Kodak grey card usage

[Diapositivo]

Wiltw, your demonstration would have been more demonstrative if the single images had been clickable in order to obtain a single image on screen. That would clearly force the eye to adjust to the blackpoint and whitepoint it sees on the image and settle the middle immediately.

By seeing the images all together, the eye tends to choose the whitest white as whitepoint and the blackest black as blackpoint, and they are from two different images.

But I get your point, which is a point we have in common

 

[wiltw]

Wiltw, your demonstration would have been more demonstrative if the single images had been clickable in order to obtain a single image on screen. That would clearly force the eye to adjust to the blackpoint and whitepoint it sees on the image and settle the middle immediately.

By seeing the images all together, the eye tends to choose the whitest white as whitepoint and the blackest black as blackpoint, and they are from two different images.

But I get your point, which is a point we have in common

I am already bucking an anti-digital "give me film or give me death" attitude via my comparative postings!

In my mind, APUG is "All Photography User Group" and there are more commonalities than differences. After all B&W neg and print is very, very different from color transparencies, yet they coexist on APUG.

 

[RobC]

but the common denominator is negs and trannies are both Analogue as in Analogue Photography User Group.

 

[wiltw]

but the common denominator is they are both Analogue.

And because the point is about human perceptions of tones and 'middle' vs the two extremes of tone, even when none are correct for absolute densitometer readings, I posted those photos to demonstrate principles of color and tonal perception. If the same photos were published in a book, you would not necessarily even be told the original photo acquistion methods.

 

[RobC]

I'm not knocking your use of digital for illustration purposes. I haven't been reading all the posts as the topic's gone off topic and into an area I'm simply not interested in discussing. The argument has become pointless and is not going to be solved by a formula. If Diapositivo can't see that, it's his problem. If he wants to know exactly how his meter works then he can contact its manufacturer technical support and ask them ( like I did with Minolta when they made spot meters)

 

[markbarendt]

In post #268 Diapositivo made a statement, which I fundamentally agree with and provide an illustration of the point: "Of course there are different light conditions in which you see a print or a slide. But your eye/brain is smart enough to compensate, as far as possible, for the variations and you end up with a mental image of that image which is always the same."

If that were true when having a slideshow or watching a movie in the theater we wouldn't bother turning of the lights. For trannies turning off the lights in the room makes blacks look blacker and whites whiter. In reality though every tone on the screen gets darker.

Visual perception is relative. Brightness in photography is more about the difference in tones, rather than the luminance.

 

[wiltw]

If that were true when having a slideshow or watching a movie in the theater we wouldn't bother turning of the lights. For trannies turning off the lights in the room makes blacks look blacker and whites whiter. In reality though every tone on the screen gets darker..

You are unfortunately confusing issues....if the room lights were up, our eyes would not be dark adapted sufficiently to see the details in the darker scenes of a movie or slide show. That brings to mind a Clint Eastwood movie about a stolen warplane, which was so dark you could not see very well even in a darkened theater.

Our brain perceives 'black' on TV screens, in spite of different TVs in different room brightnesses, and in spite of the TVs themselves having different adjustments for 'black level'...LCD vs. LED vs. plasma.

 

[RobC]

And here is what Minolta have to say in the manual of a spotmeter F.

Like all reflected-light meters, the spotmeter F is calibrated to a reading that falls on the mid-point of the films characteristic curve, and will be reproduced as a mid-tone (medium) density on the film. In other words, the meter gives a reading that will result in "normal" exposure for an "average" subject. An average subject is generally defined as one that reflects about 18% of the light that strikes it.

So what you can take from that?
Well firstly it places the exposure in the middle of the curve and not on an 18% reflectance in the subject.
Secondly it can't place the exposure in the middle of the curve if it doesn't know what log range the curve is and since neg film and slide film and different neg films and slides films all use different film log ranges its impossible for the meter out of the box to place anything in the middle of the curve unless the curve exactly matches the curve that the meter manufacturer used in its testing and setting of the meter.
Thirdly, if the subject doesn't have average reflectance of 18% the meter can't get it in the middle of the curve anyway.
And fourthly, if you use a grey card in a subject which is not 18% average overall then you won't get your subject exposed on film correctly. It will be shifted to suits the card and not the subject. i.e. the idea of the meter is to get everything on film and not put 18% in the middle of the curve unless by pure chance 18% happens to fit the meters calibration.
Fithly they ar eassuming the midpoint of the curve will be reproduced as a middle grey. Well it might or it might not be. That is outside of the meter manufacturers control and is wholly dependant on whoever and how they are reproducing it and whether they think it should be 18% grey.

So is a reflected light meter using the curve from neg film or transparecy film in its internal calibration to place exposure on the middle of the curve(NOT necessarily 18% grey reflectance). You tell me.

Diapositivo, your understanding and expectations of what a reflection meter does are wrong. Tough luck, you will have to do testing and calibration to get to where you want be.

Further more, every time your SBR changes you will have to adjust the offset you use. Suggest you go buy yourself a colour meter if you want exact colour matching but methinks you will still have to calibrate that.

But if it keeps you happy just keep chasing the end of the rainbow but don't tell me about it.

 

[wiltw]

And here is what Minolta have to say in the manual of a spotmeter F.


So what you can take from that?
Well firstly it places the exposure in the middle of the curve and not on an 18% reflectance in the subject.
Secondly it can't place the exposure in the middle of the curve if it doesn't know what log range the curve is and since neg film and slide film and different neg films and slides films all use different film log ranges its impossible for the meter out of the box to place anything in the middle of the curve unless the curve exactly matches the curve that the meter manufacturer used in its testing and setting of the meter.
Thirdly, if the subject doesn't have average reflectance of 18% the meter can't get it in the middle of the curve anyway.
And fourthly, if you use a grey card in a subject which is not 18% average overall then you won't get your subject exposed on film correctly. It will be shifted to suits the card and not the subject. i.e. the idea of the meter is to get everything on film and not put 18% in the middle of the curve unless by pure chance 18% happens to fit the meters calibration.

So is a reflected light meter using the curve from neg film or transparecy film in its internal calibration to place exposure on the middle of the curve(NOT necessarily 18% grey reflectance). You tell me. Its aiming at putting 18% grey in middle of curve and not reproducing the same colour accurately.

Diapositivo, your understanding and expectations of what a reflection meter does are wrong. Tough luck, you will have to do testing and calibration to get to where you want be.

Further more, every time your SBR changes you will have to adjust the offset you use. Suggest you go buy yourself a colour meter if you want exact colour matching with ease.

Rob, if we drop the idea that meters are EXACT and always indicate the 'right exposure' in spite of the differences between types of film we might shoot with the same meter, and we keep in mind the concept that meters get us merely 'in the ballpark', then you can understand the meter and its attempt to put midtone (or 18% grey) 'in the middle' so to speak. Our underexposure or overexposure may move the entire curve to the left or right, but 'in the middle' is always the figurative placement for the middle tone.

The incident meter failed to properly expose my 'flat art', the whites were blown out.
The incident meter might similarly fail to get vivid colors of the sunset, which is why bracketing may often be needed. And certainly the incident meter also fails in a sunny scene with areas of detail in the shade, losing that shade detail.

 

[Diapositivo]

If that were true when having a slideshow or watching a movie in the theater we wouldn't bother turning of the lights. For trannies turning off the lights in the room makes blacks look blacker and whites whiter. In reality though every tone on the screen gets darker.

Turning off the lights widens the distance between blackpoint and whitepoint. The wider the distance, the more "snap" the photograph is. We see more vivid colours, a much more pleasant, and more realistic, tonal range.
When you turn on the lights, the white point and the black point come very near, because the black point is raised to the general light level of the room (the white point stays the same, because normally the white on the screen is "whiter" than the lights of the room).
What matters is not how white is the white point, but how distant is blackpoint from whitepoint.

The reason why slides properly projected are generally speaking more pleasant than colour prints is that the actual blackpoint and whitepoint of a print are nearer.

