Kodak grey card usage

[Stephen Benskin]

Thanks Stephen, I stand corrected on the specifics.

Where I got that thought was not from reading the standard but from here https://en.wikipedia.org/wiki/Film_speed

Sorry Mark, I'm not intending to nit-pick. It's just that this point has come up a few times lately. The ISO film speed parameters produce good correlation with the fractional gradient speed point using the Delta-X Criterion. It's only for determining the film speed, not for film/paper matching.

 

[markbarendt]

Sorry Mark, I'm not intending to nit-pick. It just that this point has come up a few times lately. The ISO film speed parameters produce good correlation with the fractional gradient speed point using the Delta-X Criterion. It's only for determining the film speed, not for film/paper matching.

No sweat, really thank you.

 

[RobC]

RobC, is that aimed at me???

No it was aimed at anyone who is struggling with with why they cannot get meter readings from different types of meters or comparisons with grey card readings from one setup to another to agree.
I don't use grey cards or incident meters and have no problems. If you know your film and one meter (I opt for a spot meter) then thats all you need for accurate meter readings and tone placement. Given this it seems pointless to be concerend about what appears to be inconsistencies between meters and readings. Simplest by far is throw out the things and processes causing the inconsistencies. ie. grey cards and incident verses reflective comparisons which achieve nothing.
The only thing a grey card is good for in my book is putting a neutral gray in a test shot to get colour balance right or close when analog or digital printing. And with either print medium, neutral grey isn't likely to be best if you have skin tones to worry about so even that is a moot point. It may be important if you need exact colour matching for a clients product packaging or logo or some such, but if you are working to that level then you likely have a colour densitometer to check exact match anyway. A neutral grey may get you close quicker but isn't likely the final colour balance reuiqred since film response is so different from film to film anyway.

 

[RobC]

I think you are normal, Drew. I asked, in one of these threads, who still uses a gray/grey card... and I don't recall a single person responding in the affirmative.

It's amazing it provokes so many responses then.

 

[markbarendt]

Methinks my comments are taken too literally in this comparison of film vs. digital.

If I take a digital photograph, the red line is less contrasty than the curve...I can see that effect, and LR even calls the curve 'moderate contrast' or 'high contrast'. That was the point I was making. If I have a toe and a shoulder, far more things are perceived as 'black' or 'white' than the reality of the tonal level...we have some degree of control in the darkroom over the perceived contrastiness, just as we can do this in digital postprocessing.

I don't care that digital may have its own ISO standard than film...the camera manufacturer engineers the response so that a rated ISO 100 behaves in a manner which appears similar to ISO 100...in the early days, cameras were critiqued because "ISO 100 in the XYZ camera really is ISO 125", for example. The camera manufacturers fixed these differences, so today they behave similarly enough that an old film photographer like me does not have to relearn photography entirely to shoot digital. The more I use digital, the more I perceive it 'MORE LIKE FILM' rather than 'LESS LIKE FILM'. Differences may indeed exist, but materially for day to day purposes, it is 'the same' with many parallels in the two worlds, so that most ordinary photographers care not about the invisible (or hard to notice) differences.

The digital straight line in LR is part of a histogram and not comparable directly to an H&D curve.

 

[markbarendt]

I've pushed the forums limits of discussing digital here enough. Not going any farther.

 

[wiltw]

No it was aimed at anyone who is struggling with with why they cannot get meter readings from different types of meters or comparisons with grey card readings from one setup to another to agree.
I don't use grey cards or incident meters and have no problems. If you know your film and one meter (I opt for a spot meter) then thats all you need for accurate meter readings and tone placement. Given this it seems pointless to be concerend about what appears to be inconsistencies between meters and readings. Simplest by far is throw out the things and processes causing the inconsistencies. ie. grey cards and incident verses reflective comparisons which achieve nothing.
The only thing a grey card is good for in my book is putting a neutral gray in a test shot to get colour balance right or close when analog or digital printing. And with either print medium, neutral grey isn't likely to be best if you have skin tones to worry about so even that is a moot point. It may be important if you need exact colour matching for a clients product packaging or logo or some such, but if you are working to that level then you likely have a colour densitometer to check exact match anyway. A neutral grey may get you close quicker but isn't likely the final colour balance reuiqred since film response is so different from film to film anyway.

I whole heartedly agree. Especially when you look at the examples in Post 71...how you could expect incident to agree all the time with the reflected target, given the variability demonstrated, is unrealistic.

We have all heard about readings from meters being a 'guideline' for exposure, the grey card variability along with the 'flat art' good exposure NOT being consistent with the flat disk incident readings -- whether flat disk is flat to the surface or it is aimed at the light source -- are both demonstrations of "Your readings will vary" When I do shoot a grey card, it is as you say for color neutrality, but for exposure setting I know it gets me close, but not necessarily exactly what I need. And now I know cases in which even the incident meter only 'gets me close'.

 

[tedr1]

I use a gray card

The Kodak reflectance card I use as a target (6.6 stops brightness range) has gray 18% background and patches ranging between white 0.0 and black 2.0. A test frame of this subject was made using Ilford Delta 100 exposed using spotmeter reading of the 18% background of the reflectance card and processed for normal contrast in ID11. I find this neg has a density range approximately 0.3 (= black 2.0) and 1.2 (= white 0.0). Based on these approximate density values is this a well exposed negative? It prints well. It seems to my limited knowledge of densitometry that there is a lot of unused density, film can go to density values of 2 or higher. Is this an underexposed neg? If so by how much? At what density values does the film "shoulder" begin?

 

[Diapositivo]

In this text, authored in 1990 by three distinguished faculty and one professor emirtus of RIT, it never states any differences in numeric ratings of ISO vs. ASA.

I must have expressed myself incorrectly or ambiguously. Mr ISO and Mr ASA are basically twin brothers. It's Mr. DIN who is different from Mr. ISO ;-)

What I wanted to stress, in any case, is that the "speed point" is "m", but the general use of determining the speed of the film is the overall speed. For a certain speed at point "m" there is a certain density at middle grey. Density at point "m" indirecty defines density at middle grey.

 

[Bill Burk]

You probably got 5 stops of range onto the film, with headroom for one more stop of light on the highlights end.

Here's how I figure it.

I have one of those grayscales and mine measures 0.04 to 1.95 which is really 1.91 and then suppose you develop to 0.58 contrast which I consider normal.

You have a density range on negative of 1.2 - 0.3 = 0.9

You can take the 0.9 density range on negative and divide by 0.58 contrast and arrive at the figure of 1.55 which is the subject brightness range corresponding to the negative density range. The difference between what you shot and what you got is 1.91 - 1.55 = 0.36 which would be attributable to flare that brought your shadow density to 0.3 instead of being 0.0 or 0.1 where you thought it would be.

So... This is a classic example which fits the model (though I projected some of my assumptions onto your numbers... you can confirm or counter).

In a standard 7 1/3 stop scene with something in the shade (yours is 6 1/3 all in the main light) the 0.36 flare would bring the subject luminance range at the film plane to 1.84 which would give you a density range on negative at 0.58 ... 1.06... which is all pretty reasonable.

 

[RobC]

I use a gray card

The Kodak reflectance card I use as a target (6.6 stops brightness range) has gray 18% background and patches ranging between white 0.0 and black 2.0. A test frame of this subject was made using Ilford Delta 100 exposed using spotmeter reading of the 18% background of the reflectance card and processed for normal contrast in ID11. I find this neg has a density range approximately 0.3 (= black 2.0) and 1.2 (= white 0.0). Based on these approximate density values is this a well exposed negative? It prints well. It seems to my limited knowledge of densitometry that there is a lot of unused density, film can go to density values of 2 or higher. Is this an underexposed neg? If so by how much? At what density values does the film "shoulder" begin?

There is no hard and fast answer to this question. However, what you should know is that if you are wet printing to B&W paper then the paper requires an approx 1.3log film density range when printed at G2. That will give you a print density range of 2.1log of print reflection. Or to put it in simple terms, at Grade 2 a B&W print gives 7 stops of reflection density(when measured with a densitometer). So in a literal translation of the subject which just fits the paper, you would be looking for a subject which has a subject brightness range of 7 stops which is captured on film in a density range of 1.3log which prints to B&W paper exactly.
Of course these are all ball park values but film exposed at ISO speed and developed using manufacturers standard dev dilution, temp and times will give you fairly close to that.
So your 6.6 stops brightness range is fairly close to that. But the question for you is what print reflection density did the 18% grey of the card produce? Was it 0.75 or was it darker or lighter?

And of course with B&W we are rarely looking for a literal translation from subject to print. well I'm not, others may be.
And if you take the test print you already have and put it on a wall in your house in normal room lighting during the day and then meter it with a spot meter, I think you'll find it only shows you a reflection density of upto 5 stops from black to white unless its in direct sunlight. And in the evening with room lights on it will probably be less, maybe only 3 or 4 stops range. Only if you have it illuminated with a bright spot lamp will it show you something approaching a 7 stop reflection range.

So that suggests that if you want a literal reproduction of subject range in a print on your wall, you should be looking to select a subject with a 5 stop range, develop it to a 1.3log film density range and print that at Grade 2 to give a 2.1log print density range which in normal indoor daylight viewing will actually give about a 5 stop range.

There's an awful lot of variables to consider and it really depends how you like your prints to look and the viewing conditions you show them in. i.e. natural lighting or under artifical light designed to get the maximum reflection/contrast range to show.

If you work back from your expected viewing conditions then you have some realistic targets to aim for rather than working forwards from some arbitrary theoretical numbers which mean nothing without considering the final viewing conditions. Have you ever wondered why galleries put spot lights on all their photos? It increases the visual contrast range of the prints.

I hope that's clearer than mud for you.

p.s. high gloss prints give the best contrast reproduction when viewed on wall. Matt papers and even FB glossy paper diffuse a lot of light and therefore look lower contrast when viewed on your wall (unless they are viewed with spot lighting on them to bring the contrast up)

 

[Diapositivo]

Digital can have a true straight line tone reproduction, where film always has a bit of an s-curve.

Absolutely yes, but even if it is true that digital and film differ at the extreme of their tonal range, they both are supposed to show "middle grey" in the "middle" between white and black, where white is the "whitepoint" and black is the "blackpoint" of the image.
We know that a certain scene will have different blackpoint and whitepoint with digital and analogue. But given a certain whitepoint as "white" and blackpoint as "black" on both media, "middle" grey is always the same "perceptual middle" between black and white.

Given a certain studio scene with a SBR of let's say only 1.5 EV on both sides of middle grey film and digital behave in an extraordinarily similar way in the final result.

Also I would like to stress that just like films are supposed to be developed according to the instructions of the manufacturer in the ISO standard, so raw images can be imagined as developed "according to the instructions of the manufacturer" by just taking the default JPEG produced by the camera. That's nothing else than a standard, or manufacturer prescribed, development.

Differences between digital and film are, in my opinion, not such as to deny the tests shown but I do agree a densitometric analysis of shots of middle grey cards would certainly be extremely interesting. Sadly I have no desitometre.

 

[RobC]

I must have expressed myself incorrectly or ambiguously. Mr ISO and Mr ASA are basically twin brothers. It's Mr. DIN who is different from Mr. ISO ;-)

What I wanted to stress, in any case, is that the "speed point" is "m", but the general use of determining the speed of the film is the overall speed. For a certain speed at point "m" there is a certain density at middle grey. Density at point "m" indirecty defines density at middle grey.

Negatives don't have a middle grey. They have a transparency density. The assumption that will translate to a middle grey in a print is very dependant on how you print it.
Even with slide film it depends on how you project it and the reflctivity of what its projected onto. You may be able to measure it in the slide. Or you may digitize the slide but then its down to your digitizing settings as to what colour you get. So who says the slide density is right? It really doesn't matter. Its what you do with it to view it that really counts.

 

[Diapositivo]

It's about the parameters of the speed determination method of both B&W negative and Color Reversal and the relationship of the different speed points to the metered exposure point. I could explain it but it would take math and math tends to upset people for some reason.

Stephen, I would be grateful if you could "distill" in verbal logic a common English explanation of the math. Give us a "squeeze" of the math, taking only the conceptual juice. Math is demonstration. Give us the conclusions without the demonstration, if need be. Those who want the demonstration of the theorem, will go to the math. I would be content with the enonciation of the theory.

 

[Diapositivo]

You probably got 5 stops of range onto the film, with headroom for one more stop of light on the highlights end.

Here's how I figure it.

I have one of those grayscales and mine measures 0.04 to 1.95 which is really 1.91 and then suppose you develop to 0.58 contrast which I consider normal.

You have a density range on negative of 1.2 - 0.3 = 0.9

You can take the 0.9 density range on negative and divide by 0.58 contrast and arrive at the figure of 1.55 which is the subject brightness range corresponding to the negative density range. The difference between what you shot and what you got is 1.91 - 1.55 = 0.36 which would be attributable to flare that brought your shadow density to 0.3 instead of being 0.0 or 0.1 where you thought it would be.

So... This is a classic example which fits the model (though I projected some of my assumptions onto your numbers... you can confirm or counter).

In a standard 7 1/3 stop scene with something in the shade (yours is 6 1/3 all in the main light) the 0.36 flare would bring the subject luminance range at the film plane to 1.84 which would give you a density range on negative at 0.58 ... 1.06... which is all pretty reasonable.

Bill, if I read you correctly I think that you here, like RobC in another thread, are considering "middle grey" as "middle grey of the scene". Thus, if the scene has a different black and white points, the "middle" on the scene (and on the negative) can fall in points of different densities.

What wiltw and I are interested in is the behaviour of film, and of reflected light meters, when confronted with a certain shade of grey, which let's say is 18%, that we call "middle perceptual grey", regardless of the scene.
To make an extreme example, I might have a scene which is solely composed of something which I know has a 36% reflectivity and I want that 36% shade of grey to be naturally recogniseable in my final image. That means the SBR of the image is actually 0 EV or very near there, if it is enlightened with a flat light. The "middle grey" of the scene is 36%.

If I know that my light meter is calibrated to 18%, I know that, by reading the meter indication, and by exposing for 1 more step, I will obtain 36% grey which is the desired result, in my slide developed according to the well-known standard method prescribed by the manufacturer.

But in order to make that reasoning, I must know in advance to what grey is my light meter calibrated. If my light meter was calibrated, for instance, at an 8% grey, I would have to open more than 2 steps, instead of 1 step, in order to obtain that 36% reflectivity grey of my subject.

So knowing to which grey the light meter is calibrated is inescapable if one wants to use the light meter (the TTL averaging light meter and the spot light meter as well, or especially) and be able to "place" tones in a slide.
(I stress "in a slide" in order to avoid the objection of testing with different development and printing processing variations, and find what "works for you". You cannot do that with slides. Slides either work or don't work).

 

[Diapositivo]

Methinks my comments are taken too literally in this comparison of film vs. digital.

The camera manufacturers fixed these differences, so today they behave similarly enough that an old film photographer like me does not have to relearn photography entirely to shoot digital. T

I would like to support this idea by stressing that light meter manufacturers only know ISO, there's not such a thing as ISO for cameras and ISO for film. You buy a Sekonic or Gossen or Kenko meter, you set the ISO, and you obtain the correct exposure values to obtain this "mystery shade of grey" on your picture regardless of technology used. That is possible because, ultimately, it's shades of grey (or tones of colours) we are talking about. If your final picture gives the intended shade of grey, you have the correct exposure.

 

[Diapositivo]

Negatives don't have a middle grey. They have a transparency density. The assumption that will translate to a middle grey in a print is very dependant on how you print it.
Even with slide film it depends on how you project it and the reflctivity of what its projected onto. You may be able to measure it in the slide. Or you may digitize the slide but then its down to your digitizing settings as to what colour you get. So who says the slide density is right? It really doesn't matter. Its what you do with it to view it that really counts.

Rob, slides have a middle grey, which is the perceptual middle between what in a slide appears white (let's say, the frame of a window) and what in a slide appears black (it might not be black, imagine the shade under a bush, but it's the "blackpoint" of the image, your eye sees it as black).
Whether you see it with a 150W or a 250W lamp, at 2 meters or at 5 meters, even though the absolute amount of light projected on the screen will be different, your "perception" of middle grey will remain the same, because your brain will "center it" between the two extremes of blackpoint and whitepoint which are present on the image.
If white is presente, and black is present, middle grey on the slide must fall where the human vision would expect it to fall.

And, more in general, if I project an image of "middle grey clouds" on the screen, I will perceive them to be the same middle grey both with a 150W lamp or a 250W lamp etc. (the intensity of the light on the screen will be different, but my vision will adjust for that, just like it adjusts for the different light falling on a print).

