Color temperature variations in C-41 depending on exposure?

[Juan Valdenebro]

Hi,

I think there are color temperature variations on a given scene depending on exposure of C-41 color negative films... Grays can become colder or warmer...

Have you found the same results? For example +1 or +2 giving more real colors, closer in warmth to what we see...? Or too blue, too cool results metering at box speed?

Thanks for your comments!

This is being discussed here too:

http://www.apug.org/forums/forum40/80083-portra-160-vc-color-negative-exposure-experts.html

But soon there will be 1000 views there, and apart from RFF member Tim Gray kind and spot on comments, little information has come...

Cheers,

Juan

 

[bmasonoh]

I usually shoot b&w but recently tried shooting Ektar for a change. I wasn't really very happy with how the colors turned out. After a little research I discovered that this film is very finicky with regard to exposure. Underexpose and you're likely to get a cyan cast. Overexpose and you're likely to get oversaturated color. Many describe it as being like slide film. I think I may try some Portra next but may just stick to b&w. I'm curious to hear others experiences.

edited:
I just finished reading some of the thread on APUG. I understand your frustration. Their answer appears to be fix it in post (either digital or analog). I don't have a scientific response to the color shift other than to say I've observed the same.

 

[Finder]

The difference in exposure translates into changes in saturation/contrast. The warmth or coolness is a product of the shooting conditions and the scanning/printing.

Color balance (not white balance) is a post processing quality. Scanners automatically do this, but have a hard time because the subtraction of the orange mask is tricky. When using color enlargers (which I do), you will find each film requires different filtration to color balance to neutral. If I want the image warmer or cooler, I simply change filtration. (Negative films do not have a color balance in the strict sense that slide films do.) There are only a few situations where the scene color temperature is so extreme where one of the color layers is so underexposed that balancing becomes impossible--this cannot be compensated for with exposure as another color layer will be blown out.

Color film are fundamentally different from B&W which is why you have no development control. B&W films are made by creating an emulsion and then figuring out a development which will give a target contrast. Color films have a target development and the emulsion is designed to fit that. Messing with color development can be ugly.

 

[Tim Gray]

Scanners automatically do this, but have a hard time because the subtraction of the orange mask is tricky.

I actually don't think this is true at all. Why would it be? You filter it out the same way you do in the enlarger. You use a filter pack to offset R, G, and respectively to negate the orange mask. You can do the exact same thing in Photoshop, plus more complicated acrobatics.

That being said, I suspect there are interactions between the spectral response of the dyes used in the layers of the film and spectral response of the RGB filters used in the scanner's detector. Not all films use the same dyes, nor do all scanners use the same RGB filter specifications, so the file you end up with can be different depending on what you use. From what I understand (I've never printed it), Cibachrome interacted differently with Kodachrome as compared to E6 films for similar reasons; the matching of the film dyes and the spectral sensitivity of the paper.

I'm not trying to be argumentative, just curious to know where this idea comes from...

 

[Juan Valdenebro]

Interesting points, Tim...

Well, yet I think a scanner can find a way to see if one or two more stops of light generate ON FILM a warmer scene... I'm just not that expert scanning... I just imagined some people wet printing 160VC could have noticed it too, but maybe these days too few wet print color...

Cheers,

Juan

 

[Finder]

It can be hard to separate the blue/yellow layer from the mask. And the mask is not consistent between films. Since the scanner cannot color balance directly from the scanner illuminator unlike with a slide film, it has to make a judgement on color. This can make a film scan a little warm or a little cool.

I also had the pleasure of working with the engineers who designed film scanners. They said how hard it was to design for color negative scans.

 

[Finder]

Interesting points, Tim...

Well, yet I think a scanner can find a way to see if one or two more stops of light generate ON FILM a warmer scene... I'm just not that expert scanning... I just imagined some people wet printing 160VC could have noticed it too, but maybe these days too few wet print color...

Cheers,

Juan

I wet print color negatives, although not 160VC, and I do not see what you describe. All I see is increased contast/saturation.

 

[Tim Gray]

Interesting. Sorry if I'm derailing this thread.

I know the mask is not consistent from film type to film type. I was more thinking of scanning as a positive and doing the rest of the process in Photoshop. Also just trying to understand the process. Ultimately both scanners and RA4 printing depend on the tri-stimulus model of color vision, right? The scanner has certain RGB filters built in, and the paper has certain spectral sensitivities and dyes analogous to them (CMY). How does RA4 paper and printing filters separate the orange mask from the blue/yellow layer differently than from adjusting for it in photoshop?

I'm not disagreeing with you about the difficulty of designing a color neg scanner 🙂 I'm guessing designing RA4 paper is tough too though.

 

[Juan Valdenebro]

Thanks. Finder... Maybe it's all a scanning thing...

But honestly I think I remember with underexposure films used to go a bit blue...

And all the recent talk about differences in color on Ektar... Aren't real on film? Just a scan thing?

I can believe we can filter N-2 and N+3 and we can get an image from them, but yet I can't believe that media gives identical color if we give it x light or 32 times that light... I refuse to believe it's just contrast and saturation... Layers must suffer to some degree at some exposure points affecting balance...

But well, It seems it's almost impossible to be sure about what's on film unless we see our positive way...

Cheers,

Juan

 

[Finder]

Interesting. Sorry if I'm derailing this thread.

I know the mask is not consistent from film type to film type. I was more thinking of scanning as a positive and doing the rest of the process in Photoshop. Also just trying to understand the process. Ultimately both scanners and RA4 printing depend on the tri-stimulus model of color vision, right? The scanner has certain RGB filters built in, and the paper has certain spectral sensitivities and dyes analogous to them (CMY). How does RA4 paper and printing filters separate the orange mask from the blue/yellow layer differently than from adjusting for it in photoshop?

