well, plots look fine, i'll add to git master and see if anyone complains
(i think the profile is fine, just the benchmark isn't).
thanks for profiling!
-jo
On Tue, Apr 9, 2013 at 7:10 AM, johannes hanika <[email protected]> wrote:
> hm. something sounds fishy with those numbers.maybe the reference isn't
> whitebalanced or something similar. iso 100 should really be ~ 30 PSNR in
> that case, denoised or not. probably the profile is fine though :)
>
> -jo
>
>
> On Tue, Apr 9, 2013 at 6:02 AM, Daniel Schury <[email protected]> wrote:
>
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> 4 pictures ISO 100 and one witch each ISO 200 - 1600. Create a HDRI with
>> 3 ISO 100 pictures, safe as reference.pfm. Export one picture with ISO
>> 100 - 1600 each, unprocessed as iso($iso).pfm. Repeat with the formerly
>> created presets, save as iso($iso)_denoised.pfm. Run benchmark.sh
>>
>> Am 08.04.2013 18:56, schrieb johannes hanika:
>> > hm, the curves look okay (variance fails to stabilize for very dark
>> > regions a little, but usually that doesn't matter as much). how did you
>> > create the PSNR numbers?
>> >
>> > j.
>> >
>> >
>> > On Tue, Apr 9, 2013 at 3:59 AM, Daniel Schury <[email protected]
>> > <mailto:[email protected]>> wrote:
>> >
>> > Hi,
>> >
>> > attached my presets for the mentioned camera. benchmark output was
>> > iso100_denoised.pfm : 19.5366
>> > iso100.pfm : 19.5331
>> > iso1600_denoised.pfm : 9.32569
>> > iso1600.pfm : 9.31272
>> > iso200_denoised.pfm : 17.0771
>> > iso200.pfm : 17.0707
>> > iso400_denoised.pfm : 13.0347
>> > iso400.pfm : 13.0239
>> > iso800_denoised.pfm : 10.1878
>> > iso800.pfm : 10.1742
>> > but I'm not sure, if the results are ok. Gimme a hint if a factor of 2-3
>> > ist "the same region". And the psnr is rising not very much.
>> >
>> > If the profiling should be redone, I would happily take tips for better
>> > results.
>> >
>> >
>> >
>> ------------------------------------------------------------------------------
>> > Minimize network downtime and maximize team effectiveness.
>> > Reduce network management and security costs.Learn how to hire
>> > the most talented Cisco Certified professionals. Visit the
>> > Employer Resources Portal
>> > http://www.cisco.com/web/learning/employer_resources/index.html
>> > _______________________________________________
>> > darktable-devel mailing list
>> > [email protected]
>> > <mailto:[email protected]>
>> > https://lists.sourceforge.net/lists/listinfo/darktable-devel
>> >
>> >
>>
>> - --
>> Freundliche Grüße
>> Daniel Schury
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v1.4.12 (GNU/Linux)
>> Comment: Using GnuPG with undefined - http://www.enigmail.net/
>>
>> iEYEARECAAYFAlFjBjYACgkQk7XSfqVgB0oRagCgtbEkhPc4j5+4+6uAbvSgkt1c
>> QkcAn2YzQj1QGQ0QeMycdmpi94n3d39Q
>> =D3KE
>> -----END PGP SIGNATURE-----
>>
>>
>> ------------------------------------------------------------------------------
>> Minimize network downtime and maximize team effectiveness.
>> Reduce network management and security costs.Learn how to hire
>> the most talented Cisco Certified professionals. Visit the
>> Employer Resources Portal
>> http://www.cisco.com/web/learning/employer_resources/index.html
>> _______________________________________________
>> darktable-devel mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/darktable-devel
>>
>
>
------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
darktable-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/darktable-devel
