Page 1 of 1

Blue color cast in darkfield images

Posted: Mon Feb 01, 2016 4:58 pm
by gekko
In a previous post ( viewtopic.php?f=27&t=2365 ). I mentioned that I also get a blue cast in darkfield images. Below I post an image of a bdelloid rotifer taken in darkfield (10x objective, DF stop of phase condenser). My microscope is a Nikon Optiphot with a 50 W, 12 V tungsten-halogen lamp operated at about 11 V; the camera body is an Olympus E-P1. I set the camera's custom white balance using brightfield illumination on a clear area of the slide before taking photographs. [added in Edit]: looking through the microscope eyepiece, I do not notice any blue cast: only the images captured by the camera shows this I would appreciate any further thoughts that you may have on possible reasons for the blue color. Thank you.

Image

[Edit:] In order to justify posting in this section of the forum, I will add below a version of the same image in which I have desuterated the blue/cyan color using Photoshop Elements:

Image

Re: Blue color cast in darkfield images

Posted: Mon Feb 01, 2016 6:42 pm
by 75RR
Is this similar to the post you link to in that you seen one thing and the camera records another?

Re: Blue color cast in darkfield images

Posted: Mon Feb 01, 2016 7:38 pm
by zzffnn
Sorry, I don't have experience on that. You may need to isolate/replace one factor at a time, take a photo and compare.

For example, remove all polarizers, if you could, take a photo. If your analyzer cannot be removed from scope head, then replace that head with a plain head from Nikon model S ($40 shipped from eBay, optical length is the same between Nikon model S/G and Labophot/Optiphot).

Change your camera (use a phone camera) and take another photo.

Or change your light source, use a LED tourch reflected by mirror under condenser. Take another photo.

Also, do you have a blue filter in your scope somewhere?

Re: Blue color cast in darkfield images

Posted: Mon Feb 01, 2016 7:46 pm
by gekko
75RR wrote:Is this similar to the post you link to in that you seen one thing and the camera records another?
Yes, sorry I forgot to say that :( I'll add it now. Thanks!!

Re: Blue color cast in darkfield images

Posted: Mon Feb 01, 2016 8:01 pm
by gekko
zzffnn wrote:Sorry, I don't have experience on that. You may need to isolate/replace one factor at a time, take a photo and compare.

For example, remove all polarizers, if you could, take a photo. If your analyzer cannot be removed from scope head, then replace that head with a plain head from Nikon model S ($40 shipped from eBay, optical length is the same between Nikon model S/G and Labophot/Optiphot).

Change your camera (use a phone camera) and take another photo.

Or change your light source, use a LED tourch reflected by mirror under condenser. Take another photo.

Also, do you have a blue filter in your scope somewhere?
Thank you zzffnn. I do not use a blue filter, and I did not have a polarizer in. In fact the reason I posted this was because, unlike the DIC case, here there were no polarizers. There is no blue cast in brightfield (the light looks "orange" as you would expect when using a halogen source). Thanks for your suggestion: I will try afocal mode using a point-&-shoot camera through the eyepiece and see what I get. I will also try auto white balance as well as other preset WB settings just to see what happens.

Thanks again!

Re: Blue color cast in darkfield images

Posted: Mon Feb 01, 2016 8:27 pm
by zzffnn
Definitely try camera's auto WB first. I use that most of the time, sometimes it is off slightly, but most of the times it is fine. I only use LED though. The following is what a nearly transparent ciliate looks like (though the coleps could be slightly green - I did not look at it under brightfield), under LED and darkfield, through my Nikon G's light path, recorded by Olympus E-PM2's auto WB (without any processing): https://youtu.be/u5D3qsWQpyQ

If your brightfield blank background looks organe and you use that as custom white, then would it possible that your camera adds in some blue hue to produce true white? Then camera remembers (that as custom white) and add additional blue to all subsequent imaging?

Re: Blue color cast in darkfield images

Posted: Wed Feb 03, 2016 1:52 am
by 75RR
I think that a potential cause is that the camera does not see the background when not on AUTO until you tell it to, by either One-Touch White Balance or setting the color temperature manually by Preset White Balance or Custom White Balance.
Until you do it "sees" whatever it saw previously, i.e. it works on the previous setting no matter what we see on the screen.

Re: Blue color cast in darkfield images

Posted: Wed Feb 03, 2016 1:57 am
by gekko
75RR wrote:I think that a potential cause is that the camera does not see the background when not on AUTO until you tell it to, by either One-Touch White Balance or setting the color temperature manually by Preset White Balance or Custom White Balance.
Until you do it "sees" whatever it saw previously, i.e. it works on the previous setting no matter what we see on the screen.
Thank you, 75RR, for your comment. I am embarrassed that I do not quite understand it, though. Would you please explain what you meant? What I do is take custom (= One-Touch) white balance on a clear area of the slide under bright field conditions, then I never change it when I use DIC or DF (nor do I change the lamp voltage). I have not yet had the time to do some tests as suggested by zzffnn. Thanks again!