 

[Diapositivo]

Like all reflected-light meters, the spotmeter F is calibrated to a reading that falls on the mid-point of the films characteristic curve, and will be reproduced as a mid-tone (medium) density on the film. In other words, the meter gives a reading that will result in "normal" exposure for an "average" subject. An average subject is generally defined as one that reflects about 18% of the light that strikes it.

What Mr. Minolta is trying to tell us, although I deem he does it with an unfortunate choice of words (just as my Italian translated into English shows its limits, imagine Mr. Minolta thinking in Japanese and translating into English...) si that all light meters are calibrated to give you the exposure that Mr. ASA/ISO calls Hg, and that that exposure is correct ("normal" exposure) for my 18% cat shade of grey ("average" subject), and not for your 90% cat shade of white, and he is also telling us that that exposure will be reproduced as a certain middle density of film, not a casual one, but one that, with a normal standard industrial process, will result in a final image which reproduce both my cat and your cat as an 18% middle grey cat, only that mine is correctly reproduced, and yours is badly underexposed.

But Mr. Minolta did not know that people in this forum question the concept of middle grey, and light metering calibration to a middle grey, so he did not elaborate the subject more than in a passing way.

Basically Mr. Minolta, I think, agrees with me on all regards, but does not substantiate his agreement with me with sufficient means of demonstration.

Mind you, Mr. Minolta says that all reflected light meters are calibrated to a certain tone, a certain shade of grey.

That's Y = kX in English words, and k is that certain tone to which any reflected light meter must be calibrated. In the case of the Spotmeter F, Mr. Minolta is telling you that k = 0.18. (That's not K as the K constant in the exposure equation).

You seem to believe Mr Minolta usually

 

[Diapositivo]

As a passing note I would like to stress that my interest is for k not for K, as above defined.

K is an engineering detail that is totally irrelevant for me. It's "under the hood". I don't need to know it more than I need to know how the tension is stabilized inside the device, or which condensers are used. I only need to know how to use the device.
When you drive a car, you must know how to use clutch, brake, accelerator, gear lever, and steering wheel. You don't care about ignition, injection, and lambda sensor.

K to me is interesting only because, known K, you basically know k.

k is important because all "placing" you do on the film curve of the reading you make with the spot meter are relative to middle grey, are off-sets of middle grey, so you must know where middle grey is.

For instance white (as a white shirt) is 2.5 EV above middle grey, if middle grey is 18%.
If you think that middle grey is 8%, in order to place white you must place it more than 3.5 EV above middle grey. How can you make a placement if you don't know where do you start from?

 

[RobC]

Rob, if we drop the idea that meters are EXACT and always indicate the 'right exposure' in spite of the differences between types of film we might shoot with the same meter, and we keep in mind the concept that meters get us merely 'in the ballpark', then you can understand the meter and its attempt to put midtone (or 18% grey) 'in the middle' so to speak. Our underexposure or overexposure may move the entire curve to the left or right, but 'in the middle' is always the figurative placement for the middle tone.

The incident meter failed to properly expose my 'flat art', the whites were blown out.
The incident meter might similarly fail to get vivid colors of the sunset, which is why bracketing may often be needed. And certainly the incident meter also fails in a sunny scene with areas of detail in the shade, losing that shade detail.

No it doesn't move the curve left or right, it shifts the point up or down the curve dependant on the curve slope which are very different for slide and neg films. a 3 stop adjustment on a steep curve will place the film density at a different value than a 3 stop adjustment on a shallower curve.

 

[markbarendt]

Turning off the lights widens the distance between blackpoint and whitepoint. The wider the distance, the more "snap" the photograph is. We see more vivid colours, a much more pleasant, and more realistic, tonal range.
When you turn on the lights, the white point and the black point come very near, because the black point is raised to the general light level of the room (the white point stays the same, because normally the white on the screen is "whiter" than the lights of the room).
What matters is not how white is the white point, but how distant is blackpoint from whitepoint.

The reason why slides properly projected are generally speaking more pleasant than colour prints is that the actual blackpoint and whitepoint of a print are nearer.

My point exactly, our big brains can't make the lights on picture seem like the lights out picture.

 

[RobC]

As a passing note I would like to stress that my interest is for k not for K, as above defined.

