ".MP4" container to ".M4A" for your audio-only muxes.Nothing is changed, same effect

here is sample (http://www.mytempdir.com/874411), download last winamp, and try.

[Drachir]

If there is a reproducible problem, write a pugreport here: http://sourceforge.net/tracker/?atid=571738&group_id=84101&func=browse

But it can take some time until you get an answer.

[SeeMoreDigital]

File is HE-AAC and is backward compatible (i use -sbr switch)...Thanks for the confirm. At least that removes one possibility ;)

With respect to your "hardware" player. Which device is this?

I'm not familiar with any hardware devices that fully support AAC-HE...


Cheers

[shon3i]

With respect to your "hardware" player. Which device is this?
I now have two players one SkyVision (NeroDigital ASP+AAC support), and second is Xoro 310 (aslo NeroDigital ASP+AAC support), both players plays fine movies with aac audio but not muxed with mp4box, only recode files, aslo plays fine HE-AAC files, only direct encoded to mp4 from winamp or nero cli encoder, if i remux that files or make aac than mux after with mp4box, files can't be played anymore, same thing now with winamp

[SeeMoreDigital]

Well, technically your Xoro HSD 310 stand-alone player can't "fully" decode AAC-HE audio because its ESS Vibratto II chip-set does not support SBR.... That said, it does sound very good :)

[shon3i]

it does sound very good Not like skyvision, but not that noticable, but nevermind on this, this is not problem, aslo LC-AAC created by mp4box have same error.

[Drachir]

The file is messed up. It's specifying each MP4 frame's duration as being 1024, when it's actually 2048. This causes Winamp to only play back half of each frame. I've fixed for now (by only checking the duration on the last frame of the file), but you need to report the bug back to the author of the tool you used to encode this file.


I am not an audio/aac expert but isn't the chunk size of aac normally 1024? Where does the 2048 com from?
Maybe someone with more knowledge about this should write a comment.

Like a told you, if your are sure there is an error, write a bugreport.

[SeeMoreDigital]

Not like skyvision, but not that noticable, but nevermind on this, this is not problem, aslo LC-AAC created by mp4box have same error.Can you give this 2Ch_AAC-HE_Test_01 (http://82.10.220.174/Uploaded_Files/Doom9_Forum_files/2Ch_AAC-HE_Test_01.7z) a try?


Cheers

[shon3i]

Can you give this 2Ch_AAC-HE_Test_01 (http://82.10.220.174/Uploaded_Files/Doom9_Forum_files/2Ch_AAC-HE_Test_01.7z) a try?


Cheers
Same error SMD, file won't play in standalone, and in winamp plays with same fast forward efect.


Like a told you, if your are sure there is an error, write a bugreport.

I repack that file with mp4ui and file plays fine. So than

Nero files works
Winamp files works
mp4UI files works

then must be mp4box bug, or i missing some option.

[bond]

make a bug report on the gpac bugtracker plz

[Drachir]

Bug reports can be filled out here: http://sourceforge.net/tracker/?atid=571738&group_id=84101&func=browse

I don't think that Jean Le Feuvre or some one from ENST often take a look into this forum. If you want it to get fixed you have to report the bug.

[shon3i]

OK, thanks, i will report bug.