[FFmpeg-devel] [PATCH 10/11] avfilter/vf_ssim: use log10 instead of log()/log(10)
gajjanag at mit.edu
Fri Oct 30 00:04:07 CET 2015
On Thu, Oct 29, 2015 at 2:07 PM, Ganesh Ajjanagadde <gajjanag at mit.edu> wrote:
> On Thu, Oct 29, 2015 at 1:37 PM, Michael Niedermayer
> <michael at niedermayer.cc> wrote:
>> On Thu, Oct 29, 2015 at 12:20:08AM -0400, Ganesh Ajjanagadde wrote:
>>> This is likely more precise and conveys the intent better.
>>> Signed-off-by: Ganesh Ajjanagadde <gajjanagadde at gmail.com>
>>> libavfilter/vf_ssim.c | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>> diff --git a/libavfilter/vf_ssim.c b/libavfilter/vf_ssim.c
>>> index ce1e3db..6b2a8d7 100644
>>> --- a/libavfilter/vf_ssim.c
>>> +++ b/libavfilter/vf_ssim.c
>>> @@ -176,7 +176,7 @@ static float ssim_plane(SSIMDSPContext *dsp,
>>> static double ssim_db(double ssim, double weight)
>>> - return 10 * (log(weight) / log(10) - log(weight - ssim) / log(10));
>>> + return 10 * (log10(weight) - log10(weight - ssim));
>> note, this can be simplified further but thats maybe off topic in
>> relation to switching to log10
> I did note that you can rewrite as log10(weight/(weight-ssim)), but
> avoided it deliberately as I did not know what people want with
> respect to it. Since you brought it up and think it may be good, I
> will change it.
> I personally don't consider it too off topic, and prefer this over
> sending a separate patch for the simplification and dealing with
> another wm4 rant about it.
> Will push all later, once everything is reviewed and ok'ed.
pushed with the further simplification, thanks.
>> Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
>> You can kill me, but you cannot change the truth.
>> ffmpeg-devel mailing list
>> ffmpeg-devel at ffmpeg.org
More information about the ffmpeg-devel