K is an engineering detail that is totally irrelevant for me. It's "under the hood". I don't need to know it more than I need to know how the tension is stabilized inside the device, or which condensers are used. I only need to know how to use the device.
When you drive a car, you must know how to use clutch, brake, accelerator, gear lever, and steering wheel. You don't care about ignition, injection, and lambda sensor.

K to me is interesting only because, known K, you basically know k.

k is important because all "placing" you do on the film curve of the reading you make with the spot meter are relative to middle grey, are off-sets of middle grey, so you must know where middle grey is.

For instance white (as a white shirt) is 2.5 EV above middle grey, if middle grey is 18%.
If you think that middle grey is 8%, in order to place white you must place it more than 3.5 EV above middle grey. How can you make a placement if you don't know where do you start from?

all a meter can do is make an adjust based on the assumption that either its starting from zero or the speedpoint or the reading it gets in cd/m2, it can't magically know how long the curve is for any particular film is unless you meter the darkest point and the lightest point and average them.
But the middle film density will still be dependant on the steepness of its slope which varies with each film type. So that doesn't get you where you want to be unless you calibrate for it. Sure you can get it in the middle of the curve by averaging which will give a decent expsoure but not an exact colour match unless you force it in reproduction.

 

[Diapositivo]

No it doesn't move the curve left or right, it shifts the point up or down the curve dependant on the curve slope which are very different for slide and neg films. a 3 stop adjustment on a steep curve will place the film density at a different value than a 3 stop adjustment on a shallower curve.

ISO prescribes that the film be correctly developed so that the slope curve to be obtained is fixed. Given an A point of 0.1 density, the density point corresponding to a higher exposure of 1.3 LogH must be 0.80 more than density at point A, with a small tolerance of 0.05 on both sides. That means ISO defines pretty exactly the slope of the characteristic curve (for negatives, at least, but I do believe also for slides).

If that was not the case, we would have films that increase (or decrease) tonal relations thus giving us a very unnatural, and unexpected rendering of the scene.
A chess table with light grey and dark grey boxes would be rendered as black and white boxes in a slide.

(I know you are very advanced photographers and do these tricks all time with B&W negatives, playing with contrast and tonal value in exposure, development, and printing stage with various paper exposure, using papers of various grades and various paper development times, but that's not "normal" or "standard" development as defined in ISO).

 

[Diapositivo]

My point exactly, our big brains can't make the lights on picture seem like the lights out picture.

But my point and wiltw's is that once you see an image correctly formed, the middle grey will be settled by the eye in the middle between highpoint and blackpoint. That's true for prints, slides, television, etc.
Like all technologies, you have to make them work in the conditions for which they were engineered.
You can't enjoy Hi-Fi in a noisy room and you cannot say in a noisy room your Hi-Fi has very poor dynamic range ;-)

 

[Diapositivo]

all a meter can do is make an adjust based on the assumption that either its starting from zero or the speedpoint or the reading it gets in cd/m2, it can't magically know how long the curve is for any particular film is unless you meter the darkest point and the lightest point and average them.
But the middle film density will still be dependant on the steepness of its slope which varies with each film type. So that doesn't get you where you want to be unless you calibrate for it. Sure you can get it in the middle of the curve by averaging which will give a decent expsoure but not an exact colour match unless you force it in reproduction.

No, the light meter never starts from zero the way you think it. The ISO speed determination starts from point M but don't be mislead by that. The lightmeter gives you Hm which is the exposure creating middle grey on your ISO film. What "magically" tells the light meter where middle grey is, is the ISO speed that you plug in it (plus the K constant which is an "adjusting factor" that the maker plugs in it).

The fact that ISO is determined, for negative films, on the shadows leads you and somebody else to view the exposure data as something that is calculated by the lightmeter from the shadows, from the "speed point".
The speed point is used for the ISO speed determination, and there its function ends;
The ISO speed determination is used to determine Hg which creates the normal exposure for the average 18% cat with a reflected light meter, as Mr. Minolta says. And that both for slides and for negatives.

An incident light meter has no k factor because it needs not assume a certain reflectivity of the subject. All reflected light meters must assume a certain reflectivity (a certain "tone") of the subject.

