Re: [FFmpeg-user] [FFMPEG] Audio/Video Syncing

[Guillermo Martinez]

Hi Everyone,

Good morning.

I would like to follow-up if there are other solutions I do based on the
logs and information I have provided? Thanks!

Regards,

Guillermo

On 1/15/2016 4:27 PM, Guillermo Martinez wrote:

Hi Carl,

Good afternoon.

Thank you very much for the reply.

On 1/13/2016 9:49 PM, Carl Eugen Hoyos wrote:

Guillermo Martinez <94dlve1izy52e5t  gmail.com> writes:

ffmpeg -y -ihttp://192.168.0.11:8090/VIDEO_CAM1.mpg  
-itsoffset 1.0 -ihttp://192.168.0.11:8090/AUDIO.mp2  
-map 0:0 -map 1:0 test.mp4

Complete, uncut console output missing.
Sorry about that. The following are console outputs from the client 
with and without the "*itsoffset*" option:

==
*1. **ffmpeg -y -i **http://192.168.0.11:8090/VIDEO_CAM1.mpg**-i 
**http://192.168.0.11:8090/AUDIO.mp2**-map 0:0 -map 0:0 -t 15 test.mp4 
2> ffmpeg_client_without_itsoffset.txt*

- without the "itsoffset" option
- 
https://www.dropbox.com/s/hfo2newv7x4o12t/ffmpeg_client_without_itsoffset.txt?dl=0


***2. **ffmpeg -y -i 
**http://192.168.0.11:8090/VIDEO_CAM1.mpg**-itsoffset 1.0 -i 
**http://192.168.0.11:8090/AUDIO.mp2**-map 0:0 -map 0:0 -t 15 test.mp4 
2> ffmpeg_client_with_itsoffset.txt*

- with the "itsoffset" option
- 
https://www.dropbox.com/s/dirfkb3yndwtjni/ffmpeg_client_with_itsoffset.txt?dl=0

==

I would also like to share my "*ffserver.conf*" and "*input ffmpeg 
commands*"

==
*1. ffserver.conf
* https://www.dropbox.com/s/cf3kgyfimq6dgcr/ffserver.conf?dl=0*
*
*2. ffmpeg input commands (to ffserver)*
https://www.dropbox.com/s/mdl5gufzuc59wmz/start.ffmpeg.sh?dl=0
==

Hope this helps in getting further insight.

Unfortunately, I suspect your http sources do not provide timestamps, 
so there is no easy solution;-( 
I see. The stream that seems to always start at *dts=0* is the audio 
stream.


Please refer to the following outputs of ffprobe:
==
*1. ffprobe -i http://192.168.0.11:8090:AUDIO.mp2 -show_frames*
- https://www.dropbox.com/s/xb5t88rc3vdyc0y/ffprobe_audio.txt?dl=0
- pkt_dts and pkt_pts always start at 0

*2. ffprobe -i http://192.168.0.11:8090:VIDEO_CAM1.mpg -show_frames*
- https://www.dropbox.com/s/ywxgznv8xjbqozj/ffprobe_video.txt?dl=0
- pkt_dts and pkt_pts always start non-zero
==

I think that the solution is to add timestamp to the audio stream. But 
I already tried
using the "*use_wallclock_as_timestamp*" option in ffmpeg audio feed. 
Nothing happened.

Is there any other way to do this?

I am very sorry for the inconvenience. I am not much of a programmer, 
but more
of an enthusiast. I am using FFmpeg in my home for personal security 
and surveillance.


Regards,

Guillermo Martinez


Carl Eugen

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






---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] How to pan hozitonally on a video?

[Ryan Williams]

According to http://ffmpeg.org/ffmpeg-filters.html#crop the expression 
variables for t (time in seconds) and n (frame number) are available.

# X distance per frame = (400 - 0) / (10 * 25) = 1.6
# Y distance per frame = (10 - 0) / (10 * 25) = 0.04

I suspect you'll want something similar to the following.

-vf "crop=enable=lt(t\,10):w=1200:h=980, 
crop=enable=between(t\,10\,20):w=1200:h=980:x='(t-10)*1.6':y='(t-10)*.04', 
crop=enable=gte(t\,20):w=1200:h=980:x=400:y=10"

-Original Message-
From: ffmpeg-user [mailto:ffmpeg-user-boun...@ffmpeg.org] On Behalf Of Paul B 
Mahol
Sent: Wednesday, 27 January 2016 05:35
To: FFmpeg user questions
Subject: Re: [FFmpeg-user] How to pan hozitonally on a video?

On 1/26/16, Sub Phil  wrote:
> RE:https://www.ffmpeg.org/ffmpeg-filters.html#zoompan
>
> Hi,
>
> I have a 30 sec video in 1920x1080@25fps
>
> I want to pan/crop with a constant rectangle of 1200x980
>
> It rectangle starts at position (0,0) till t=10 sec than from t=10 sec 
> to t=20sec I wish my rectangle to move at a constant/linear speed from 
> position (0,0) to (400,10)
>
> Than from t=20 sec onwards, rectangle stays at (400,10).
>
> How can I do that??

Using crop filter and some kind of expressions for x and y.
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

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

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Andy Furniss wrote:


One thing that is curious if I convert to 422 interlace aware no
flags then to rgb it looks better that going straight to rgb. flags
will still make it better on the second conversion.


Another observation - if I grab (xwd) a frame produced by mplayer +
ilpack then it looks better than the both the interlaced in my
screenshot - it's as good as the best looking (progressive + flags) one
without the chroma field bleed of course.

mplayer -vf ilpack,scale -noaspect -vo gl ~/snooker-short.ts

I have to use gl as 422 vdpau is broken on my gpu (radeonsi)

Forcing ilpack into dumb (unweighted filter) mode - ilpack=0 produces
the banding/misaligned chroma similar to what's seen on ffmpegs worse
output.


https://drive.google.com/file/d/0BxP5-S1t9VEEcEd0WndFVlFENU0/view?usp=sharing

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

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[liupin.2...@gmail.com]

liupin.2...@gmail.com

From: Andy Furniss
Date: 2016-01-27 02:34
To: ffmpeg-user
Subject: Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?
Carl Eugen Hoyos wrote:
> Andy Furniss  gmail.com> writes:
>
>> After reminding my self recently just how bad ffmpeg yuv -> rgb can
>> look with interlaced
>
> Command line, complete, uncut console output and a sample missing.

It's too soon really I am still thinking about it but until I get more
time here's commands from bash history the source file and a screen of
what I get. On left is interlace aware on right is same converted as
progressive. Zoomed is stationary ball - the moving ball shoes that the
interlaced scale did it's thing OK = no chroma spill between fields
unlike the progressive versions on the right.

png best viewed unscaled = download.

>> (without using extra flags).
>
> Why is this an issue?

Maybe should be default but needs more investigation.

https://drive.google.com/file/d/0BxP5-S1t9VEEUUR0QnVYRU8yczQ/view?usp=sharing

https://drive.google.com/file/d/0BxP5-S1t9VEEVkQ2dkhJVGVwVzg/view?usp=sharing

ffmpeg -i ~/snooker-short.ts -ss 11.2 -sws_flags 
+accurate_rnd+full_chroma_int -vf scale=interl=1 -vframes 1 snk-i-flags.png

ffmpeg -i ~/snooker-short.ts -ss 11.2 -vf scale=interl=1 -vframes 1 
snk-i.png

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

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Paul B Mahol wrote:

On 1/26/16, Andy Furniss  wrote:

Andy Furniss wrote:
Here's the assert - after seeing swscale in the first test I tried
format=rgb24 ...


use copy filter before zscale filter.
Anyway using y4m is wrong as it doesnt have frame metadata as
primaries,matrix coefficients etc.


As I said I did also try a "real" input = the transport stream I linked.


I added support for overriding input params in zscale.


Thank you.


so something like this should work:

ffmpeg -i input.y4m zscale=min=709:range=full,format=gbrp out.png

If it asserts, add copy:

ffmpeg -i input.y4m copy,zscale=min=709:range=full,format=gbrp out.png


This works - it will assert without the copy.

Is interlace aware 420 -> * possible?

I notice in the web help -

https://www.ffmpeg.org/ffmpeg-filters.html#toc-zscale

That there are several options at the end that don't appear in ffmpeg -h 
full.


Earlier I tried the [o]hsub/vsub so I know they are unrecognised -
 maybe they and any others that don't parse should be removed.
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Paul B Mahol]

On 1/26/16, Andy Furniss  wrote:
> Andy Furniss wrote:
> Here's the assert - after seeing swscale in the first test I tried
> format=rgb24 ...

use copy filter before zscale filter.

Anyway using y4m is wrong as it doesnt have frame metadata as
primaries,matrix coefficients etc.

I added support for overriding input params in zscale.

so something like this should work:

ffmpeg -i input.y4m zscale=min=709:range=full,format=gbrp out.png

If it asserts, add copy:

ffmpeg -i input.y4m copy,zscale=min=709:range=full,format=gbrp out.png

This assumes input colorspace matrix is 709 and range is full.
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Andy Furniss wrote:
Here's the assert - after seeing swscale in the first test I tried 
format=rgb24 ...


ffmpeg -loglevel debug -i snk-422i-flags.y4m -vf zscale,format=rgb24 
zimg-422.png
ffmpeg version N-78059-g2e31434 Copyright (c) 2000-2016 the FFmpeg 
developers

  built with gcc 5.3.0 (GCC)
  configuration: --prefix=/usr --disable-doc --enable-gpl 
--enable-libzimg --enable-libvpx --enable-libx265 --enable-libdcadec 
--enable-libmp3lame --enable-libx264

  libavutil  55. 15.100 / 55. 15.100
  libavcodec 57. 22.101 / 57. 22.101
  libavformat57. 23.100 / 57. 23.100
  libavdevice57.  0.100 / 57.  0.100
  libavfilter 6. 27.100 /  6. 27.100
  libswscale  4.  0.100 /  4.  0.100
  libswresample   2.  0.101 /  2.  0.101
  libpostproc54.  0.100 / 54.  0.100
Splitting the commandline.
Reading option '-loglevel' ... matched as option 'loglevel' (set logging 
level) with argument 'debug'.
Reading option '-i' ... matched as input file with argument 
'snk-422i-flags.y4m'.
Reading option '-vf' ... matched as option 'vf' (set video filters) with 
argument 'zscale,format=rgb24'.

Reading option 'zimg-422.png' ... matched as output file.
Finished splitting the commandline.
Parsing a group of options: global .
Applying option loglevel (set logging level) with argument debug.
Successfully parsed a group of options.
Parsing a group of options: input file snk-422i-flags.y4m.
Successfully parsed a group of options.
Opening an input file: snk-422i-flags.y4m.
[yuv4mpegpipe @ 0x1fec0a0] Format yuv4mpegpipe probed with size=2048 and 
score=100
[yuv4mpegpipe @ 0x1fec0a0] Before avformat_find_stream_info() pos: 52 
bytes read:32768 seeks:0

[yuv4mpegpipe @ 0x1fec0a0] All info found
[yuv4mpegpipe @ 0x1fec0a0] After avformat_find_stream_info() pos: 829498 
bytes read:829498 seeks:0 frames:1

Input #0, yuv4mpegpipe, from 'snk-422i-flags.y4m':
  Duration: 00:00:00.04, start: 0.00, bitrate: 165899 kb/s
Stream #0:0, 1, 1/25: Video: rawvideo, 1 reference frame (Y42B / 
0x42323459), yuv422p, 720x576, 1/25, SAR 64:45 DAR 16:9, 25 fps, 25 tbr, 
25 tbn, 25 tbc

Successfully opened the file.
Parsing a group of options: output file zimg-422.png.
Applying option vf (set video filters) with argument zscale,format=rgb24.
Successfully parsed a group of options.
Opening an output file: zimg-422.png.
Successfully opened the file.
detected 4 logical cores
[Parsed_format_1 @ 0x1ff3ec0] compat: called with args=[rgb24]
[Parsed_format_1 @ 0x1ff3ec0] Setting 'pix_fmts' to value 'rgb24'
[graph 0 input from stream 0:0 @ 0x1ff3d40] Setting 'video_size' to 
value '720x576'

[graph 0 input from stream 0:0 @ 0x1ff3d40] Setting 'pix_fmt' to value '4'
[graph 0 input from stream 0:0 @ 0x1ff3d40] Setting 'time_base' to value 
'1/25'
[graph 0 input from stream 0:0 @ 0x1ff3d40] Setting 'pixel_aspect' to 
value '64/45'
[graph 0 input from stream 0:0 @ 0x1ff3d40] Setting 'sws_param' to value 
'flags=2'
[graph 0 input from stream 0:0 @ 0x1ff3d40] Setting 'frame_rate' to 
value '25/1'
[graph 0 input from stream 0:0 @ 0x1ff3d40] w:720 h:576 pixfmt:yuv422p 
tb:1/25 fr:25/1 sar:64/45 sws_param:flags=2
[format @ 0x1ff4720] compat: called with 
args=[rgb24|rgba|rgb48be|rgba64be|pal8|gray|ya8|gray16be|ya16be|monob]
[format @ 0x1ff4720] Setting 'pix_fmts' to value 
'rgb24|rgba|rgb48be|rgba64be|pal8|gray|ya8|gray16be|ya16be|monob'

[auto-inserted scaler 0 @ 0x1ffdfc0] Setting 'flags' to value 'bicubic'
[auto-inserted scaler 0 @ 0x1ffdfc0] w:iw h:ih flags:'bicubic' interl:0
[Parsed_format_1 @ 0x1ff3ec0] auto-inserting filter 'auto-inserted 
scaler 0' between the filter 'Parsed_zscale_0' and the filter 
'Parsed_format_1'
[AVFilterGraph @ 0x1ff2d40] query_formats: 5 queried, 3 merged, 1 
already done, 0 delayed
[Parsed_zscale_0 @ 0x1ff3720] w:720 h:576 fmt:yuv422p sar:64/45 -> w:720 
h:576 fmt:yuv422p sar:64/45
[auto-inserted scaler 0 @ 0x1ffdfc0] w:720 h:576 fmt:yuv422p sar:64/45 
-> w:720 h:576 fmt:rgb24 sar:64/45 flags:0x4

