[Bruce Clark]

Hi.

I am doing a course and one of the assignments involves manually setting a white balance and taking stepped images of an 18% grey card. So far so good with setting the white balance but the initial photo of grey card is coming out very patchy with significant colour variation across its surface. In theory, the Red, Blue and Green channels should be equal. I am getting all sorts of variations across the resultant photo. I am using a K-7 with Sigma 105mm Macro. Shooting in RAW using Photoshop Elements and Adobe Camera Raw. Settings are ISO 100 and f8 and manually metering. I have disable the WB Adjustable Range (Menu C 11)

Any suggestions welcome.

[Lowell Goudge]

I have found while photographing documents, that the exposure across the subject is highly variable depending upon the angle of incidence of light on the document and also the angle of the camera to the document. although it looks uniformly lit to the eye, it just is not so uniform across the entire surface as you would like.

I suspect the same happens here.

Care to post some examples?

[Bruce Clark]

I have found while photographing documents, that the exposure across the subject is highly variable depending upon the angle of incidence of light on the document and also the angle of the camera to the document. although it looks uniformly lit to the eye, it just is not so uniform across the entire surface as you would like.

I suspect the same happens here.

Care to post some examples?

I have deleted my last attempts as they were just too bad. I will do another run in the morning, trying to get even light as you suggest. Thanks for your response.

[Lowell Goudge]

I have deleted my last attempts as they were just too bad. I will do another run in the morning, trying to get even light as you suggest. Thanks for your response.

One thing I have learned is that no image is worth deleting, and In general, the best lessons come out of mistakes

[Bruce Clark]

Test photo attached. This is the first shot taken immediately after a manual white balance. Taken with full sun over the right shoulder I can't 100% discount the possibility of reflections but this one is much better than previous attempts Running the cursor over the image I see readings in the order of.

Red 170 - 176
Green 180 - 185
Blue 168-174

ISO 100, F8 at 1/400 sec. On full size image with extreme magnification I can see coloured specks. Digital noise at ISO 100 or merely the individual RGB pixels? How the downsized image fares compared to the original is another thing I suppose. I don't delete very much but some things are a waste of disk space

[Bruce Clark]

Hi again. I have just discovered Adobe Camera Raw is applying a "fix" by default to the images. Damn, its hard to go manual. First the camera screws things around, then ACR has a go. Scrap the lot and start again.

[Bruce Clark]

Back once more and this time I think I have another part of the answer. - In camera sharpening. The in camera sharpening creates noise.
A few images shot with image tone of Natural with all presets set to 0 except for the sharpening present. 0 here actually adds some sharpening and noise. Wind this back one notch below 0 & the indicator turns green and consistently grey grey cards are there for the asking. I have only tried with a few hand held shots today. Obviously I need to do a more rigorous test which I shall attempt tomorrow.

Attached Images

View Picture EXIF

File Date

 

[Digitalis]

One thing I have learned is that no image is worth deleting, and In general, the best lessons come out of mistakes

I agree.

the variations you are seeing in the colour channels can be attributed to Noise as well as what is called metamerism - not all the colour filters used in the bayer array transmit 100% of the light that goes through them. Often the red channel is often one of the most problematic because of Infrared light contamination - because Silicon has a surprisingly high sensitivity to IR - but it isn't as sensitive to Blue and Ultraviolet wavelengths* and there is the strong Green component in sunlight** so a balance has to be struck. A perfect sensor with linear response to visible light coupled with a perfect bayer filter would render middle grey as R=128 G=128 B=128. But AFAIK such a sensor has yet to be made - at least on a commercial scale. And that is why professional photographers use colour calibration software and hardware so these metameric errors can be corrected or controlled.


*which is why Foveon CCD sensors often have such weak blue sensitivity.

**arguably the most common light source