The reason I don't use auto white balance is that in that case the camera will automatically adjust the white balance for a "typical" scene using an internal algorithm, so if what it is looking at has just red or just green, the white balance will be off because it things there is too much red or green and compensates (I think).

Re: Blue color cast in darkfield images

Posted: Wed Feb 03, 2016 2:10 am
by 75RR
What I do is take custom (= One-Touch) white balance on a clear area of the slide under bright field conditions, then I never change it when I use DIC or DF (nor do I change the lamp voltage).
I think you should take a One-Touch White Balance (not the same as CWB) before every photo except when taking multiple images for stacking, as one can assume there will be no need as conditions will remain the same between each frame.

Re: Blue color cast in darkfield images

Posted: Wed Feb 03, 2016 2:23 am
by gekko
75RR wrote:
What I do is take custom (= One-Touch) white balance on a clear area of the slide under bright field conditions, then I never change it when I use DIC or DF (nor do I change the lamp voltage).
I think you should take a One-Touch White Balance (not the same as CWB) before every photo except when taking multiple images for stacking, as one can assume there will be no need as conditions will remain the same between each frame.
I don't think I have both a CWB and a One-Touch balance on my camera: all I have is custom WB and it is set by pressing the "one-touch" button, and it compensates for the color temperature of the light. At least that is what I know. I don't quite see the point of doing this before every image. The purpose is to adjust the camera WB so that with the light used, white as seen under white light is white, grey is grey, and so on, regardless of what the image is or what the contrast enhancement (DIC, DF, etc.) is. For example, in phase contrast, the background I see through the eyepiece is usually dark reddish brown or purplish, and with my method of adjusting WB, the images come out with the correct color. If I adjust the WB while using phase, then the background will become grey and all the other colors will be changed too, which is not what I see through the microscope. Am I missing something? Sorry for being dense :) .

Re: Blue color cast in darkfield images

Posted: Wed Feb 03, 2016 2:53 am
by 75RR
This is what the E-p2 manual says on page 68

"Color reproduction differs depending on light conditions. For instance, when daylight or tungsten lighting is reflected on white paper, the shade of white produced will be slightly different for each. With a digital camera, white color can be adjusted to reproduce more natural white with a digital processor."
I think using DIC or Phase counts as a different lighting condition.
Additionally, I change the light intensity all the time, which with a tungsten bulb changes the lighting conditions.
You could try it and see. Digital film is cheap!

Re: Blue color cast in darkfield images

Posted: Wed Feb 03, 2016 3:29 am
by gekko
75RR wrote:This is what the E-p2 manual says on page 68

"Color reproduction differs depending on light conditions. For instance, when daylight or tungsten lighting is reflected on white paper, the shade of white produced will be slightly different for each. With a digital camera, white color can be adjusted to reproduce more natural white with a digital processor."
I think using DIC or Phase counts as a different lighting condition.
Additionally, I change the light intensity all the time, which with a tungsten bulb changes the lighting conditions.
You could try it and see. Digital film is cheap!
Thank you. Now I think I understand what you mean. However, I do not want to compensate for what phase does, for example; rather I want the image to reflect what I see through the eyepiece. So I want to see the background of a phase image as purplish not grey. I know very well that, with tungsten, when you change the temperature, the color temperature changes, which is why I set the voltage at about 11 V and keep it there. In other words, I do not want to compensate for the blue cast that I get with DF and DIC (which I usually do in software at present, the aim being not to also alter the 'legitimate' colors in the image, which I think adjusting WB under DF and DIC would), rather I want to know why I get this color cast and try to prevent it from happening if I can. Apologies for my unclear attempt at explaining my aim. Thanks again!

Re: Blue color cast in darkfield images

Posted: Mon Feb 08, 2016 3:37 am
by gekko
gekko wrote:
75RR wrote:
What I do is take custom (= One-Touch) white balance on a clear area of the slide under bright field conditions, then I never change it when I use DIC or DF (nor do I change the lamp voltage).
I think you should take a One-Touch White Balance (not the same as CWB) before every photo except when taking multiple images for stacking, as one can assume there will be no need as conditions will remain the same between each frame.
I don't think I have both a CWB and a One-Touch balance on my camera: all I have is custom WB and it is set by pressing the "one-touch" button, and it compensates for the color temperature of the light. At least that is what I know. I don't quite see the point of doing this before every image. The purpose is to adjust the camera WB so that with the light used, white as seen under white light is white, grey is grey, and so on, regardless of what the image is or what the contrast enhancement (DIC, DF, etc.) is. For example, in phase contrast, the background I see through the eyepiece is usually dark reddish brown or purplish, and with my method of adjusting WB, the images come out with the correct color. If I adjust the WB while using phase, then the background will become grey and all the other colors will be changed too, which is not what I see through the microscope. Am I missing something? Sorry for being dense :) .
75RR, I owe you an apology. In fact my camera has both "custom white balance" and "one touch white balance" just as you said, and as you correctly stated, they are not the same thing. I should have read the instruction manual! What Olympus calls "one touch white balance" is what is usually called manual or custom white balance, which is why I was confused. Again, I apologize.