Output #0, image2, to 'zimg-422.png':
  Metadata:
encoder : Lavf57.23.100
Stream #0:0, 0, 1/25: Video: png, 1 reference frame, rgb24, 720x576 
[SAR 64:45 DAR 16:9], 1/25, q=2-31, 200 kb/s, 25 fps, 25 tbn, 25 tbc

Metadata:
  encoder : Lavc57.22.101 png
Stream mapping:
  Stream #0:0 -> #0:0 (rawvideo (native) -> png (native))
Press [q] to stop, [?] for help
cur_dts is invalid (this is harmless if it occurs once at the start per 
stream)
[Parsed_zscale_0 @ 0x1ff3720] w:720 h:576 fmt:yuv422p sar:64/45 -> w:720 
h:576 fmt:yuv422p sar:64/45
ffmpeg: src/zimg/api/zimg.cpp:485: zimg_error_code_e 
zimg_filter_graph_process(const zimg_filter_graph*, const 
zimg_image_buffer_const*, const zimg_image_buffer*, void*, 
zimg_filter_graph_callback, void*, zimg_filter_graph_callback, void*): 
Assertion `(!(src->plane[0].data) || 
reinterpret_cast(src->plane[0].data) % zimg::ALIGNMENT == 0) 
&& ("pointer not aligned")' failed.


___
ffmpeg-user mailing list
ffmp

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Paul B Mahol wrote:

On 1/26/16, Andy Furniss  wrote:

Paul B Mahol wrote:

On 1/26/16, Andy Furniss  wrote:



Maybe there is some way - or maybe not, I thought I would ask
before doing my usual ffmpeg game of trying many many
options/variants  in the hope that eventually something will work
:-)


ffmpeg -i yuv -vf zscale,format=gbrp out.png


Thanks there is hope then - doesn't seem to work, but I am using git
z.img perhaps that's not best. I even managed to get it to assert at
one point trying random things :-).

[Parsed_zscale_0 @ 0x2502720] w:720 h:576 fmt:yuv422p sar:64/45 -> w:720
h:576 fmt:gbrp sar:64/45
[Parsed_zscale_0 @ 0x2502720] code 3074: no path between colorspaces


Can you post what you tried but full output not trimmed everything and leave
the most useless info?


The example you gave gives that, I have started trying to add more 
options but not found a working combination yet.


So for now no options =

ffmpeg -loglevel debug -i snk-422i-flags.y4m -vf zscale,format=gbrp 
zimg-422.png
ffmpeg version N-78059-g2e31434 Copyright (c) 2000-2016 the FFmpeg 
developers

  built with gcc 5.3.0 (GCC)
  configuration: --prefix=/usr --disable-doc --enable-gpl 
--enable-libzimg --enable-libvpx --enable-libx265 --enable-libdcadec 
--enable-libmp3lame --enable-libx264

  libavutil  55. 15.100 / 55. 15.100
  libavcodec 57. 22.101 / 57. 22.101
  libavformat57. 23.100 / 57. 23.100
  libavdevice57.  0.100 / 57.  0.100
  libavfilter 6. 27.100 /  6. 27.100
  libswscale  4.  0.100 /  4.  0.100
  libswresample   2.  0.101 /  2.  0.101
  libpostproc54.  0.100 / 54.  0.100
Splitting the commandline.
Reading option '-loglevel' ... matched as option 'loglevel' (set logging 
level) with argument 'debug'.
Reading option '-i' ... matched as input file with argument 
'snk-422i-flags.y4m'.
Reading option '-vf' ... matched as option 'vf' (set video filters) with 
argument 'zscale,format=gbrp'.

Reading option 'zimg-422.png' ... matched as output file.
Finished splitting the commandline.
Parsing a group of options: global .
Applying option loglevel (set logging level) with argument debug.
Successfully parsed a group of options.
Parsing a group of options: input file snk-422i-flags.y4m.
Successfully parsed a group of options.
Opening an input file: snk-422i-flags.y4m.
[yuv4mpegpipe @ 0x24fb0a0] Format yuv4mpegpipe probed with size=2048 and 
score=100
[yuv4mpegpipe @ 0x24fb0a0] Before avformat_find_stream_info() pos: 52 
bytes read:32768 seeks:0

[yuv4mpegpipe @ 0x24fb0a0] All info found
[yuv4mpegpipe @ 0x24fb0a0] After avformat_find_stream_info() pos: 829498 
bytes read:829498 seeks:0 frames:1

Input #0, yuv4mpegpipe, from 'snk-422i-flags.y4m':
  Duration: 00:00:00.04, start: 0.00, bitrate: 165899 kb/s
Stream #0:0, 1, 1/25: Video: rawvideo, 1 reference frame (Y42B / 
0x42323459), yuv422p, 720x576, 1/25, SAR 64:45 DAR 16:9, 25 fps, 25 tbr, 
25 tbn, 25 tbc

Successfully opened the file.
Parsing a group of options: output file zimg-422.png.
Applying option vf (set video filters) with argument zscale,format=gbrp.
Successfully parsed a group of options.
Opening an output file: zimg-422.png.
Successfully opened the file.
detected 4 logical cores
[Parsed_format_1 @ 0x2502ec0] compat: called with args=[gbrp]
[Parsed_format_1 @ 0x2502ec0] Setting 'pix_fmts' to value 'gbrp'
[graph 0 input from stream 0:0 @ 0x2502d40] Setting 'video_size' to 
value '720x576'

[graph 0 input from stream 0:0 @ 0x2502d40] Setting 'pix_fmt' to value '4'
[graph 0 input from stream 0:0 @ 0x2502d40] Setting 'time_base' to value 
'1/25'
[graph 0 input from stream 0:0 @ 0x2502d40] Setting 'pixel_aspect' to 
value '64/45'
[graph 0 input from stream 0:0 @ 0x2502d40] Setting 'sws_param' to value 
'flags=2'
[graph 0 input from stream 0:0 @ 0x2502d40] Setting 'frame_rate' to 
value '25/1'
[graph 0 input from stream 0:0 @ 0x2502d40] w:720 h:576 pixfmt:yuv422p 
tb:1/25 fr:25/1 sar:64/45 sws_param:flags=2
[format @ 0x2503720] compat: called with 
args=[rgb24|rgba|rgb48be|rgba64be|pal8|gray|ya8|gray16be|ya16be|monob]
[format @ 0x2503720] Setting 'pix_fmts' to value 
'rgb24|rgba|rgb48be|rgba64be|pal8|gray|ya8|gray16be|ya16be|monob'

[auto-inserted scaler 0 @ 0x250c5a0] Setting 'flags' to value 'bicubic'
[auto-inserted scaler 0 @ 0x250c5a0] w:iw h:ih flags:'bicubic' interl:0
[format @ 0x2503720] auto-inserting filter 'auto-inserted scaler 0' 
between the filter 'Parsed_format_1' and the filter 'format'
[AVFilterGraph @ 0x2501d40] query_formats: 5 queried, 3 merged, 1 
already done, 0 delayed

[auto-inserted scaler 0 @ 0x250c5a0] picking rgb24 out of 9 ref:gbrp alpha:0
[Parsed_zscale_0 @ 0x2502720] w:720 h:576 fmt:yuv422p sar:64/45 -> w:720 
h:576 fmt:gbrp sar:64/45
[swscaler @ 0x250db00] Forcing full internal H chroma due to input 
having non subsampled chroma
[auto-inserted scaler 0 @ 0x250c5a0] w:720 h:576 fmt:gbrp sar:64/45 -> 
w:720 h:576 fmt:rgb24 sar:64/45 flags:0x4

Output #0, image2, to 'zimg-422.png':
  Metadata:
encod

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Paul B Mahol]

On 1/26/16, Andy Furniss  wrote:
> Paul B Mahol wrote:
>> On 1/26/16, Andy Furniss  wrote:
>
>>> Maybe there is some way - or maybe not, I thought I would ask
>>> before doing my usual ffmpeg game of trying many many
>>> options/variants  in the hope that eventually something will work
>>> :-)
>>
>> ffmpeg -i yuv -vf zscale,format=gbrp out.png
>
> Thanks there is hope then - doesn't seem to work, but I am using git
> z.img perhaps that's not best. I even managed to get it to assert at
> one point trying random things :-).
>
> [Parsed_zscale_0 @ 0x2502720] w:720 h:576 fmt:yuv422p sar:64/45 -> w:720
> h:576 fmt:gbrp sar:64/45
> [Parsed_zscale_0 @ 0x2502720] code 3074: no path between colorspaces

Can you post what you tried but full output not trimmed everything and leave
the most useless info?

I guess you did not specify correct params to zscale, like output
color space and so on.

>
> I did try with 420 output from my command posted in this thread and got
> the same.
>
>
>
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Andy Furniss wrote:

Carl Eugen Hoyos wrote:

Andy Furniss  gmail.com> writes:


After reminding my self recently just how bad ffmpeg yuv -> rgb
can look with interlaced


Command line, complete, uncut console output and a sample missing.


It's too soon really I am still thinking about it but until I get
more time here's commands from bash history the source file and a
screen of what I get. On left is interlace aware on right is same
converted as progressive. Zoomed is stationary ball - the moving ball
shoes that the interlaced scale did it's thing OK = no chroma spill
between fields unlike the progressive versions on the right.

png best viewed unscaled = download.


(without using extra flags).


Why is this an issue?


Maybe should be default but needs more investigation.

https://drive.google.com/file/d/0BxP5-S1t9VEEUUR0QnVYRU8yczQ/view?usp=sharing



https://drive.google.com/file/d/0BxP5-S1t9VEEVkQ2dkhJVGVwVzg/view?usp=sharing



ffmpeg -i ~/snooker-short.ts -ss 11.2 -sws_flags
+accurate_rnd+full_chroma_int -vf scale=interl=1 -vframes 1
snk-i-flags.png

ffmpeg -i ~/snooker-short.ts -ss 11.2 -vf scale=interl=1 -vframes 1
snk-i.png



One thing that is curious if I convert to 422 interlace aware no flags
then to rgb it looks better that going straight to rgb. flags will still
make it better on the second conversion.


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

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Paul B Mahol wrote:

On 1/26/16, Andy Furniss  wrote:



Maybe there is some way - or maybe not, I thought I would ask
before doing my usual ffmpeg game of trying many many
options/variants  in the hope that eventually something will work
:-)


ffmpeg -i yuv -vf zscale,format=gbrp out.png


Thanks there is hope then - doesn't seem to work, but I am using git
z.img perhaps that's not best. I even managed to get it to assert at
one point trying random things :-).

[Parsed_zscale_0 @ 0x2502720] w:720 h:576 fmt:yuv422p sar:64/45 -> w:720 
h:576 fmt:gbrp sar:64/45

[Parsed_zscale_0 @ 0x2502720] code 3074: no path between colorspaces

I did try with 420 output from my command posted in this thread and got 
the same.




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

Re: [FFmpeg-user] Pixel Format 10-bit YUV (v210) and 10-bit RGB (r210)

[Dave Rice]

> On Jan 26, 2016, at 1:56 PM, Joe Volpe  wrote:
> 
> Hi,
> 
> Forgive me if I'm not going through the proper protocol in contacting the
> list users for this forum, I'm new to this site and to FFMPEG.
> I was curious to know if you could please help me out.
> 
> I came across your thread on FFmpeg User. I see that it is dated roughly 3
> years ago.
> 
> Did you ever find a solution to your problem mentioned in the thread?  I'm
> attempting to do something similar and capture from tape through a black
> magic deckling 4k pro card into a 10-bit YUV (v210).

See the documentation here http://ffmpeg.org/ffmpeg-devices.html#Options-3 
, particularly the bm_v210 
option.
Dave Rice
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

[FFmpeg-user] Pixel Format 10-bit YUV (v210) and 10-bit RGB (r210)

[Joe Volpe]

Hi,

Forgive me if I'm not going through the proper protocol in contacting the
list users for this forum, I'm new to this site and to FFMPEG.
I was curious to know if you could please help me out.

I came across your thread on FFmpeg User. I see that it is dated roughly 3
years ago.

Did you ever find a solution to your problem mentioned in the thread?  I'm
attempting to do something similar and capture from tape through a black
magic deckling 4k pro card into a 10-bit YUV (v210).

Any help you can provide would be awesome!

Thanks so much,

Joe

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

Re: [FFmpeg-user] x265 on osx

[Jim Worrall]

> On 2016 Jan 19, at 7:52 PM, G A  wrote:
> 
> i probably made the mistake of installing x265 lib via homebrew, ugh, i’m 
> transcoding some video
> and get this from the display
> 
> x265 [info]: HEVC encoder version 1.8+1-5dcc9d3a928c400b
> x265 [info]: build info [Mac OS X][clang 7.0.2][64 bit] 8bit
> 
> so if i compile the lib with 10 and 12-bits will fmpeg recognize this?

Since you got no answer from the experts . . . please try it and let us know.
I would also be interested to see all the commands that lead to a 10-bit 
compilation.
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] How to pan hozitonally on a video?

[Paul B Mahol]

On 1/26/16, Sub Phil  wrote:
> RE:https://www.ffmpeg.org/ffmpeg-filters.html#zoompan
>
> Hi,
>
> I have a 30 sec video in 1920x1080@25fps
>
> I want to pan/crop with a constant rectangle of 1200x980
>
> It rectangle starts at position (0,0) till t=10 sec
> than from t=10 sec to t=20sec I wish my rectangle to move at a
> constant/linear speed from position (0,0) to (400,10)
>
> Than from t=20 sec onwards, rectangle stays at (400,10).
>
> How can I do that??

Using crop filter and some kind of expressions for x and y.
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Carl Eugen Hoyos wrote:

Andy Furniss  gmail.com> writes:


After reminding my self recently just how bad ffmpeg yuv -> rgb can
look with interlaced


Command line, complete, uncut console output and a sample missing.


