I was being very silly. The input pix_fmt to the encoder was 'YUV422P';
however, I was setting it to 'YUV420P' :/ I think I was distracting myself
with the color ranges/primaries configuration.
After setting it to the correct pix_fmt, the quality was very poor, so I
altered the qmin/qmax to 2. Voila!
A few of my questions remain:
1. The color characteristics of the DV's AVCodecContext and AVFrame are not
explicitly set; they are all 'not specified'. I tried a MPEG clip from the
same set of test clips and the same result. Is that common for the other
decoders too? Is this considered a bug? If they're not set, should I
explicitly set the values?
2. DV video is bottom field. How does FFmpeg handle this? On decoding can I
transition all the lines down by 1 line to change the order? Is there a
function to perform this operation?
3. Out of interest, can I use SWScale to convert the color primaries/ranges?
Gavin.
From: Gavin Smith <gcs584-at-yahoo....@ffmpeg.org>
To: "This list is about using libavcodec, libavformat, libavutil, libavdevice
and libavfilter." <libav-user@ffmpeg.org>
Sent: Friday, 19 May 2017, 11:24
Subject: Re: [Libav-user] DV50 to M2V Artefacts (Color Ranges)
Carl, I think you have given me something to go on.
I tried both:
ffmpeg -i file.mxf -vcodec libx264 -b:v 1200k ./output.tsffmpeg -i file.mxf
-vcodec mpeg2ideo -b:v 1200k ./output.ts
...and the transcoded clip was fine. I didn't need to lower the default
quantizer. I'll try write a sample application to try uncover my problem.
Gavin.
_______________________________________________
Libav-user mailing list
Libav-user@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/libav-user