 

[Stephen Benskin]

The eye adapts to changes in illumination, but there are other factors involved like surround, local adaptation, and simultaneous contrast. The Theory of the Photographic Process 3rd edition has a very nice section on perception and subjective tone reproduction. It's a good but difficult read.

Here's an example of simultaneous contrast. The gray bar is a single tone against a gradient background.



And another example



For a quick experiment, take a transparency (larger format the better). Place it in the middle of a light box. Note how it looks, then quickly place a mask around the transparency. It will look different.

 

[Diapositivo]

Kodak instructions for using a grey card say that the orientation of the card should be as follows.

Looking from above, the plan view from top, angle the card 1/3 of the way from the subject to camera axis towards the main light source.
AND
Looking from the side, angle the card 1/3 of the way from subject to camera axis towards the main light source.

So you need to tilt the card to the side and upwards assuming your light source is higher than the camera to subject axis.

Note: just becasue the lighting is diffuse doesn't mean you can forget about the angles because you can't. The angle the light is reflected from the card makes a big difference. Its not just the angle of light falling on the card that counts.

If you are not adhering to the correct angles then your card is not reflecting 18%. It becomes an unknown reflectance and is no use as a known reflectance.

Furthermore, if lighting is low or high then adjustments need to be made.

Personally I don't use a grey card since doing my own film speed, dev and printing calibration means I get more accurate readings and exposure placement than I would when using a grey card.

I think the right answer is: the grey card is to be kept in such a way that the light source is outside of the family of angles that would create a direct reflection (a "specular" reflection).
A direct reflection is created when the light source is at an angle such that you would see the light source reflected in the subject if the subject is a mirror.
If the light source is outside the family of angles that create a direct reflection, you only have a diffuse reflection. That gives more or less the same light (and the same light reading at the light meter) at whichever angle you keep it provided the condition above is satisfied.

Kodak, I deem, does not feel like explaining what a family of angles is and what a direct reflection is, so it gives these instructions.
Basically, avoid specular reflection of the light source and you should obtain a very good value while using a grey card.
(Or, forget the grey card, and use an incident light meter!).

Grey card remain interesting tools for photographing artwork (as wiltw demonstrated) and other flat subject. You place them over the artwork. It goes without saying the light source/s is outside of the family of angles creating a direct reflection otherwise you would obtain a very poor art reproduction (but maybe you can claim your art reproduction is "artsy" ).

 

[Diapositivo]

The eye adapts to changes in illumination, but there are other factors involved like surround, local adaptation, and simultaneous contrast. The Theory of the Photographic Process 3rd edition has a very nice section on perception and subjective tone reproduction. It's a good but difficult read.

Here's an example of simultaneous contrast. The gray bar is a single tone against a gradient background.

View attachment 153462

And another example

View attachment 153463

For a quick experiment, take a transparency (larger format the better). Place it in the middle of a light box. Note how it looks, then quickly place a mask around the transparency. It will look different.

Yes, but that belongs to the quirks of human vision and brain reconstruction of reality.
This link shows other things like False colour (the eye sees colours that don't exist in the image) and Mach Bands (your eyes sees shades of grey that don't exist in reality)

http://www.cambridgeincolour.com/tutorials/cameras-vs-human-eye.htm

Normally though we don't see images of grey squares in white squares. In normal average images (of cats ) we see a "complex" scene, with a black point, a white point, and a grey point in the middle. That's true for 99% of images, I reckon.

 

[RobC]

the slope of slides and negatives is different as far as I'm concerned. Here is direct comparison of provia against tri-x. the kodak trix chart has been scaled to match the spacing of the provia. ie the spacing of density and exposure show the same size grid and values. Clearly the tri-x gives a much shallowr curve than the provia. You can take the range of exposure across the straight line portion of both curves and divide it into the density range that the exposure range covers and get the films Contrast index.
The provia gives approx 1.6 and the TRI-X gives approx 0.55. Clearly the slope is very different and therefore exposure needs to be calibrated to find the middle film density depending on its slope. We know B&W film requires a film density range 1.3log. provia goes all the way upto well over 2.0log The middle of those curves is obviosuly going to be a different density. Does the exposure required to hit the middle of the two curves differ? Well simply looking at the charts I think we can say yes. If the slide film accepts 6 stops of SBR the middle of curve will be approx 3 stop adjustment from either end. If neg film accepts 7 1/3 stops then middle of curve will be approx 3 2/3 adjustment from either end. They are different but the meter doesn't know that so how can it know how much to offset for all types of film? It doesn't. Its only calibrated to one curve which doesn't fit all. You must calibrate exposure if you want accuracy.

