Vittorio Giovara <[email protected]> writes:

>> On 13/05/2011 23:00, Ronald S. Bultje wrote:
>>> Hi,
>>>
>>> On Fri, May 13, 2011 at 3:59 AM, Vittorio Giovara
>>> <[email protected]>  wrote:
>>>> weekly ping :)
>>>> as I explained in the first email this patch is not doing the right thing 
>>>> as
>>>> the "right thing"(tm) would be to actually crop the resulting frame to the
>>>> dimensions specified in the cropping rectangle (tested on the JM); however
>>>> this patch takes a saner approach for the current handling in which the
>>>> "unusual" cropping values are always ignored and don't modify the final
>>>> resolution.
>>> So this breaks fate-h264-interlace-crop, and not for the better.
>>> Before the patch, the values are properly used and the bottom of the
>>> image is cropped off. After the patch, the bottom looks like crap. I
>>> don't think that's right.
>>
>> Ronald
>
> So just to reflect the discussion on the chan, it is fine to reject
> this patch as it breaks FATE.
>
> However I still believe that the behaviour introduced in my patch
> would be much more consistent for different videos: I mean why is it
> fine for a video to crop 24 lines because it is interlaced, but you
> can crop max 14 in case of progressive? Imho it is better and more
> consistent to always crop max 8 lines independent of resolution and
> framerate.

Please explain how limiting to 8 lines could ever make any sense
whatsoever.

-- 
Måns Rullgård
[email protected]
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to