And If that wasn't the case, how could we go to the cinema and be satisfied by what we see? Things look natural because white looks white, black looks black, and middle look middle, IMO.

 

[RobC]

what you really want to know is that whatever you meter you will end up on whatever negative density or slide density you want it to.

See first post of following topic which is the practical evaluation I use. I'm not saying you should use this methodolgy but I do. You can alter it to use however many stops of range to fit the paper as you like but for me 10 stops of range works fine.
If you're using slide film then assume each "zone" is half a stop instead of 1 stop and you'll be fairly close.

(there was a url link here which no longer exists)

 

[tedr1]

Thank you gentlemen, I'll take that as a no, it is not an underexposed negative. Printed on grade 2 paper the neg reproduces the reflection density range of the original pretty closely. This is basic calibration work to orient myself with a new enlarger and darkroom after a 20 year break.

 

[RobC]

Rob, slides have a middle grey, which is the perceptual middle between what in a slide appears white (let's say, the frame of a window) and what in a slide appears black (it might not be black, imagine the shade under a bush, but it's the "blackpoint" of the image, your eye sees it as black).
Whether you see it with a 150W or a 250W lamp, at 2 meters or at 5 meters, even though the absolute amount of light projected on the screen will be different, your "perception" of middle grey will remain the same, because your brain will "center it" between the two extremes of blackpoint and whitepoint which are present on the image.
If white is presente, and black is present, middle grey on the slide must fall where the human vision would expect it to fall.

And, more in general, if I project an image of "middle grey clouds" on the screen, I will perceive them to be the same middle grey both with a 150W lamp or a 250W lamp etc. (the intensity of the light on the screen will be different, but my vision will adjust for that, just like it adjusts for the different light falling on a print).

And If that wasn't the case, how could we go to the cinema and be satisfied by what we see? Things look natural because white looks white, black looks black, and middle look middle, IMO.

yes but becasue your eye is constantly and subconciously altering its aperture you can never calibrate anything perceptually and get it right unless viewing conditions are always identical and even then its subjective. So you either do it by numbers or do a pratical test to see what you're actually getting which is what I do.

 

[Diapositivo]

One of the great doubts I had was solved by this statement of the Wikipedia voice referred to beforehand:

"
ISO speed for color reversal film is determined from the middle rather than the threshold of the curve; it again involves separate curves for blue, green, and red, and the film is processed according to the film manufacturer’s recommendations.
"
That corresponds to what I was thinking. Giving the toe and shoulder, and given the way slide film work and are used (totally different from negatives, especially B&W negatives) how could sensitivity be measured from point m, which is just an arbitrary point in the highlights (in the case of slides) of the image? And how could one derive the "slope" from so near the base+fog for a slide?

The answer here is clear. Slide film speed is calculated from the middle. The method must be entirely different than for negatives. If I get it right "speed point" for slides is somewhere in the middle of the curve, and that is probably also the case for digital, where certainly you cannot make any reasoning starting from highlights or shadows, for known technological reasons.

My doubts are:
Is "middle grey" for slides always at density 1?
Why it is not at a density which corresponds to an opacity of around 72% (giving a transparence of 18%)?

If I get it right, D is the Log with base 10 of opacity. Opacity is the reciprocal of transmission. If transmission is 0.18, opacity is 5.555.
Density of that transmission is the Logarithm to base 10 of 5.555 which is 0.745.
I would expect slides to have curves which are "centered" at 0.745 but they are not*.

The mother of all question for me is: where (the bloody hell) is the density corresponding to the "middle grey" for slides?

(Please Benskin or anybody else give some answer. Short and long, if you prefer ).

* Please note it's 1:35 at night, I did not sleep much last night, I had a hike of 800 metres of climb which at the moment it's a lot for me, I have a slight headache, and I could be making a huge mess at the moment with any number.

 

[RobC]

you can NOT work it out accurately like that. Each film is different and had a different dynamic range it will accept. The middle of that range will be a hit by a different %age reflectance depending on its ueable dynamic range.

So what you do is, for each film type, you sacrifice a film by photographing a middleish even toned subject such as a rendered wall which has some texture in it. You meter it with your spot meter and close down 4 stops and expose it. You then open up 1/2 a stop and expose it. Then another 1/2 stop and expose it and repeat until you have exposed 17 frames. Now you develop your slides and inspect them. Exposure number 9 should have been the metered value. That is what your meter is producing. If you have a known reflectance subject on the wall at the correct angles (but not used for metering from) then you know from the resulting density of the wall compared to the density of the reference subject how close or far apart they are. If you can then measure the density of the reflectance reference card in the slide, then you know how much you need to adjust exposure by so that the reference card would appear same as its actual colour is. You will have also determined what the actual useable dynamic range of the film is from the resulting 17 frames and where the middle exposure from the actual metered reading is in that range. It might not be the middle. It is likely shifted 1/2 to 1 stop towards one end of the useable dynamic range.
Its a simple test to do and will tell you everything you want to know without resorting to theoretical functioning of your meter.

 

[tedr1]

Rob C: regarding your question about the density of the 18% subject area the answer is approximately 0.7, however this is only an estimate based on interpolating between the steps of a Kodak step wedge transparency having 0.3 increments.

 

[Bill Burk]

What wiltw and I are interested in is the behaviour of film, and of reflected light meters, when confronted with a certain shade of grey, which let's say is 18%, that we call "middle perceptual grey", regardless of the scene.
To make an extreme example, I might have a scene which is solely composed of something which I know has a 36% reflectivity and I want that 36% shade of grey to be naturally recogniseable in my final image. That means the SBR of the image is actually 0 EV or very near there, if it is enlightened with a flat light. The "middle grey" of the scene is 36%.

If I know that my light meter is calibrated to 18%, I know that, by reading the meter indication, and by exposing for 1 more step, I will obtain 36% grey which is the desired result, in my slide developed according to the well-known standard method prescribed by the manufacturer.

But in order to make that reasoning, I must know in advance to what grey is my light meter calibrated. If my light meter was calibrated, for instance, at an 8% grey, I would have to open more than 2 steps, instead of 1 step, in order to obtain that 36% reflectivity grey of my subject.

I like wiltr's methodology and his findings. Even though the shots are digital and digital is calibrated to 18% and has a straight line while film is calibrated to 12% (or maybe 7% but you have to use 12% if you want to compare it to incident) and has a curve... The tests demonstrate the setup and procedure that you can use to verify film.

You can prove the calibration point of film very easily with a gray card if you have a TTL meter. Just put the gray card in the frame, light it evenly, and whatever the in-camera meter wants to put it at, it will put it at that point. Doesn't even have to be gray. It can be any neutral shade between white and black. The meter will adjust exposure until it puts the amount of light on film that it is designed to put at the mid-tone exposure point. Now I would boldly say that I hope a camera meter and a handheld meter are both designed to put the same amount of light on film. If I'm wrong, then there is another whole discussion we have to go into. With all this talk about the difference between incident and reflective... nobody's talked about how both are different from in-camera metering. The in-camera meter doesn't have to assume light loss of the optical system because it is measuring through the taking lens - so instead of using a generic value for light loss, in-camera meters get to account for light loss precisely.

Color slides certainly have the tightest tolerance. Even tighter is movies where scene to scene continuity is required. Slide shows look better when you maintain high quality exposures so the more you know about where your exposures are going to fall the better off you are.

tedr1 explained the results of shooting black and white film of a good long-range grayscale that stretches the limits of what a flat, evenly lit subject can reflect without having 3-D shadows to provide cavities of even darker tones. So 6 2/3 stops is really a good benchmark to keep in mind... for flat copywork. Until that grayscale was pointed out, I was looking at another grayscale from Stouffer that only had 5 1/2 stops... So I had a different benchmark in mind until today.

 

[Bill Burk]

Thank you gentlemen, I'll take that as a no, it is not an underexposed negative. Printed on grade 2 paper the neg reproduces the reflection density range of the original pretty closely. This is basic calibration work to orient myself with a new enlarger and darkroom after a 20 year break.

You can tell your negative is properly exposed because you can see all the steps. An underexposed negative, for example, might show 0.10 density for step number 2. For your grayscale with 0.10 steps, an underexposure where step 2 was where "A" should be (the steps on that Kodak grayscale are A=white, 1, 2 etc) would mean you are one stop underexposed.