 

[markbarendt]

Yes, but that belongs to the quirks of human vision and brain reconstruction of reality.
This link shows other things like False colour (the eye sees colours that don't exist in the image) and Mach Bands (your eyes sees edges that don't exist in reality)

http://www.cambridgeincolour.com/tutorials/cameras-vs-human-eye.htm

Normally though we don't see images of grey squares in white squares. In normal average images (of cats ) we see a "complex" scene, with a black point, a white point, and a grey point in the middle. That's true for 99% of images, I reckon.

We don't see just the image, we see the frame, matting, wall, ...

 

[Diapositivo]

the slope of slides and negatives is different as far as I'm concerned. Here is direct comparison of provia against tri-x. the kodak trix chart has been scaled to match the spacing of the provia. ie the spacing of density and exposure show the same size grid and values. Clearly the tri-x gives a much shallowr curve than the provia. You can take the range of exposure across the straight line portion of both curves and divide it into the density range that the exposure range covers and get the films Contrast index.
The provia gives approx 1.6 and the TRI-X gives approx 0.55. Clearly the slope is very different and therefore exposure needs to be calibrated to find the middle film density depending on its slope. We know B&W film requires a film density range 1.3log. provia goes all the way upto well over 2.0log The middle of those curves is obviosuly going to be a different density. Does the exposure required to hit the middle of the two curves differ? Well simply looking at the charts I think we can say yes. If the slide film accepts 6 stops of SBR the middle of curve will be approx 3 stop adjustment from either end. If neg film accepts 7 1/3 stops then middle of curve will be approx 3 2/3 adjustment from either end. They are different but the meter doesn't know that so how can it know how much to offset for all types of film? It doesn't. Its only calibrated to one curve which doesn't fit all. You must calibrate exposure if you want accuracy.

View attachment 153464

The light meter gives you a value of exposure, Hg, that, given a standard process, will create a middle grey (18%) on the final image. That is not necessarily in the exact middle of the curve of film. Slide film has slightly more linear curve on the left (shadows) than on the right (highlights) of middle grey. That's why Mr. Minolta made your Spotmeter F in such a way that it places the highlights 2.3 EV above middle grey, and the shadows 2.7 below middle grey. Because he imagines the "good detail rendition" of slide film to be 5 EV of range, but not equally distributed. There is more linear range on the shadows side than on the highlights side.
Yet, the light meter will give you the exposure of middle grey. You paid it for that purpose.

For negative film, the exposure latitude is way bigger on the right side (highlight side, higher density) than on the lower side (shadows) of the curve. So for negatives the Hg is "further away" from the middle of the curve.
That's why it is important to understand where Hg sits on the characteristic curve.
The exposure your lightmeter gives you is always a certain POINT in the characteristic curve (Hg) where g means, I suppose, grey. When you modify the exposure you place the metered spot at a certain offset, along the LogH axis, from that point.

If you look back at post #205, page 9, by Stephan, the bottom image gives a placement for Hg, the exposure indicated by your light meter. That's 1.27 LogH in this case. If you want skin tone, which is at a certain offset from that point, you read the indication of the light meter, and you offset it of the same offset you see, in the graph, between Hm and Hskin tone.

 

[RobC]

which means you must calibrate your exposure if you want accurate colour since you don't know where on the curve for any particular film it will be. From calcs in earlier post I said its 3 stops probably from speed point. And that means as you rightly say it won't be in middle of curve unless its fits manufacturers test curve which most of the time it won't. So how can it be exact middle grey? It can't.
You now seem to be saying meters are calibrated for slide film. Minolta may be better for slide film but I don't think others are.