[FFmpeg-user] Audio Only Dolby Digital Plus

[Ronak]

Hi All,

I'm trying to get Ffmpeg to package and generate a valid fMP4 file for HLS of 
an Audio Only Dolby Digital Plus file. However, the stream that I get is 
unplayable and has no audio in its sample tables.

Here's an example command:

ffmpeg -i atmosTest.mp4 -codec copy -hls_time 0.993 -hls_segment_type fmp4 
-hls_flags single_file -hls_playlist_type vod atmosTestffmpeg.m3u8
ffmpeg version N-92272-g0e9c01f Copyright (c) 2000-2018 the FFmpeg developers
  built with Apple LLVM version 10.0.0 (clang-1000.11.45.2)
  configuration: --prefix=/Users/ronakp/ffmpeg_build 
--pkg-config-flags=--static --extra-cflags=-I/Users/ronakp/ffmpeg_build/include 
--extra-ldflags=-L/Users/ronakp/ffmpeg_build/lib --extra-libs=-lpthread 
--extra-libs=-lm --bindir=/Users/ronakp/bin --enable-libfdk_aac 
--enable-libmp3lame --enable-libopus --enable-nonfree
  libavutil  56. 20.100 / 56. 20.100
  libavcodec 58. 34.100 / 58. 34.100
  libavformat58. 19.102 / 58. 19.102
  libavdevice58.  4.106 / 58.  4.106
  libavfilter 7. 38.100 /  7. 38.100
  libswscale  5.  2.100 /  5.  2.100
  libswresample   3.  2.100 /  3.  2.100
Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'atmosTest.mp4':
  Metadata:
major_brand : mp42
minor_version   : 1
compatible_brands: mp42dby1isom
creation_time   : 2018-09-07T19:17:49.00Z
  Duration: 00:02:40.80, start: 0.00, bitrate: 769 kb/s
Stream #0:0(und): Audio: eac3 (ec-3 / 0x332D6365), 48000 Hz, 5.1(side), 
fltp, 768 kb/s (default)
Metadata:
  creation_time   : 2018-09-07T19:17:49.00Z
  handler_name: sound handler
Side data:
  audio service type: main
[hls @ 0x7fbfc4825c00] Opening 'atmosTestffmpeg.m4s' for writing
[mp4 @ 0x7fbfc4827600] track 0: codec frame size is not set
Output #0, hls, to 'atmosTestffmpeg.m3u8':
  Metadata:
major_brand : mp42
minor_version   : 1
compatible_brands: mp42dby1isom
encoder : Lavf58.19.102
Stream #0:0(und): Audio: eac3 (ec-3 / 0x332D6365), 48000 Hz, 5.1(side), 
fltp, 768 kb/s (default)
Metadata:
  creation_time   : 2018-09-07T19:17:49.00Z
  handler_name: sound handler
Side data:
  audio service type: main
Stream mapping:
  Stream #0:0 -> #0:0 (copy)
Press [q] to stop, [?] for help
size=N/A time=00:02:40.76 bitrate=N/A speed= 397x
video:0kB audio:15075kB subtitle:0kB other streams:0kB global headers:0kB 
muxing overhead: unknown

When I try to run Apple's mediastreamvalidator on this:


a45e60e46a79:Dolby ronakp$ mediastreamvalidator atmosTestffmpeg.m3u8
mediastreamvalidator: Version 1.2(180612)

[atmosTestffmpeg.m3u8] Started loading root playlist
[atmosTestffmpeg.m3u8] Started loading media playlist
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting segment data
Error injecting 

[FFmpeg-user] Audio Only Fragmented MP4 - SIDX

[Ronak]

Hey All,

I noticed that ffmpeg always seems to generate a SIDX box when it's not really 
required for Audio Only content.

Example command:

ffmpeg -i atmosTest.mp4 -codec copy -hls_time 0.993 -hls_segment_type fmp4 
-hls_flags single_file -hls_playlist_type vod atmosTestffmpeg.m3u8
ffmpeg version N-92272-g0e9c01f Copyright (c) 2000-2018 the FFmpeg developers
  built with Apple LLVM version 10.0.0 (clang-1000.11.45.2)
  configuration: --prefix=/Users/ronakp/ffmpeg_build 
--pkg-config-flags=--static --extra-cflags=-I/Users/ronakp/ffmpeg_build/include 
--extra-ldflags=-L/Users/ronakp/ffmpeg_build/lib --extra-libs=-lpthread 
--extra-libs=-lm --bindir=/Users/ronakp/bin --enable-libfdk_aac 
--enable-libmp3lame --enable-libopus --enable-nonfree
  libavutil  56. 20.100 / 56. 20.100
  libavcodec 58. 34.100 / 58. 34.100
  libavformat58. 19.102 / 58. 19.102
  libavdevice58.  4.106 / 58.  4.106
  libavfilter 7. 38.100 /  7. 38.100
  libswscale  5.  2.100 /  5.  2.100
  libswresample   3.  2.100 /  3.  2.100
Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'atmosTest.mp4':
  Metadata:
major_brand : mp42
minor_version   : 1
compatible_brands: mp42dby1isom
creation_time   : 2018-09-07T19:17:49.00Z
  Duration: 00:02:40.80, start: 0.00, bitrate: 769 kb/s
Stream #0:0(und): Audio: eac3 (ec-3 / 0x332D6365), 48000 Hz, 5.1(side), 
fltp, 768 kb/s (default)
Metadata:
  creation_time   : 2018-09-07T19:17:49.00Z
  handler_name: sound handler
Side data:
  audio service type: main
[hls @ 0x7fbfc4825c00] Opening 'atmosTestffmpeg.m4s' for writing
[mp4 @ 0x7fbfc4827600] track 0: codec frame size is not set
Output #0, hls, to 'atmosTestffmpeg.m3u8':
  Metadata:
major_brand : mp42
minor_version   : 1
compatible_brands: mp42dby1isom
encoder : Lavf58.19.102
Stream #0:0(und): Audio: eac3 (ec-3 / 0x332D6365), 48000 Hz, 5.1(side), 
fltp, 768 kb/s (default)
Metadata:
  creation_time   : 2018-09-07T19:17:49.00Z
  handler_name: sound handler
Side data:
  audio service type: main
Stream mapping:
  Stream #0:0 -> #0:0 (copy)
Press [q] to stop, [?] for help
size=N/A time=00:02:40.76 bitrate=N/A speed= 397x
video:0kB audio:15075kB subtitle:0kB other streams:0kB global headers:0kB 
muxing overhead: unknown

When you look at the generated .m4s file, you will see ftyp, moov, followed by 
a collection of sidx, moof and mdat atoms. 


However, when I run a similar command using Apple's Tooling:


a45e60e46a79:Dolby ronakp$ mediafilesegmenter -a -r -s atmosTest.ec3
Nov 12 2018 11:58:52.713: ISO fragmented mode, forcing segments to start with 
I-Frame
Nov 12 2018 11:58:52.714: Processing file 
/Users/ronakp/Desktop/Dolby/atmosTest.ec3
Nov 12 2018 11:58:52.743: Wrote 0 to 578 to main.mp4
Nov 12 2018 11:58:52.748: Wrote 578 to 959266 to main.mp4
Nov 12 2018 11:58:52.749: segment bitrate 768.18 kbits/sec is new max
Nov 12 2018 11:58:52.756: Wrote 959266 to 1917954 to main.mp4
Nov 12 2018 11:58:52.763: Wrote 1917954 to 2876642 to main.mp4
Nov 12 2018 11:58:52.772: Wrote 2876642 to 3835330 to main.mp4
Nov 12 2018 11:58:52.777: Wrote 3835330 to 4794018 to main.mp4
Nov 12 2018 11:58:52.784: Wrote 4794018 to 5752706 to main.mp4
Nov 12 2018 11:58:52.793: Wrote 5752706 to 6711394 to main.mp4
Nov 12 2018 11:58:52.800: Wrote 6711394 to 7670082 to main.mp4
Nov 12 2018 11:58:52.880: Wrote 7670082 to 8628770 to main.mp4
Nov 12 2018 11:58:52.924: Wrote 8628770 to 9587458 to main.mp4
Nov 12 2018 11:58:52.966: Wrote 9587458 to 10546146 to main.mp4
Nov 12 2018 11:58:52.973: Wrote 10546146 to 11504834 to main.mp4
Nov 12 2018 11:58:53.017: Wrote 11504834 to 12463522 to main.mp4
Nov 12 2018 11:58:53.060: Wrote 12463522 to 13422210 to main.mp4
Nov 12 2018 11:58:53.102: Wrote 13422210 to 14380898 to main.mp4
Nov 12 2018 11:58:53.147: Wrote 14380898 to 15339586 to main.mp4
Nov 12 2018 11:58:53.147: Wrote 15339586 to 15441066 to main.mp4
Nov 12 2018 11:58:53.147: segment bitrate 768.24 kbits/sec is new max
Nov 12 2018 11:58:53.147: Unknown video dynamic range value 0
Nov 12 2018 11:58:53.147: average bit rate is 768.21 kbits/sec - max file bit 
rate is 768.24 kbits/sec

It generates a file without any SIDX atoms; only ftyp, moov & a collection of 
moof & mdat atoms.