It's too soon really I am still thinking about it but until I get more
time here's commands from bash history the source file and a screen of
what I get. On left is interlace aware on right is same converted as
progressive. Zoomed is stationary ball - the moving ball shoes that the
interlaced scale did it's thing OK = no chroma spill between fields
unlike the progressive versions on the right.

png best viewed unscaled = download.


(without using extra flags).


Why is this an issue?


Maybe should be default but needs more investigation.

https://drive.google.com/file/d/0BxP5-S1t9VEEUUR0QnVYRU8yczQ/view?usp=sharing

https://drive.google.com/file/d/0BxP5-S1t9VEEVkQ2dkhJVGVwVzg/view?usp=sharing

ffmpeg -i ~/snooker-short.ts -ss 11.2 -sws_flags 
+accurate_rnd+full_chroma_int -vf scale=interl=1 -vframes 1 snk-i-flags.png


ffmpeg -i ~/snooker-short.ts -ss 11.2 -vf scale=interl=1 -vframes 1 
snk-i.png


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

[FFmpeg-user] How to pan hozitonally on a video?

[Sub Phil]

RE:https://www.ffmpeg.org/ffmpeg-filters.html#zoompan

Hi,

I have a 30 sec video in 1920x1080@25fps

I want to pan/crop with a constant rectangle of 1200x980

It rectangle starts at position (0,0) till t=10 sec
than from t=10 sec to t=20sec I wish my rectangle to move at a
constant/linear speed from position (0,0) to (400,10)

Than from t=20 sec onwards, rectangle stays at (400,10).

How can I do that??

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

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Paul B Mahol]

On 1/26/16, Andy Furniss  wrote:
> Moritz Barsnick wrote:
>> On Tue, Jan 26, 2016 at 15:43:45 +, Andy Furniss wrote:
>>> Is it possible via ffmpeg with libzimg to convert yuv -> rgb or is
>>>  zscale just for scaling and the Colorspace bits not
>>> implemented/need some different filter?
>>
>> From looking at the code, it seems to be implemented. (I don't use
>> zimg.) I have no idea how libzimg handles interlaced...
>>
>> Have you tried? Or are you considering whether to integrate libzimg
>> into your build?
>
> Well I got as far as building with the lib doing
>
> ffmpeg -h full | grep zscale -A 100 plus looking at the web help but
> though I can see plenty of options for colorspace "stuff" I couldn't
> really see anything that specified I wanted yuv -> rgb or how to specify
> interlaced.
>
> The github readme does specifically mention yuv -> rgb and the
> doc/ecamples folder mentiones interlaced.
>
> I did a quick ffmpeg -loglevel debug -i yuv -vf zscale out.png in case
> ffmpeg would automagically use it for the transform but it doesn't.
>
> Maybe there is some way - or maybe not, I thought I would ask before
> doing my usual ffmpeg game of trying many many options/variants  in the
> hope that eventually something will work :-)

ffmpeg -i yuv -vf zscale,format=gbrp out.png

for hight bit depth:

ffmpeg -i yuv -vf zscale,format=gbrp16le out.png
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

Moritz Barsnick wrote:

On Tue, Jan 26, 2016 at 15:43:45 +, Andy Furniss wrote:

Is it possible via ffmpeg with libzimg to convert yuv -> rgb or is
 zscale just for scaling and the Colorspace bits not
implemented/need some different filter?


From looking at the code, it seems to be implemented. (I don't use
zimg.) I have no idea how libzimg handles interlaced...

Have you tried? Or are you considering whether to integrate libzimg
into your build?


Well I got as far as building with the lib doing

ffmpeg -h full | grep zscale -A 100 plus looking at the web help but
though I can see plenty of options for colorspace "stuff" I couldn't
really see anything that specified I wanted yuv -> rgb or how to specify
interlaced.

The github readme does specifically mention yuv -> rgb and the
doc/ecamples folder mentiones interlaced.

I did a quick ffmpeg -loglevel debug -i yuv -vf zscale out.png in case
ffmpeg would automagically use it for the transform but it doesn't.

Maybe there is some way - or maybe not, I thought I would ask before
doing my usual ffmpeg game of trying many many options/variants  in the
hope that eventually something will work :-)
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] Compiling ffmpeg with libx264

[Farid Zakaria]

The website has some pretty clear documents on how to do it.

Just add --libx264 and make sure the shared libraries are in a default path
(or you can add to the LD_LIBRARY_PATH)
On 25 Jan 2016 10:21 p.m., "Rayden" 
wrote:

> Is there anyone who could help in this?
>
>
>
> --
> View this message in context:
> http://ffmpeg-users.933282.n4.nabble.com/Compiling-ffmpeg-with-libx264-tp4674076p4674112.html
> Sent from the FFmpeg-users mailing list archive at Nabble.com.
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Carl Eugen Hoyos]

Andy Furniss  gmail.com> writes:

> After reminding my self recently just how bad ffmpeg yuv -> rgb 
> can look with interlaced 

Command line, complete, uncut console output and a sample missing.

> (without using extra flags).

Why is this an issue?

Carl Eugen

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

Re: [FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Moritz Barsnick]

On Tue, Jan 26, 2016 at 15:43:45 +, Andy Furniss wrote:
> Is it possible via ffmpeg with libzimg to convert yuv -> rgb or is
> zscale just for scaling and the Colorspace bits not implemented/need
> some different filter?

From looking at the code, it seems to be implemented. (I don't use
zimg.) I have no idea how libzimg handles interlaced...

Have you tried? Or are you considering whether to integrate libzimg
into your build?

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

[FFmpeg-user] zimg colorspace yuv -> rgb possible?

[Andy Furniss]

After reminding my self recently just how bad ffmpeg yuv -> rgb can look
with interlaced (without using extra flags). I wanted to try something
else -

Is it possible via ffmpeg with libzimg to convert yuv -> rgb or is
zscale just for scaling and the Colorspace bits not implemented/need
some different filter?

TIA

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

Re: [FFmpeg-user] ffmpeg convert to wrong duration using dvd2concat

[Etienne Desautels]

On 2016-01-21, at 05:36, Nicolas George wrote:
> There is so much in the VOB files of a DVD image that anything can go wrong,
> especially with so-called anti-piracy measures. Full-fledged players have
> counter-measures, but not simple tools.

Yes, DVDs are a bit (too) complex. But in my present case there's no 
anti-piracy measure, at least...

>> file 
>> 'subfile,,start,0,end,312870912,,:/home/mmac/iso/mountpoint/VIDEO_TS/VTS_01_1.VOB'
>> duration 00:04:57.734
>> 
>> file 
>> 'subfile,,start,312870912,end,625534976,,:/home/mmac/iso/mountpoint/VIDEO_TS/VTS_01_1.VOB'
>> duration 00:04:57.433
> 
> I suggest you extract the given segments of this file into temp files and
> try to decode them to see if they match the declared duration.

I did some tests. If I extract the stream directly from de .VOB or from the 
.iso, there's no problem. Output matches the declared duration. But if I 
extract it using concat, that gives me many buffer underflow errors and then 
the time jumps and duration is wrong on playback, stall on 1 frame for the 
added duration. If I extract the stream to .mkv instead of .mpg that gives me 
the same results with the exception that instead of the underflow errors I got 
this warning: Starting new cluster due to timestamp.

