Sharpening in 1.0.0 SVN
Locked
 
|
Hi all,
what happened to sharpening filter in 1.0.0 (SVN 22/06/2009)? It seems to be much more aggressive than in 0.10.0. See http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#5350414477017003490 http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#5350414482636616850 the same image processed with 0.10.0 (first) and 1.0.0 svn (second). Same parameters but result is quite different. Thanks Leonardo -- Leonardo Giordani Tele-Rilevamento Europa - T.R.E. s.r.l. Sensing the planet Via Vittoria Colonna, 7 20149 Milano - Italia tel.: +39.02.4343.121 fax: +39.02.4343.1230 e-mail: leonardo.giordani (at) treuropa.com web: www.treuropa.com _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Re: Sharpening in 1.0.0 SVN
|
|
No one has information about this? I found it really strange.
2009/6/23 Leonardo Giordani <[hidden email]> Hi all, -- Leonardo Giordani Tele-Rilevamento Europa - T.R.E. s.r.l. Sensing the planet Via Vittoria Colonna, 7 20149 Milano - Italia tel.: +39.02.4343.121 fax: +39.02.4343.1230 e-mail: leonardo.giordani (at) treuropa.com web: www.treuropa.com _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
|
|
Nothing has changed from me in code relevant...
Gilles Caulier 2009/6/25 Leonardo Giordani <[hidden email]>: > No one has information about this? I found it really strange. > > 2009/6/23 Leonardo Giordani <[hidden email]> >> >> Hi all, >> >> what happened to sharpening filter in 1.0.0 (SVN 22/06/2009)? It seems to >> be much more aggressive than in 0.10.0. >> See >> >> >> http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#5350414477017003490 >> >> http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#5350414482636616850 >> >> the same image processed with 0.10.0 (first) and 1.0.0 svn (second). Same >> parameters but result is quite different. >> Thanks >> >> Leonardo >> >> -- >> Leonardo Giordani >> >> Tele-Rilevamento Europa - T.R.E. s.r.l. >> Sensing the planet >> Via Vittoria Colonna, 7 >> 20149 Milano - Italia >> tel.: +39.02.4343.121 >> fax: +39.02.4343.1230 >> e-mail: leonardo.giordani (at) treuropa.com >> web: www.treuropa.com > > > > -- > Leonardo Giordani > > Tele-Rilevamento Europa - T.R.E. s.r.l. > Sensing the planet > Via Vittoria Colonna, 7 > 20149 Milano - Italia > tel.: +39.02.4343.121 > fax: +39.02.4343.1230 > e-mail: leonardo.giordani (at) treuropa.com > web: www.treuropa.com > > _______________________________________________ > Digikam-users mailing list > [hidden email] > https://mail.kde.org/mailman/listinfo/digikam-users > > Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
|
|
In reply to this post by Leonardo Giordani
Looking at the original parts of the images, it seems to me that they are not
completely equal - could this be the reason for differing sharpening results? Markus -- Photoblog: http://spring2life.blogspot.com _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Re: Sharpening in 1.0.0 SVN
|
|
They have been processed with the same steps and the same parameters; they are a little shifted because the two processing have been made in two different time moments. Do you refer to this or did you noticed other differences?
2009/6/25 Markus Spring <[hidden email]> Looking at the original parts of the images, it seems to me that they are not -- Leonardo Giordani Tele-Rilevamento Europa - T.R.E. s.r.l. Sensing the planet Via Vittoria Colonna, 7 20149 Milano - Italia tel.: +39.02.4343.121 fax: +39.02.4343.1230 e-mail: leonardo.giordani (at) treuropa.com web: www.treuropa.com _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
|
|
In both cases, save result images and compare side by side in a
separated software. Gilles 2009/6/25 Leonardo Giordani <[hidden email]>: > They have been processed with the same steps and the same parameters; they > are a little shifted because the two processing have been made in two > different time moments. Do you refer to this or did you noticed other > differences? > > 2009/6/25 Markus Spring <[hidden email]> >> >> Looking at the original parts of the images, it seems to me that they are >> not >> completely equal - could this be the reason for differing sharpening >> results? >> >> Markus >> -- >> Photoblog: http://spring2life.blogspot.com >> _______________________________________________ >> Digikam-users mailing list >> [hidden email] >> https://mail.kde.org/mailman/listinfo/digikam-users > > > > -- > Leonardo Giordani > > Tele-Rilevamento Europa - T.R.E. s.r.l. > Sensing the planet > Via Vittoria Colonna, 7 > 20149 Milano - Italia > tel.: +39.02.4343.121 > fax: +39.02.4343.1230 > e-mail: leonardo.giordani (at) treuropa.com > web: www.treuropa.com > > _______________________________________________ > Digikam-users mailing list > [hidden email] > https://mail.kde.org/mailman/listinfo/digikam-users > > Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
|
|
In reply to this post by Leonardo Giordani
Leonardo Giordani schrieb:
> They have been processed with the same steps and the same parameters; > they are a little shifted because the two processing have been made in > two different time moments. Do you refer to this or did you noticed > other differences? The 2nd image seems to show marginally more detail. This is really hard to notice, and I do doubt that this is the reason for the drastic effect the screenshot shows. The Parameters you've chosen of 5/2/,05 seem pretty high to me - I never go beyond 2 for the radius. _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Re: Sharpening in 1.0.0 SVN
|
|
In reply to this post by Leonardo Giordani
I can confirm this, 0.10 results look very different from 1.0.0.
I will check this. Andi On Tuesday 23 June 2009 09:12:38 Leonardo Giordani wrote: > Hi all, > > what happened to sharpening filter in 1.0.0 (SVN 22/06/2009)? It seems to > be much more aggressive than in 0.10.0. > See > > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#535041447701 >7003490 > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#535041448263 >6616850 > > the same image processed with 0.10.0 (first) and 1.0.0 svn (second). Same > parameters but result is quite different. > Thanks > > Leonardo _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Re: Sharpening in 1.0.0 SVN
|
|
I reverted all changes we made this year to the sharpening tool and filter.
Still the same result... weird! I will need to use "git bisect" to find the problem. Andi On Thursday 25 June 2009 13:43:52 Andi Clemens wrote: > I can confirm this, 0.10 results look very different from 1.0.0. > I will check this. > > Andi > > On Tuesday 23 June 2009 09:12:38 Leonardo Giordani wrote: > > Hi all, > > > > what happened to sharpening filter in 1.0.0 (SVN 22/06/2009)? It seems to > > be much more aggressive than in 0.10.0. > > See > > > > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#5350414477 > >01 7003490 > > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#5350414482 > >63 6616850 > > > > the same image processed with 0.10.0 (first) and 1.0.0 svn (second). Same > > parameters but result is quite different. > > Thanks > > > > Leonardo > > _______________________________________________ > Digikam-users mailing list > [hidden email] > https://mail.kde.org/mailman/listinfo/digikam-users _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Re: Sharpening in 1.0.0 SVN
|
|
I found the problem, it appears since we changed our blur method to CImg::blur
(commit #953278). The radius we set in the widget is wrong, I guess it always has been wrong. A radius of 5 in other programs doesn't look like our results in digiKam with old code. So actually the label is wrong: It should have been called "diameter" in the past, instead of "radius". So in order to get the same result, you need to set a radius of 2.5 now. This is not possible with the current widget, we need to accept double values here. I will change the widget to accept double values. Andi On Thursday 25 June 2009 14:02:29 Andi Clemens wrote: > I reverted all changes we made this year to the sharpening tool and filter. > Still the same result... weird! > > I will need to use "git bisect" to find the problem. > > Andi > > On Thursday 25 June 2009 13:43:52 Andi Clemens wrote: > > I can confirm this, 0.10 results look very different from 1.0.0. > > I will check this. > > > > Andi > > > > On Tuesday 23 June 2009 09:12:38 Leonardo Giordani wrote: > > > Hi all, > > > > > > what happened to sharpening filter in 1.0.0 (SVN 22/06/2009)? It seems > > > to be much more aggressive than in 0.10.0. > > > See > > > > > > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#53504144 > > >77 01 7003490 > > > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#53504144 > > >82 63 6616850 > > > > > > the same image processed with 0.10.0 (first) and 1.0.0 svn (second). > > > Same parameters but result is quite different. > > > Thanks > > > > > > Leonardo > > > > _______________________________________________ > > Digikam-users mailing list > > [hidden email] > > https://mail.kde.org/mailman/listinfo/digikam-users > > _______________________________________________ > Digikam-users mailing list > [hidden email] > https://mail.kde.org/mailman/listinfo/digikam-users _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Re: Sharpening in 1.0.0 SVN
|
|
Fixed with commit #987077.