Why does ffmpeg do this? How hard would it be to remove this behavior from 
ffmpeg?

Thanks,

Ronak




___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-user] can't reorder streams and concatenate vob's

[Nicolas George]

sean darcy (2018-11-12):
> So for vob's from a video dvd :
> 
> cat VTS* | ffmpeg -i - ... out.mp4
> 
> or
> 
> ffmpeg -f concat -i list.of.vobs  .. out.mp4
> 
> is not a good idea ? Even though it does not generate any errors.

Not a good idea at all. It generates garbage video.

> I should be using a complex filter for any vob's from a video dvd ? Who knew
> ?

No. You need to use a tool aware of the specifics of the DVD Video
format. The VOB files there are not video containers, they are a
container for a kind of virtual filesystem on top of the real
filesystem, indexed by the IFO files. You need a tool that can
understand that.

The dvd2concat script shipped with FFmpeg is an example of such a tool,
but it is very rudimentary.

Please do not Cc me.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-user] can't reorder streams and concatenate vob's

[sean darcy]

On 11/8/18 5:34 PM, Nicolas George wrote:

sean darcy (2018-11-08):

Ok. So is it possible to concatenate 2 vob's where the audio and video don't
have the same stream numbers ?


VOB do not have stream numbers (FFmpeg is inventing them on the fly), so
the question does not make any sense.


for input in *.VOB ; do
   ffmpeg -i $input -c copy  $(basename $input .VOB).[???]
   echo $(basename $input .VOB).[???] >> concat.txt
done
ffmpeg -f concat -i concat.txt ..out.mp4

Bigger question : is the concatenation problem here the stream numbers ?


In this instance, if the files are the ones you have shown earlier,
yes, the order in which FFmpeg detects the streams is the problem.

I suggest you read the documentation of the concat demuxer to the end,
the answer to your question is there.

Note that VOB files are also concatenable at the container level,
provided they do not contain extra junk at the beginning or the end (if
you got them directly from a DVD Video filesystem, they do and you
should NOT be using them directly).

Regards,



On 11/8/18 5:34 PM, Nicolas George wrote:
> sean darcy (2018-11-08):
>> Ok. So is it possible to concatenate 2 vob's where the audio and 
video don't

>> have the same stream numbers ?
>
> VOB do not have stream numbers (FFmpeg is inventing them on the fly), so
> the question does not make any sense.
>
>> for input in *.VOB ; do
>>ffmpeg -i $input -c copy  $(basename $input .VOB).[???]
>>echo $(basename $input .VOB).[???] >> concat.txt
>> done
>> ffmpeg -f concat -i concat.txt ..out.mp4
>>
>> Bigger question : is the concatenation problem here the stream numbers ?
>
> In this instance, if the files are the ones you have shown earlier,
> yes, the order in which FFmpeg detects the streams is the problem.
>
> I suggest you read the documentation of the concat demuxer to the end,
> the answer to your question is there.
>
> Note that VOB files are also concatenable at the container level,
> provided they do not contain extra junk at the beginning or the end (if
> you got them directly from a DVD Video filesystem, they do and you
> should NOT be using them directly).
>
> Regards,
>

Thanks for the help. I've set up a complex filter. That works.

But following up on:

> if you got them directly from a DVD Video filesystem ... you
> should NOT be using them directly

So for vob's from a video dvd :

cat VTS* | ffmpeg -i - ... out.mp4

or

ffmpeg -f concat -i list.of.vobs  .. out.mp4

is not a good idea ? Even though it does not generate any errors.

I should be using a complex filter for any vob's from a video dvd ? Who 
knew ?



___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-user] HW acceleration and ffmpeg

[vpstranger]

So I assume there's no way of making it use HWA automatically via some build 
parameters without getting into the code, right? 
I'm using a video source filter which wraps itself around ffmpeg and I need it 
to do the HWA, which it does not. So if I had an auto-HWA-enabled ffmpeg build, 
I could build that filter against it and HWA would work.

