Olympus-OM
[Top] [All Lists]

Re: [OM] ( OM ) Evaluating colour in digital images

Subject: Re: [OM] ( OM ) Evaluating colour in digital images
From: dreammoose <dreammoose@xxxxxxxxx>
Date: Sun, 16 Jun 2002 17:02:34 -0700
Before we even get to the specific problem you ask about, Don't you often look a a print and say "That's not at all what I remember it looking like!"? I have some shots of rolling green hills dotted with trees of different greens. Everything in the scene was shades of green, but the automated prints have no true greens at all. The computer tried to balance it to some standard, be it 18 0ray or a proprietary standard for regular snapshots and failed. When I scan and print these, I have to rely on my memory of what green spring hills in California look like to guide me back to something approaching the scenes I saw. My TOPE 10 entry is of a scene predominantly darker, with some highlights. The autoprinter dutifully brightened it up like it was in the sun.

People who shoot slides (as I mostly did for years), tend to use the slide as reference to what the scene really looked like when printing them. However, both experience, reading al the discussions here and elsewhere about the color, saturation, contrast, etc., etc. characteristics and reading photo magizine film tests with the color reproductions of standard targets shot with different films make it clear that this is not really a reliable standard either. The truth is that there is no way on earth to know just what that scene looked like when the picture was taken. There is no reference color capturing system which is perfect.

I have the luxury of scanning directly from the negatives with all balancing turned off, but still have essentially the same problem you do. Some scans come out looking just the way I remember the scene and others don't match my memory/idea at all. My answer is to try to reproduce the scene in some kind of combo of my memory and what I think looks realistic for the kind of image at hand. Trying to match what's on the screen to a print that's not right anyway sounds crazy making. I've been working on developing the ability to see the potential below the surface in dull looking or weird colored standard prints.

Remember Ansel Adams and many other famous photographers manipulated their images extensively to get the picture the saw in their heads, rather than the imperfect image the science of photography produced unaided by human artistry and imagination. My rule that I commend to you is make it look real and/or make it look good, in whatever combination suits you. In the last 4 shots on <http://home.attbi.com/0.000000E+00dreammoose/wsb/html/view.cgi-photos.html-MerchantID-50215-Publish-t-makestatic-true-skip-10.html> , 3 follow the 'realisitc' side of the equation and one the 'look good' side.All the original prints look crummy.

Moose

Brian Swale wrote:

Hi folk,

I wonder if any of you have faced and dealt with this problem.

When I scan images, they are always from prints, on a flat-bed scanner.

So I have no idea if the print-maker has the colour right. The scanning software and the post-scanning imaging software then throw a few wobblies into the image-making process. Not least of these is during the conversion from whatever patented format the program uses, to jpeg.

Another source of error arises in the colour temperature of the light I use when comparing the print with the on-screen image. Unless I have access to good daylight from a window to the side, I have only an incandescent bulb which emits light that is obviously yellow.

All of these things make it difficult to adjust the light level, the contrast and the colours on the digital image so that it represents the reality at the time of exposure.

What do you use for a white-light source?

Brian

< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >





< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >


<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz