How to interpret film characteristics data?

[davidfink_photography]

I need some help understanding the color data given in film

characteristics tables such as this: <a

href=http://creekin.net/films.htm>click here</a>. Specifically, do

the data listed for Red, Green, and Blue allow one to judge the

relative color balance of a film? That would seem to be a

reasonable inference, but when I examine the data for slide films in

the table I am confused. For example, comparing Kodak E100G to

E100GX, we find for the former Blue=3.5, Green=3.5, Red=3.5, while

for the latter Blue=3.8, Green=3.6, Red=3.2. Since we know that

E100GX is designed to have a warmer color balance than E100G, we are

led to conclude that a lower value of Red (3.2) vs. Blue (3.8)

results in a warmer balance. BUT......when looking at the listed

values for Provia 100F (Blue=3.6, Green=3.4, Red=3.3), a film with a

noticeable blue balance (at least in my experience), we see that it

too has a lower value of Red than Blue. So....what am I failing

to understand about these issues? Thanks in advance for helpful

information. Best wishes, David

[pdumais]

I don't see that those numbers are of much use for evaluating the color balance. They apparently tell you what the maximum density each layer can achieve. They tell you nothing about the relative sensitivity of each for, say, a "normal" exposure.

[ken_burns1]

We don't see numbers, we see colors. So quit looking at the numbers and shoot test rolls. Compare them on a good light box and see which colors look best. Characteristic curves and logarithms are for scientists, test rolls are for photographers!

[andrew_somerset1]

Those numbers are useless. They're telling you the max density for each layer, which only comes into play in the blacks. That has nothing to do with how the films respond to different colours.

[davidfink_photography]

Thanks Patrick and Andrew,

 

It's good to know that my mystification about interpretation of those color numbers results from asking the wrong question of them! Stupid of me to assume that data published in such a table might actually prove useful. (Then again, perhaps there are folks who *are* interested in density info, and so for them the numbers may have worth.)

 

In any case, thanks again for clarifying this issue for me.

 

Best wishes,

 

David

[kenneth_logan]

Trying interpreting the numbers in context, given at the bottom of the chart: "Maximum values for blue, green, and red are listed as a rough guide to the color saturation of different films." Don't push the numbers beyond the clearly-stated context and they can be quite useful to the practical photographer.

[andrew_somerset1]

Those numbers tell you nothing about saturation. They're the d-max read from the published characteristic curves, not the purity of the colours.

 

That page is full of errors when it comes to what the various numbers mean. Use with caution.

[bill_tuthill]

Andrew, please be specific about errors so I can correct them.

(I'm the guy who maintains that film characteristics table.)

David, when recording 100G and 100GX numbers from the Kodak datasheet,

I remember thinking it was bizarre that 100GX appeared to have

greater blue response. I also posted a comment here on photo.net

saying 100G is the only film I've never seen with RGB curves

so close together, but nobody responded to that.

[davidfink_photography]

Hi Bill,

 

Thanks very much for responding.

 

And I'm glad you appreciate my dilemma when looking at the color numbers for E100G vs. E100GX. Of course it's only a dilemma if those numbers really DO correspond to a film's relative color balance, which you seem to be suggesting that they do.

 

After your response, it looks like I'm back to my original question, then.

 

As you've seen, others in this thread have maintained that the Blue, Green, Red numbers have nothing to do with relative color response.....but rather with density. If so, this would make them considerably less useful, to say the least!, for purposes of judging a film's color balance.

 

I'd love to know definitively what the color numbers actually mean.....and how they are (or are not) meaningful in a practical sense.

 

Thanks again for responding, Bill.

 

Best wishes,

 

David

[andrew_somerset1]

Bill, without going to the page to review it all, I recall that the explanation of the RMS granularity number was incorrect -- something to do with the actual size of the silver grains in the emulsion.

 

RMS granularity is actually simply a measurement of the variability in density in a densitometer trace across an area of even tone. i.e. you shoot a standard target like a grey card and then measure the degree to which density varies from gray on your slide -- so it's a close cousin to a visual assessment of graininess.

 

The saturation/dmax thing is another -- if you really want to look at saturation on a data sheet then the spectral sensitivity curves are probably the place to go, but frankly the best thing to do is shoot some film and look at it.

[bill_tuthill]

Thanks Andrew, I added your definition of (how to obtain) RMS.

David, the spectral dye density graph might be a better place to look

for color density. Years ago my film comparison table showed high

points on the spectral sensitivity curves, but somebody knowledgeable

(can't remember who) told me they don't mean as much as characteristic

curves. However sensitivity curves do show the smoothness of tonal

response. Flat films are probably better for conversion to B&W.

 

For E100G and GX there are three possibilities here: 1) Kodak TechPubs

reversed the 100G and 100GX characteristic curve graphs, 2) 100GX

does not have such a warm color balance when underexposed, and 3) we

are all part of a mass delusion. I vote for #1 because the graph of

supposedly 100G is warmer (less blue) the whole way.

[andrew_somerset1]

I'm not sure that datasheets can really get you to saturation, because saturation is partly a function of the dyes that are used in each layer, and all the charts are really giving you is density information. There's no way to measure how green the green layer is.

 

It would be interesting to hear from someone like Scott Eaton on that.

[davidfink_photography]

Good thoughts, Bill and Andrew both.

 

Yes, it would be interesting to hear from Scott on these issues.

