Page 5 of 5

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-06-04T04:57:35-07:00
by NicolasRobidoux
henriwho: I could not get rid of the green tinge (or at least minimize it significantly) with rawtherapee without introducing some other annoyance. My incompetence, probably.
But I tried all the presets, and then modified the most obvious candidates (including Neutral 2).

-----

I know that many people swear by Amaze, but it seemed more "noisy" on this image, and I did not see obvious moire in the apartments image. I'll read the article though. Don't know about the other method.

P.S. Interesting article. I've also read somewhere that with some cameras, vng4 (1 colour per bayer array sensor block) is less artifact prone. Don't remember where. Forum posts about processing raw from specific cameras. Some Nikon model, I believe.

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-06-04T06:48:20-07:00
by NicolasRobidoux
henriwho: IMHO, the top part of the article is not relevant to current DSLRs because there is not a lens that can keep up with current top quality sensors. In other words, the level or sharpness that created the moire in these tests is unattainable in practice.

Have you ever seen moire in full size digital images? I haven't.

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-06-19T04:55:46-07:00
by henrywho
Do you have any tool to demosaic the RAW samples at
http://www.fotopolis.pl/index.php?n=14580&p=1

P.S. The landscape sample seems nice (http://www.dpreview.com/reviews/nikon-d800-d800e/27)

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-06-19T07:31:50-07:00
by NicolasRobidoux
Henry: I'm busy with other things, but the free ones I've used are dcraw, photivo (which I like quite a lot), rawtherapee and darktable. I would guess that recent versions would all read Nikon D800 and D800E?

Several of them had major updates, I believe, around February-March. Linux Mint knows about these recent updates.

Before I use one of them I usually do a search like "google rawtherapee D800E" to find out what people recommend as far as settings go. It's been hit and miss, so I don't recommend following whatever advice is out there blindly, but I like to know what "buttons" I should play with first.

P.S. I also am enclined to leave the test image bank alone for a while at least, unless I have a strong reason not to.

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-06-19T07:38:53-07:00
by NicolasRobidoux
Henry: Yes, the landscape image is really nice. If I was in the mood, I'd ask the author for permission to include it.

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-07-11T05:04:58-07:00
by NicolasRobidoux
Henry:
Bruno George Moraes (Brazilian embedded systems software engineer with an interest in image processing) is really impressed by the main demosaicing method of http://www.my-spot.com/RHC/RHC_Demosiac.htm.

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-07-11T19:29:11-07:00
by henrywho
It looks great! Hope it would support my Panasonic GX1 in some future.

http://www.my-spot.com/RHC/RHC_Cameras.htm

Re: [SOLVED] Donate a high quality raw image to science?

Posted: 2012-07-13T11:26:45-07:00
by Pictus
NicolasRobidoux wrote:Henry:
Bruno George Moraes (Brazilian embedded systems software engineer with an interest in image processing) is really impressed by the main demosaicing method of http://www.my-spot.com/RHC/RHC_Demosiac.htm.
Good links about RAW demosaicing algorithms
http://www.libraw.org/articles/bayer-moire.html
http://www.libraw.org/articles/festina-lente.html