To MPEG from .VOB

ffmpeg -i iso/mountpoint/VIDEO_TS/VTS_01_1.VOB -map 0:1 -c:v copy out.mpg
ffmpeg version N-75845-g079d553 Copyright (c) 2000-2015 the FFmpeg developers
  built with gcc 4.4.3 (Ubuntu 4.4.3-4ubuntu5.1)
  configuration: --enable-gpl --enable-libass --enable-libfdk-aac 
--enable-libfreetype --enable-libmp3lame --enable-libopus --enable-libtheora 
--enable-libvorbis --enable-libvpx --enable-libx264 --enable-nonfree
  libavutil  55.  2.100 / 55.  2.100
  libavcodec 57.  4.100 / 57.  4.100
  libavformat57.  3.100 / 57.  3.100
  libavdevice57.  0.100 / 57.  0.100
  libavfilter 6. 10.100 /  6. 10.100
  libswscale  4.  0.100 /  4.  0.100
  libswresample   2.  0.100 /  2.  0.100
  libpostproc54.  0.100 / 54.  0.100
Input #0, mpeg, from 'iso/mountpoint/VIDEO_TS/VTS_01_1.VOB':
  Duration: 00:09:55.14, start: 0.228411, bitrate: 8408 kb/s
Stream #0:0[0x1bf]: Data: dvd_nav_packet
Stream #0:1[0x1e0]: Video: mpeg2video (Main), yuv420p(tv), 720x480 [SAR 8:9 
DAR 4:3], 8000 kb/s, 29.97 fps, 29.97 tbr, 90k tbn, 59.94 tbc
Stream #0:2[0x80]: Audio: ac3, 48000 Hz, stereo, fltp, 256 kb/s
Output #0, mpeg, to 'out.mpg':
  Metadata:
encoder : Lavf57.3.100
Stream #0:0: Video: mpeg2video, yuv420p, 720x480 [SAR 8:9 DAR 4:3], q=2-31, 
8000 kb/s, 29.97 fps, 29.97 tbr, 90k tbn, 29.97 tbc
Stream mapping:
  Stream #0:1 -> #0:0 (copy)
Press [q] to stop, [?] for help
frame= 2970 fps=0.0 q=-1.0 size=   97136kB time=00:01:39.06 bitrate=8032.4kbits
frame= 6505 fps=6504 q=-1.0 size=  212748kB time=00:03:37.01 bitrate=8030.9kbits
frame=10043 fps=6694 q=-1.0 size=  328476kB time=00:05:35.06 bitrate=8030.8kbits
frame=13604 fps=6801 q=-1.0 size=  444964kB time=00:07:33.88 bitrate=8031.0kbits
frame=17194 fps=6876 q=-1.0 size=  562380kB time=00:09:33.67 bitrate=8030.7kbits
frame=17837 fps=6875 q=-1.0 Lsize=  583416kB time=00:09:55.12 
bitrate=8030.8kbits/s
video:581203kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB 
muxing overhead: 0.380809%

To MPEG from concat
--
ffmpeg -f concat -safe 0 -i concat.txt -map 0:0 -c:v copy out.mpg
ffmpeg version N-75845-g079d553 Copyright (c) 2000-2015 the FFmpeg developers
  built with gcc 4.4.3 (Ubuntu 4.4.3-4ubuntu5.1)
  configuration: --enable-gpl --enable-libass --enable-libfdk-aac 
--enable-libfreetype --enable-libmp3lame --enable-libopus --enable-libtheora 
--enable-libvorbis --enable-libvpx --enable-libx264 --enable-nonfree
  libavutil  55.  2.100 / 55.  2.100
  libavcodec 57.  4.100 / 57.  4.100
  libavformat57.  3.100 / 57.  3.100
  libavdevice57.  0.100 / 57.  0.100
  libavfilter 6. 10.100 /  6. 10.100
  libswscale  4.  0.100 /  4.  0.100
  libswresample   2.  0.100 /  2.  0.100
  libpostproc54.  0.100 / 54.  0.100
Input #0, concat, from 'concat.txt':
  Duration: 00:09:55.17, start: 0.228411, bitrate: 0 kb/s
Stream #0:0: Video: mpeg2video (Main), yuv420p(tv), 720x480 [SAR 8:9 DAR 
4:3], 8000 kb/s, 29.97 fps, 29.97 tbr, 90k tbn, 59.94 tbc
Stream #0:1: Audio: ac3, 48000 Hz, stereo, fltp, 256 kb/s
Output #0, mpeg, to 'out.mpg':
  Metadata:
encoder : Lavf57.3.100
Stream #0:0: Video: mpeg2video, yuv420p, 720x480 [SAR 8:9 DAR 4:3], q=2-31, 
8000 kb/s, 29.97 fps, 29.97 tbr, 90k tbn, 29.97 tbc
Stream mapping:
  Stream #0:0 -> #0:0 (copy)
Press [q] to stop, [?] for help
frame= 3035 fps=0.0 q=-1.0 size=   99254kB time=00:01:41.23 bitrate=8031.7kbits
frame= 6585 fps=6585 q=-1.0 size=  215376kB time=00:03:39.68 bitrate=8031.3kbits
[mpeg @ 0x35903a0] buffer underflow st=0 bufi=361 size=40148
[mpeg @ 0x35903a0] buffer underflow st=0 bufi=2390 size=40148
[mpeg @ 0x35903a0] buffer unde

Re: [FFmpeg-user] field order detection - numbering framepositions possible?

[Andy Furniss]

Christoph Gerstbauer wrote:

As we know it is possible to detect field orders with ffmpeg:

ffmpeg -i  -vf idet -f rawvideo -y NUL

Example Output:

[Parsed_idet_0 @ 003ea220] Single frame detection: TFF:0 BFF:153
Progressive:0 Undetermined:0 [Parsed_idet_0 @ 003ea220] Multi frame
detection: TFF:0 BFF:153 Progressive:0 Undetermined:0


Is it possible to MARK the framenumbers to its detected field orders?
 Via framenumber or timestamp? Or to search a "known" TFF content for
BFF frames? To force the output to lines like: Found BFF frames at
position X

(https://ffmpeg.org/pipermail/ffmpeg-user/2014-October/023995.html)


Not sure I would trust idet to call everything right, but then maybe the
fail case I vaguely remember was some artificial/genetated sample.
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] field order detection - numbering framepositions possible?

[Moritz Barsnick]

Hi Christoph,

On Tue, Jan 26, 2016 at 11:36:54 +0100, Christoph Gerstbauer wrote:
> Is it possible to MARK the framenumbers to its detected field orders? 
> Via framenumber or timestamp?
> Or to search a "known" TFF content for BFF frames? To force the output 
> to lines like:
> Found BFF frames at position X

I don't think ffmpeg provides the desired statistics, but you could
extract them yourself.

The idet filter inserts metadata into each frame as documented here:
https://ffmpeg.org/ffmpeg-filters.html#idet

You can use ffprobe to show and analyze the metadata and parse your own
conclusions:

$ ffprobe -f lavfi -i "movie=infile.avi,idet" -show_frames
and use ffprobe arguments to show only the interesting fields (frame
number, tff/bff, statistics) for the desired streams.

