[Digikam-devel] [Bug 137278] New: 16-bit/channel workspace when working with RAW conversions
Jakob Østergaard
joe at evalesco.com
Mon Nov 13 10:48:25 GMT 2006
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
http://bugs.kde.org/show_bug.cgi?id=137278
Summary: 16-bit/channel workspace when working with RAW
conversions
Product: digikamimageplugins
Version: unspecified
Platform: Ubuntu Packages
OS/Version: Linux
Status: UNCONFIRMED
Severity: wishlist
Priority: NOR
Component: Core Plugin
AssignedTo: digikam-devel kde org
ReportedBy: joe evalesco com
Version: (using KDE KDE 3.5.2)
Installed from: Ubuntu Packages
OS: Linux
As a hobby photographer I have the following workflow when I process digital images from my camera:
1) First I copy the RAW files from the CF card
2) Then I run a batch job to convert all RAW files to JPEG
3) Viewing the JPEGs allows me to delete the 90% of the images that are not "good"
4) Then, using ShowPhoto, I take the remaining RAW images, one by one, and adjust colour levels, cropping, noise reduction etc. etc.
However, in step 4, I am losing some of the image quality from the RAW file, because the way ShowPhoto works, it converts the RAW image (which has 10 or 12 bits/channel resolution) to an 8-bit/channel work-space. I then "stretch" colurs and noise-reduce on the 8-bit/channel workspace. The resulting image will have holes in the histogram, if I stretch the curves of the image.
If ShowPhoto used a 16-bit/channel workspace instead, only converting to 8-bit/channel when saving the resulting TIFF, then the resulting image histogram would be without holes (in all the common cases, except when the curves are stretched violently).
I hope the above explains the work process and the problem, and my suggestion for a significant improvement. If this feature request is misplaced (sorry I don't know ShowPhoto/DigiKam internals), please let me know.
So, to sum up, I propose the following change:
ShowPhoto currently does:
1) RAW -> 8-bit/channel conversion
2) Allows user to edit image
3) Saves 8-bit/channel image to resulting file
I propose instead:
1) RAW -> 16-bit/channel conversion
2) Allows user to edit 16-bit/channel image
3) Scales to 8-bit/channel during save to resulting file
If someone can point me in the right direction, I might even have time to see if I can make this change myself (no promises though). Thank you very much, for a very useful tool :)
More information about the Digikam-devel
mailing list