12. Nov 2018 16:30 by harshil.makwana1...@gmail.com 
:


> Hi,
>
> You need to specify hardware accelerator to ffmpeg.
>
> You can give hw acceleration name with 
> -hwaccel option to ffmpeg.
>
> And if you want to use it with any binary then its example code is located at 
> :
>
> doc/examples/hw_decode.c or doc/examples/vaapi_encode.c or
> doc/examples/vaapi_decode.c file
>
> You can compile and use its binary.
>
> Thanks 
> Harshil
>
>> On 12-Nov-2018, at 6:11 PM, <>> vpstran...@tutanota.com 
>> >> > <>> vpstran...@tutanota.com 
>> >> > wrote:
>>
>> Hello! 
>> I have a couple of questions about ffmpeg tool. 
>> When I use the ffmpeg tool that was compiled with support of one of the 
>> hardware accelerators, will the tool automatically use this acceleration 
>> while working? Or it has to be explicitly specified somehow that it must be 
>> used? 
>> And if it attempts to use it, how do I check whether that attempt was 
>> successful?
>>
>> ___
>> ffmpeg-user mailing list
>> ffmpeg-user@ffmpeg.org 
>> http://ffmpeg.org/mailman/listinfo/ffmpeg-user 
>> 
>>
>> To unsubscribe, visit link above, or email
>> ffmpeg-user-requ...@ffmpeg.org >>  
>> with subject "unsubscribe".
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org 
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user 
> 
>
> To unsubscribe, visit link above, or email
> ffmpeg-user-requ...@ffmpeg.org >  with 
> subject "unsubscribe".
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-user] HW acceleration and ffmpeg

[Harshil Makwana]

Hi,

You need to specify hardware accelerator to ffmpeg.

You can give hw acceleration name with 
-hwaccel option to ffmpeg.

And if you want to use it with any binary then its example code is located at :

doc/examples/hw_decode.c or doc/examples/vaapi_encode.c or
doc/examples/vaapi_decode.c file

You can compile and use its binary.

Thanks 
Harshil

> On 12-Nov-2018, at 6:11 PM,  
>  wrote:
> 
> Hello! 
> I have a couple of questions about ffmpeg tool. 
> When I use the ffmpeg tool that was compiled with support of one of the 
> hardware accelerators, will the tool automatically use this acceleration 
> while working? Or it has to be explicitly specified somehow that it must be 
> used? 
> And if it attempts to use it, how do I check whether that attempt was 
> successful?
> 
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
> 
> To unsubscribe, visit link above, or email
> ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".

[FFmpeg-user] HW acceleration and ffmpeg

[vpstranger]

Hello! 
I have a couple of questions about ffmpeg tool. 
When I use the ffmpeg tool that was compiled with support of one of the 
hardware accelerators, will the tool automatically use this acceleration while 
working? Or it has to be explicitly specified somehow that it must be used? 
And if it attempts to use it, how do I check whether that attempt was 
successful?

___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-user] MOV: "undetermined" as audio track language writes "eng"

[Peter B.]

On 09/11/18 16:09, Carl Eugen Hoyos wrote:
> 2018-11-09 13:21 GMT+01:00, Peter B. :
>> [...]
>> https://trac.ffmpeg.org/ticket/6103
>>
>> [...]
>> Shall I post my findings there?
> The ticket does look incomplete.

Does it look better now? ;D
(Posted commandline + output to the ticket)


Thanks in advance,
Peter
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-user] HD MXF SMPTE ST377 Standard Compliance Problem with multiple IndexTableSegments carring Unique ID twins (maybe a bug)

[Martin Vignali]

Hello,

I think this report is interesting.
Creating file for broadcast delivery, is not just about having a file which
play or doesn't play
It's mainly about passing Quality Control.

Even if a file can be play in most of software, if it doesn't pass
broadcaster QC, it will be reject.

In my use i always remux MXF generate by ffmpeg using bmxtranswrap, to have
"clean" mxf.

Mxf is a very complicated format, with very expensive "documentations".
Reason why, few people can improve it inside ffmpeg.

As suggested by Marton Balint, some developers (not me), have more chance
to work on it, with a financial support.
If you can sponsor it, you can try to contact developper who have recently
improve the mxfenc.

@Chris von Görstinger :
If you remux your file with bmx, does it still show warning in the MXF
Analyser ?

If yes, i suggest to create a bug report on trac.
Ideally, using an ffmpeg generator instead of a file as source.


Martin
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".