I'm not disagreeing with you about the difficulty of designing a color neg scanner 🙂 I'm guessing designing RA4 paper is tough too though.

I guess I am a little confused. Once in photoshop, the mask is no longer in the equation.

You are right that the scanner and printer compensate for the mask. However, when I am printing, I judge the filtration to be good. When the scanner does it, the firmware makes the judgement.

 

[Tim Gray]

Scan as a positive. Correct for the orange mask in photoshop, then invert. Or invert and then correct for the cyan cast.

Juan - color - the color *should* be the same as long as we are in the linear part of the characteristic curve. It's when we drop out of the linear part that things start to happen. Underexposing one or two stops will probably place a couple tones that were on the linear part of the curve on the toe, shifting the color balance in the shadows. Over exposing will put some tones that were on the toe up in the linear part, and might also push some that were in the linear part up into the shoulder. So color balance will change with exposure.

But if you shot a gray card at -1, 0, and +1, it should be a neutral gray in all the exposures, since it's always on the linear part of the curve.

This is of course as I understand it - I could be wrong.

 

[Juan Valdenebro]

That's a clever exposition of the subject, Tim...

Maybe the scanners can produce some changes where film is yet tonally stable... It's a shame I can't check my test frames with wet prints in any way... I can't believe no pro lab in Barcelona offers wet color printing anymore... I'll forget about it...

Cheers,

Juan

 

[Finder]

Thanks. Finder... Maybe it's all a scanning thing...

But honestly I think I remember with underexposure films used to go a bit blue...

And all the recent talk about differences in color on Ektar... Aren't real on film? Just a scan thing?

I can believe we can filter N-2 and N+3 and we can get an image from them, but yet I can't believe that media gives identical color if we give it x light or 32 times that light... I refuse to believe it's just contrast and saturation... Layers must suffer to some degree at some exposure points affecting balance...

But well, It seems it's almost impossible to be sure about what's on film unless we see our positive way...

Cheers,

Juan

Why does underexposure make the image cooler? is it the situation in which you underexpose or the underexposure itself? I am not sure I know the answer. I do know when I underexpose, it is usually around dusk or twilight when there is a lot of skylight in the shadows.

However, you can do a test. Photograph a step wedge--a series of neutral blocks going from black to white. Is the lighter portion of the wedge a different color than the darker portion? If so, then your hypothesis would be correct.

Thinking about this more, if the contrast of the blue/yellow layer did not equal the other two layers, then you could get a warmer or cooler result based on exposure. You would also end up with a cross-curve where the shadows and highlights would not be the same--cool shadows and warm highlights (the step wedge test above would also show this). My next question would be why not magenta/green or cyan/red shifts with exposure? Why is there a dominance with the blue/yellow layer? All of the cross-curves I have seen have been a result of bad development.

But then you put this in an enlarger and change any changes with a filter pack anyway. Since it is so much easier to change color balance with filtration during printing, why do your want to over- or underexpose your film to do this? Bad exposures have their own problems.


Now about shooting color. Colors in the image are really different under different lighting. If I shoot under sunny, cloudy, or skylight conditions, the colors in the image are very different. I could not color balance the other two condition to match a single one, sunny for example. The colors would simply not match or is I matched one color (red for example), the other colors would be way off. Naturally, some of my images should be warm or cool depending on the conditions. Since color conditions vary greatly and the human eye is awful at making absolute color determinations, how are you going to judge exposure to control warmth if it can be done?

Color balance is done after the fact. Color temperature, thank goodness, has already been fixed by the film manufacturer.

 

[Finder]

Scan as a positive. Correct for the orange mask in photoshop, then invert. Or invert and then correct for the cyan cast.

Have you actually tried this? I have and the results are awful. The main reason is the blue channel is so underexposed. (I recommend you letting your scanner do the work and then tweaking it in photoshop.)

The scanner/enlarger does not underexpose the blue and so comes out with good results.

 

[Juan Valdenebro]

Right, Finder... I'll give myself a break and just overexpose a bit to make sure there's light enough, and then I'll care when printing... Thanks! (The idea of shooting the neutral step wedge was cool!)

Cheers,

Juan

 

[Tim Gray]

Have you actually tried this? I have and the results are awful. The main reason is the blue channel is so underexposed. (I recommend you letting your scanner do the work and then tweaking it in photoshop.)

The scanner/enlarger does not underexpose the blue and so comes out with good results.

I have. It's not an ideal workflow - I agree, but it does work. What I tend to do instead is scan as a 'generic negative' in Vuescan which gets the RGB balance pretty close and then finish off the correction in Photoshop. There is often a residual light blue cast that is easy to remove in Photoshop. Maybe this is what you meant all along. If so, I agree with you.

I prefer doing that to letting Nikonscan (in my case) doing the complete adjustment, thought that usually does a pretty good job. It's just that I will doubtless make some adjustments anyway in Photoshop and I can be a bit more consistent doing more there.

 

[Juan Valdenebro]

Cool and friendly APUG member Mark gave me his address in Colorado and today I sent him my test negatives and he'll kindly do all the scenes' wet color printing... Great! In a few weeks he'll be sending me the prints, and then I'll be able to compare his well filtered wet color printing with the -I believe- more limited prints of my test I got here after pro scanning... I guess no sensor or scanner is as good for color tonal range as film + wet printing... But we'll see! I haven't shot any color roll during these two RFs years, so I guess I'll use some 160VC on my next christmas family trip...

Cheers,

Juan