Again: the label was wrong for the old blur method. So what you set in this dialog was never a radius, but a diameter. With this commit you can use double values for the radius. Therefore you need to bisect the old value you used in 0.10, to achieve the same results. Andi On Thursday 25 June 2009 15:42:50 Andi Clemens wrote: > I found the problem, it appears since we changed our blur method to > CImg::blur (commit #953278). > > The radius we set in the widget is wrong, I guess it always has been wrong. > A radius of 5 in other programs doesn't look like our results in digiKam > with old code. > > So actually the label is wrong: It should have been called "diameter" in > the past, instead of "radius". > > So in order to get the same result, you need to set a radius of 2.5 now. > This is not possible with the current widget, we need to accept double > values here. > > I will change the widget to accept double values. > > Andi > > On Thursday 25 June 2009 14:02:29 Andi Clemens wrote: > > I reverted all changes we made this year to the sharpening tool and > > filter. Still the same result... weird! > > > > I will need to use "git bisect" to find the problem. > > > > Andi > > > > On Thursday 25 June 2009 13:43:52 Andi Clemens wrote: > > > I can confirm this, 0.10 results look very different from 1.0.0. > > > I will check this. > > > > > > Andi > > > > > > On Tuesday 23 June 2009 09:12:38 Leonardo Giordani wrote: > > > > Hi all, > > > > > > > > what happened to sharpening filter in 1.0.0 (SVN 22/06/2009)? It > > > > seems to be much more aggressive than in 0.10.0. > > > > See > > > > > > > > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#535041 > > > >44 77 01 7003490 > > > > http://picasaweb.google.it/giordani.leonardo/DigiKamSharpening#535041 > > > >44 82 63 6616850 > > > > > > > > the same image processed with 0.10.0 (first) and 1.0.0 svn (second). > > > > Same parameters but result is quite different. > > > > Thanks > > > > > > > > Leonardo > > > > > > _______________________________________________ > > > Digikam-users mailing list > > > [hidden email] > > > https://mail.kde.org/mailman/listinfo/digikam-users > > > > _______________________________________________ > > Digikam-users mailing list > > [hidden email] > > https://mail.kde.org/mailman/listinfo/digikam-users > > _______________________________________________ > Digikam-users mailing list > [hidden email] > https://mail.kde.org/mailman/listinfo/digikam-users _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
Re: Sharpening in 1.0.0 SVN
|
|
Ok, thank you! As soon as possible I'll check it and report.
Thank you again for the quick fix Leonardo 2009/6/25 Andi Clemens <[hidden email]> Fixed with commit #987077. -- Leonardo Giordani Tele-Rilevamento Europa - T.R.E. s.r.l. Sensing the planet Via Vittoria Colonna, 7 20149 Milano - Italia tel.: +39.02.4343.121 fax: +39.02.4343.1230 e-mail: leonardo.giordani (at) treuropa.com web: www.treuropa.com _______________________________________________ Digikam-users mailing list [hidden email] https://mail.kde.org/mailman/listinfo/digikam-users |
«
Return to digikam-users
|
1 view|%1 views
| Free forum by Nabble | Edit this page |