from memory only ISO movies have the option of adding the range bit to
the ATOM, QuickTime (.mov) does not.
Kevin
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user
To unsubscribe, visit link above, or ema
Sent on the go...
> On 30 Apr 2018, at 18:58, Carl Eugen Hoyos wrote:
>
> 2018-04-30 16:33 GMT+02:00, Kevin Wheatley :
>
>> I note that a similar report was made
>> https://trac.ffmpeg.org/ticket/6865 but that was closed.
>
> Looks like ticket #6866 (which sho
Hi,
I've just built a version of ffmpeg n4.0
from ace829cb45cff530b8a0aed6adf18f329d7a98f6 linked against libvmaf
ba5356cb41cf3b19ca0bb108bd8e86e9da402f94.
I notice when running commands similar to the examples in the FFmpeg
documentation:
ffmpeg -i test2.mov -i test3.mov -lavfi libvmaf -f null
On Fri, Jan 20, 2017 at 5:53 PM, Carl Eugen Hoyos wrote:
> 2017-01-20 16:26 GMT+01:00 DEV - VFXBOAT :
>
>> But, the exr rendered by maya are in linear
>
> What is wrong if they transformed to png by FFmpeg?
>
> Please do not top-post here, Carl Eugen
typically OCIO configuration files involve com
On Fri, Jan 20, 2017 at 10:54 AM, Carl Eugen Hoyos wrote:
> 2017-01-19 3:20 GMT+01:00 Daniel Kim :
>
>> I'm trying to use ffmpeg in our pipeline and integrate it with
>> OCIO for the color space transformations.
>
> Why?
>
> What is wrong with the color space transformation in FFmpeg?
>
> What is
As a partial fix I've submitted a patch here:
https://ffmpeg.org/pipermail/ffmpeg-devel/2017-January/205267.html
That fixes the simple cases.
Kevin
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-user
To u
Answering my own question... I have not tested this but having
manually read the code in the git history I'll try code up a fix for
my theory:
looks like there is a bug in the commit
1a08758e7c4e14a9ea8d2fef6c33ad411b2d3c40 relating to the handling of
ptr in decode_frame after decode_block is call
Hi,
just wondered if this is a known problem, if I take an Open EXR with a
data window != display window:
exrheader /tmp/exr_test.0001.exr
/tmp/exr_test.0001.exr:
file format version: 2, flags 0x0
channels (type chlist):
B, 16-bit floating-point, sampling 1 1
G, 16-bit floating-point, s
Andreas,
instead of using the -colormatrix option try instead using
-vf
scale=in_range=full:in_color_matrix=bt709:out_range=tv:out_color_matrix=bt709
assuming your DPX files are utilising full range encoding
Kevin
___
ffmpeg-user mailing list
ffmpeg
For note DCP's are typically 2.6 gamma encoded X'Y'Z'.
Conversion to Rec 709 RGB might be as simple as linearise to XYZ and
then convert to 709 using the primaries matrix and apply an
appropriate transfer function for your new display environment,
however for perceptual matching you may need to ad
The ITU has the following freely available
https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.1366-0-199802-S!!PDF-E.pdf
so at best tens of hours has 2 bits of BCD
Kevin
___
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
http://ffmpeg.org/mailman/li
On Wed, Dec 16, 2015 at 11:27 PM, Rens Dijkshoorn wrote:
> Currently there is no agreed standard how this
> curve should be defined. The NHK/BBC proposal
> seems to be the most promising
>
> If the input file already has an OETF applied encoding
> with ffmpeg should be straight forward in 10 bit x
what happens is that there is an auto inserted scale filter used to
convert from RGB to YCbCr, the auto inserted defaults/features don't
work for your case you could try explicitly add the scale using
-vf scale=in_range=XXX:out_range=YYY:out_color_matrix=bt709
Replace XXX and YYY with 'full'/'tv'
You might want to use the colr atom then, adding options similar to:
-color_trc bt709 -colorspace bt709 -color_primaries bt709 -movflags +write_colr
just prior to your output filename, obviously you should select the
appropriate options, but those are usually a good starting point.
Kevin
___
On Tue, Jun 9, 2015 at 1:08 PM, Hardik Kanakia wrote:
> Hello All,
>
> Tried using the below command, still the issue of color jump persists from
> dpx (RGB) to prores422 (YUV)
>
> /usr/local/sbin/ffmpeg263/ffmpeg -r 24.00 -i "hahk_c300_400.%06d.dpx"
> -s 1920x1080 -r 24.00 -vcodec prore
You may want to also try -vf scale=out_color_matrix=bt709 instead of
the colormatrix as that might give you better quality as ProRes is 10
bit, in addition to the options Carl suggested to write the color
atom.
Kevin
___
ffmpeg-user mailing list
ffmpeg-u
I like to say there are many logarithmic encodings at least 1 per
camera type it appears somedays...
In terms of having ffmpeg support all the options that appears a
thankless and endless task especially given the need for complex
colour mappings (read typically 3D LUT) between a wide gamut space
I should have paid more attention to your command line, in that case
you are going between the same pixel format and I'm not sure that the
use of the scalar for converting between different colour difference
encodings will actually do anything as I believe it only considers
using the matrices if th
On Thu, May 14, 2015 at 4:49 PM, Kevin Wells wrote:
> Hi Kevin, what would be your suggestion for a possible work-around?
you can control the conversion ffmpeg adds automatically by using an
explicit "-vf scale=out_color_matrix=bt709" when going from RGB to
YCbCr substitute what you need for the
In case nobody realises, the colour differences you are showing, look
like a miss match between the encoding matrix and the decoding matrix
used to perform the RGB<->YCbCr conversion, like it is encoded with BT
709 and decoded with 601.
Kevin
___
ffmpeg-
running the same command here, I get an image that contains ~16 bits
worth of data and it looks 'correct' in Nuke. It also matches a PNG
made with ffmpeg
Kevin
On Mon, May 11, 2015 at 2:50 PM, tim nicholson
wrote:
> On 11/05/15 12:59, tim nicholson wrote:
>
>> Thanks for the confirmation that it
21 matches
Mail list logo