 

I suspect you're correct, Andrew, that color saturation and sensitivity is a more complex matter than can be reduced to a single numeric value for Blue, Green, and Red.

 

On the other hand, as you say Bill, the spectral curves ARE reflecting something distinctive about each emulsion, and theoretically those differences ought to be able to be expressed numerically somehow. If not in a series of single digits, than in some other fashion.

 

On the matter of E100G vs. E100GX, I, too, did think of your theory that Kodak may have inadvertently switched the curves (and resultant numbers) for the two. Wish there were a way to find out.

 

Best wishes,

 

David

[kenneth_logan]

Bill and all, As E100G is currently my film of choice, I'm really interested in getting to the bottom of this issue of the possibility of the charts being switched by Kodak. Yes, I found it interesting that the E100G values for RGB were the same. Any way, does anyone know a better way to query Kodak than to go to their website and send a tech-support-question email?

[dan_sapper]

I'll look into the possibility of the E100G/GX curves being inadvertently switched. Will take several days to get back to you on this.

 

By the way, you can't really judge color balance by relative positioning of the RGB curves on the page, unless the films are very closely related in terms of how the emulsions are designed, spectrally sensitized, and use the same imaging dyes.

 

Bill Tuthill's comment that the curves line up well is worth considering, because linearity of the curves in all three layers will provide neutral color response throughout the tone scale.

[davidfink_photography]

Hi Dan,

 

Thanks for your help on this. Will look forward to what you learn about the possibility of a switch in the E100G/GX curves.

 

Do you think this is plausible, given GX's mission as a warmer balance film? Or is it more likely that G would have the B=G=R graph, as presently represented?

[dan_sapper]

Hi again -

 

Re-reading through all the posts, I'd like to clear up a few things:

 

1. The table in question only lists D-max, which I don't find very useful for discerning any properties about how the films will reproduce color, saturation, neutrality, or spectral response.

 

2. Saturation of a film is best judged using colorimetry. You have to know the colorimetry of the object you are photographing, and then you have to compare the colorimetry of the object as recorded by the film. CIElab color mapping or L*a*b* is what is needed, and for the full range of color space (across the gamut of low-to-high chroma and across all color hues). Neither spectral sensitivity nor characteristic curves, alone or together, will tell you about saturation. If the colorimetry shows the film to produce more saturated colors than the original objects, then we would indicate the film has enhanced saturation.

 

3. I looked carefully at the characteristic curves in the technical publication for E100G and E100GX. I don't believe there should be such a drastic difference in the curves between E100G and E100GX; that is what I'll be asking some colleagues to investigate over the next few days.

 

4. It is typical for slide film characteristic curves to be presented in terms of dye density, status A densitometry. To me, this means the amount of cyan, magenta, and yellow dye in the film. With this in mind, the red curve represents cyan dye, green curve for magenta, and blue curve for yellow dye. You would expect a warm balanced film to produce more yellow dye given the same exposure; this would show as a higher blue curve, as is shown for E100GX as compared to E100G. However, I still don't believe the drastic difference in the shape of the curves at the higher densities, as these two films are more closely related than those curves indicate. Again, will investigate this anomoly.

[davidfink_photography]

Hi Dan,

 

Thank you very much indeed for that information.....it is precisely what was needed to help understand what the table data can and cannot reveal about differences among films.

 

I found your point #2 particularly interesting and helpful:

 

" 2. Saturation of a film is best judged using colorimetry. You have to know the colorimetry of the object you are photographing, and then you have to compare the colorimetry of the object as recorded by the film. CIElab color mapping or L*a*b* is what is needed, and for the full range of color space (across the gamut of low-to-high chroma and across all color hues). Neither spectral sensitivity nor characteristic curves, alone or together, will tell you about saturation. If the colorimetry shows the film to produce more saturated colors than the original objects, then we would indicate the film has enhanced saturation. "

 

To me, the issue of colorimetry begs the question of whether an industry standard could be arrived at.....both in a standardized color target, and in how "saturation" data might be expressed and reported, based on the colorimetric findings.

 

For now, in the absence of a such a standardized procedure, I accept your explanation that the present Dmax data are wholely insufficient to make judgments about color balance/saturation differences across film brands and types.

 

Again, thanks for taking the time to write such a clear and thorough explanation.

 

If there's a follow-up re: the E100G vs. GX issue, I hope you'll report what you learn in that regard.

 

Best wishes,

 

David

[dan_sapper]

Kodak Technical Publication E-4024 will be updated. It turns out that the characteristic curves for E100G and E100GX are 'technically' both correct as originally shown, but are expressed in different mathematical terms. For consistency, the E100G curve will be updated and will look much more like E100GX. This is not a change to the product, just application of a mathematical algorithm to the existing data.

[davidfink_photography]

Hi Dan,

 

Thanks for the update. Interesting that this question may have revealed a calculation inconsistency that Kodak can now fix. If you think of it, can you please let us know when the graphs have been updated in the technical document and on Kodak's site?

 

Thanks again,

 

David

[dan_sapper]

I just checked the <a href="http://www.kodak.com/global/en/professional/support/techPubs/e4024/e4024.jhtml?id=0.1.18.14.7.20.3.24&lc=en">Tech Pub for E100G and E100GX</a> and verified that the characteristic curve for E100G has been updated to be in the same mathematical terms as E100GX. Now you should be able to see that E100GX blue response (yellow dye density) is higher throughout the tone scale in comparison to E100G.

[._._z]

Thanks Dan.