[FFmpeg-devel] [PATCH] Clarifying the use of the scale filter interlace option

Mark Himsley mark at mdsh.com
Tue Apr 3 10:45:56 CEST 2012

On 03/04/12 07:33, Tim Nicholson wrote:
> On 02/04/12 19:41, Mark Himsley wrote:
>> On 02/04/2012 08:56, Tim Nicholson wrote:
>>> On 31/03/12 12:54, Mark Himsley wrote:
>>>> On 29/03/2012 20:33, Michael Niedermayer wrote:
>>>>> On Thu, Mar 29, 2012 at 08:49:38AM +0100, Tim Nicholson wrote:
>>>>>> Also unlike other cases -1 (auto) is not the default.
>>>>> this could be changed if it works better overall
>>>> I made this point on 25/07/2011, with a patch. It was rejected because
>>>> the progressive videos used in fate tests were always flagged as
>>>> progressive in DV codecs so PSNR tests failed horribly.
>>> Did you mean "flagged as interlaced". Progressive marked as progressive
>>> is surely correct.
>> No. I meant what I said :-)
> Sorry I am still confused....
>> DV is always flagged as interlaced by dvvideo_decode_frame()
> But earlier you said "..were always flagged as
>>>> progressive in DV codecs..."

You're right to be confused. My email of 31/03/12 12:54 was wrong.

I meant "I meant what I though I said"!

DV is always flagged as interlaced when decoded in FFmpeg. Even if you 
though you flagged DV as progressive, it's is flagged as interlaced when 
it's decoded in FFmpeg. I'm not necessarily saying that's wrong, I'm 
just saying that it is how it is.

I think the same is true for some of the DNx formats. The format defines 
whether the the media is flagged as interlaced when decoded.


More information about the ffmpeg-devel mailing list