Thinking about it, those "cumulative" values may even be the indicators
you are looking for. (Not sure.)

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

Re: [FFmpeg-user] Encoding from video input results in damaged video

[Moritz Barsnick]

On Tue, Jan 26, 2016 at 11:21:37 +0100, Karl Karpfen wrote:
> What I can see during encoding are a lot of these outputs;

Did you even read the mail you replied to?

This may be somewhat useful for analysis for someone who knows the
issue:
> [alsa @ 0x9194c0] ALSA buffer xrun. 1131kB time=00:00:03.16
> bitrate=2922.2kbits/s dup=32 drop=0

But Lou actually requested:
> 2016-01-25 18:58 GMT+01:00 Lou :
> > There is not enough information here to provide any answers. You should
> > include the complete console output from your ffmpeg command. Also, the
> > output of "v4l2-ctl --list-formats-ext" will provide useful info.
> >
> > What player did you test? Does it make a difference if you omit the
> > audio?

That info (command line, console output) is very useful for debug. Also
the other questions Lou asked.

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

Re: [FFmpeg-user] Video filtering for chroma dot crawl and rainbow effect

[Andy Furniss]

Christian Johannesen wrote:

On Monday, January 25, 2016, Andy Furniss 



If you don't have access to better/"real" masters I don't think
you can get good results from this.

What produced this file with height 486?



The file was captured as 720x486 29.97 fps interlaced lower field
first ProRes 422 HQ with an EVS OpenCube server from Beta SP tape.
The resolution is used to retain line 21 data that would not be
stored with a 720x480 capture.


OK, so you need to crop to 480 while it's still 422 and then do an
interlace aware scale to 420 before processing any further.

That's assuming you want to start with 420. I don't know whether or not
all the filters you need will work with 422.

Of course after yadif then you are progressive so if you did work in 422
then the final conversion wouldn't need to be interlace.

I noticed when testing the trac sample (I don't have access to others)
that I needed to specify bff for yadif - just another potential issue to
look out for.
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

[FFmpeg-user] field order detection - numbering framepositions possible?

[Christoph Gerstbauer]

As we know it is possible to detect field orders with ffmpeg:

ffmpeg -i  -vf idet -f rawvideo -y NUL

Example Output:

[Parsed_idet_0 @ 003ea220] Single frame detection: TFF:0 BFF:153 
Progressive:0 Undetermined:0
[Parsed_idet_0 @ 003ea220] Multi frame detection: TFF:0 BFF:153 
Progressive:0 Undetermined:0



Is it possible to MARK the framenumbers to its detected field orders? 
Via framenumber or timestamp?
Or to search a "known" TFF content for BFF frames? To force the output 
to lines like:

Found BFF frames at position X

(https://ffmpeg.org/pipermail/ffmpeg-user/2014-October/023995.html)

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

Re: [FFmpeg-user] Encoding from video input results in damaged video

[Karl Karpfen]

What I can see during encoding are a lot of these outputs;

[alsa @ 0x9194c0] ALSA buffer xrun. 1131kB time=00:00:03.16
bitrate=2922.2kbits/s dup=32 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 2400kB time=00:00:05.34
bitrate=3674.8kbits/s dup=41 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 2951kB time=00:00:05.70
bitrate=4237.3kbits/s dup=44 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 5735kB time=00:00:09.20
bitrate=5101.7kbits/s dup=108 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 5894kB time=00:00:09.50
bitrate=5077.8kbits/s dup=111 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 6167kB time=00:00:09.91
bitrate=5093.5kbits/s dup=117 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 8051kB time=00:00:14.18
bitrate=4648.1kbits/s dup=205 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 8318kB time=00:00:14.91
bitrate=4568.5kbits/s dup=221 drop=0
[alsa @ 0x9194c0] ALSA buffer xrun. 8482kB time=00:00:15.51
bitrate=4478.4kbits/s dup=236 drop=0

They do not appear when there is no signal at video0 (means when the video
source does not give a picture).



2016-01-25 18:58 GMT+01:00 Lou :

> On Mon, Jan 25, 2016, at 08:20 AM, Karl Karpfen wrote:
> [...]
> > Any ideas how to correct this?
>
> There is not enough information here to provide any answers. You should
> include the complete console output from your ffmpeg command. Also, the
> output of "v4l2-ctl --list-formats-ext" will provide useful info.
>
> What player did you test? Does it make a difference if you omit the
> audio?
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] Video filtering for chroma dot crawl and rainbow effect

[Christian Johannesen]

On Monday, January 25, 2016, Andy Furniss  wrote:

> Andy Furniss wrote:
>
>> Andy Furniss wrote:
>>
>>> Carl Eugen Hoyos wrote:
>>>
 Christian Johannesen  gmail.com> writes:

 ffmpeg.exe  -r 3/1001 -i
> D:\media\test_720x486i_dotcrawl.mov
>

 Please remove the input option -r unless you know exactly what
 you are doing.

 -filter_complex [0:0]crop=704:480:8:6,pullup
>

>>> Input on ticket is 4:2:0 and you can't crop vertically interlaced
>>> 4:2:0 by 6 without messing up the chroma. AFAIK you need to crop 4
>>> lines at a time.
>>>
>>> Maybe you could crop 6 if you converted (interlace aware) to 4:2:2
>>> crop and then interlace aware back to 4:2:0 (untested!).
>>>
>>
>> Hmm, after looking at the trac sample I think the chroma has already
>> been scaled progressive = broken sample.
>>
>> I may be wrong! only looked quickly.
>>
>
> Looked again and I still think the "master" is broken. As noted in the
> track it's size may be an issue by its self (as 420 interlaced would
> normally be divisible by 4).
>
> Ignoring that it seems that it has been converted/generated incorrectly
> for 420 interlaced. The chroma has bled between the fields so even
> something simple like yadif=1 alone produces corrupt output.
>
> If you don't have access to better/"real" masters I don't think you can
> get good results from this.
>
> What produced this file with height 486?


The file was captured as 720x486 29.97 fps interlaced lower field first
ProRes 422 HQ with an EVS OpenCube server from Beta SP tape. The resolution
is used to retain line 21 data that would not be stored with a 720x480
capture.


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

Re: [FFmpeg-user] RTMP output - "Past duration too large"

[Charl Wentzel]

With reference to this post:
http://stackoverflow.com/questions/30782771/what-does-past-duration-x-xxx-
too-large-mean

The problem seems to disappear if you set the input frame rate, i.e. use the 
-r flag before your input definition,  e.g.

ffmpeg -r 25 -i /dev/video0 

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

Re: [FFmpeg-user] how to combine resampler? dither and resampl

[Paul B Mahol]

On 1/26/16, Moritz Barsnick  wrote:
> G Alexander, are you "G A"? Why do you change your name? Why does
> hotmail not properly reply to mails? (Why do you top-post?)
>
> On Tue, Jan 26, 2016 at 00:33:34 -0800, G Alexander wrote:
>
>> posting a log is not beneficial, you should be able to abstract the
>> keys and associations without even seeing an output.
>
> If the person trying to help you says it's benficial, then it is!
>
>> eventually it turns out that the documentation is absolutely
>> terrible. you're better off reading the header files like i did. the
>> documentation is pitiful.
>
> The header files won't tell you *anything* about what the original
> poster (is that you?) did incorrectly. He apparently doesn't have the
> correct understanding of the way filters or the ffmpeg command line are
> set up, or he does but made an unfortunate error. *Which we can't tell*
> if we don't see the command line!
>
> Both (how filter arguments are constructed and how the general command
> line is set up) are very well documented, by the way.
>
> If you're so smart, G, you could have told us
> a) what's wrong with this:
>'aresample=resampler=soxr:cheby=1:precision=28=dither_method=shibata'
   ^ this should be :

> b) what this usually points to:
> "Unable to find a suitable output format for"
>
> Go ahead, G, now you tell us.
>
> Moritz
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] how to combine resampler? dither and resampl

[Moritz Barsnick]

G Alexander, are you "G A"? Why do you change your name? Why does
hotmail not properly reply to mails? (Why do you top-post?)

On Tue, Jan 26, 2016 at 00:33:34 -0800, G Alexander wrote:

> posting a log is not beneficial, you should be able to abstract the
> keys and associations without even seeing an output.

If the person trying to help you says it's benficial, then it is!

> eventually it turns out that the documentation is absolutely
> terrible. you’re better off reading the header files like i did. the
> documentation is pitiful.

The header files won't tell you *anything* about what the original
poster (is that you?) did incorrectly. He apparently doesn't have the
correct understanding of the way filters or the ffmpeg command line are
set up, or he does but made an unfortunate error. *Which we can't tell*
if we don't see the command line!

Both (how filter arguments are constructed and how the general command
line is set up) are very well documented, by the way.

If you're so smart, G, you could have told us
a) what's wrong with this:
   'aresample=resampler=soxr:cheby=1:precision=28=dither_method=shibata’
b) what this usually points to:
"Unable to find a suitable output format for"

Go ahead, G, now you tell us.

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

Re: [FFmpeg-user] how to combine resampler? dither and resampl

[Paul B Mahol]

On 1/26/16, G Alexander  wrote:
> posting a log is not beneficial, you should be able to abstract the keys and
> associations without even seeing an output.
>
> eventually it turns out that the documentation is absolutely terrible.
> you're better off reading the header files like i did.  the documentation is
> pitiful.
>

Why are you top-posting?
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user

Re: [FFmpeg-user] how to combine resampler? dither and resampl

[G Alexander]

posting a log is not beneficial, you should be able to abstract the keys and 
associations without even seeing an output.  

eventually it turns out that the documentation is absolutely terrible.  you’re 
better off reading the header files like i did.  the documentation is pitiful.

> Begin forwarded message:
> 
> From: Moritz Barsnick 
> Subject: Re: [FFmpeg-user] how to combine resampler? dither and resample
> Date: January 26, 2016 at 00:06:57 PST
> To: FFmpeg user discussions 
> Reply-To: FFmpeg user questions 
> 
> On Tue, Jan 19, 2016 at 18:42:27 -0800, G A wrote:
>> -filter:a aresample=resampler=soxr
>> -filter:a aresample=dither_method=shibata
>> 
>> i’ve tried these separately and combined, nothing seems to work
>> 
>> i get errors like Unable to find a suitable output format for 
>> 'aresample=resampler=soxr:cheby=1:precision=28=dither_method=shibata’
> 
> Again, you have been told this before:
> Please (always) show us the full command line and the complete, uncut
> console output. All of its information is valuable.
> 
> From the error message, I can "guess" that you have at least two syntax
> errors. If you show us the aforementioned info, I can try to point out
> what you are doing wrong and what you may need to do.
> 
> Furthermore, "nothing seems to work" does not describe the error you
> are seeing. How are we supposed to help? And you never responded to
> answers you got, i.e. whether they helped you or not. That is
> demotivating.
> 
> Moritz
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
> Begin forwarded message:
> 
> From: Moritz Barsnick 
> Subject: Re: [FFmpeg-user] how to combine resampler? dither and resample
> Date: January 26, 2016 at 00:06:57 PST
> To: FFmpeg user discussions 
> Reply-To: FFmpeg user questions 
> 
> On Tue, Jan 19, 2016 at 18:42:27 -0800, G A wrote:
>> -filter:a aresample=resampler=soxr
>> -filter:a aresample=dither_method=shibata
>> 
>> i’ve tried these separately and combined, nothing seems to work
>> 
>> i get errors like Unable to find a suitable output format for 
>> 'aresample=resampler=soxr:cheby=1:precision=28=dither_method=shibata’
> 
> Again, you have been told this before:
> Please (always) show us the full command line and the complete, uncut
> console output. All of its information is valuable.
> 
> From the error message, I can "guess" that you have at least two syntax
> errors. If you show us the aforementioned info, I can try to point out
> what you are doing wrong and what you may need to do.
> 
> Furthermore, "nothing seems to work" does not describe the error you
> are seeing. How are we supposed to help? And you never responded to
> answers you got, i.e. whether they helped you or not. That is
> demotivating.
> 
> Moritz
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user

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

Re: [FFmpeg-user] FFMPEG output to HTTP

[Moritz Barsnick]

On Mon, Jan 25, 2016 at 22:10:17 -0800, Farid Zakaria wrote:
> I'll try to see in code where HTTP method type is hardcoded.

It seems there was once a proposal how to support the PUT method, but
it has never have gone into ffmpeg. (The attachments don't seem to be
in ffmpeg.org's archive, so I'm quoting gmane instead.)

http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/115011

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

Re: [FFmpeg-user] settings?

[G A]

i figured this out days ago.

Sent from somewhere in … The Twilight Zone

> On Jan 25, 2016, at 23:59, Moritz Barsnick  wrote:
> 
> On Tue, Jan 19, 2016 at 19:05:24 -0800, G A wrote:
>> what settings for x265 would you use to convert this old avi?
> 
> I would always use the default setting if in doubt (unless using
> mpeg2video ;-)).
> 
> Honestly, the use of settings depends on what you are trying to
> achieve. You are saying nothing about this. Nor do we know the
> complexity of the input material.
> 
> Video encoding is a complex topic...
> 
> Moritz
> ___
> ffmpeg-user mailing list
> ffmpeg-user@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user

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

Re: [FFmpeg-user] how to combine resampler? dither and resample

[Moritz Barsnick]

On Tue, Jan 19, 2016 at 18:42:27 -0800, G A wrote:
> -filter:a aresample=resampler=soxr
> -filter:a aresample=dither_method=shibata
> 
> i’ve tried these separately and combined, nothing seems to work
> 
> i get errors like Unable to find a suitable output format for 
> 'aresample=resampler=soxr:cheby=1:precision=28=dither_method=shibata’

Again, you have been told this before:
Please (always) show us the full command line and the complete, uncut
console output. All of its information is valuable.

From the error message, I can "guess" that you have at least two syntax
errors. If you show us the aforementioned info, I can try to point out
what you are doing wrong and what you may need to do.

Furthermore, "nothing seems to work" does not describe the error you
are seeing. How are we supposed to help? And you never responded to
answers you got, i.e. whether they helped you or not. That is
demotivating.

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