Why is Zone V 50% gray?

Discussion in 'APUG.ORG's "Gray" Area Subforum -NOW HYBRIDPHOTO.C' started by RalphLambrecht, Aug 30, 2006.

  1. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    Why does a digital camera record a Kodak Gray Card (Zone V) as 50% gray?

    A Kodak Gray Card represents an 'average' scene of 18% reflectance, but that has nothing to do with a mean or medium on a grayscale. A gray card has an absolute log reflectance density of about 0.75. Medium gray on a 2.2-gamma monitor and a calibrated print thereof is about 0.66. So the image of the card is lighter than the card itself. Why? Who came up with this assumption or standard? Is it a mistake? Did someone assume 'average gray' meant 50%?



    Disclaimer:

    I'm aware that this is a mainly or exclusively analog forum, but I doubt to get a satisfying answer for my question in a digital forum, since they usually understand little about the Zone System. That's why I post the question here. Besides, I need this to work out some details about making digital negatives, which will be used to make pure-silver contact prints. So, it's an analog question in a way.
     
  2. David A. Goldfarb

    David A. Goldfarb Moderator Staff Member Moderator

    Messages:
    17,978
    Joined:
    Sep 7, 2002
    Location:
    Honolulu, Ha
    Shooter:
    Large Format
    A perfect question for the "Gray Area."
     
  3. Chan Tran

    Chan Tran Member

    Messages:
    2,386
    Joined:
    May 10, 2006
    Location:
    Aurora, IL
    Shooter:
    35mm
    I don't know either. I had a question for a long time and I asked many so called experts in the digital field and get no real answer. The question is if we use spotmeter and meter an evenly lit gray card and set the exposure according to this reading what do we get in term of RGB values in the resulting image file? I got chewed out for asking dumb questions by these so called experts because they said that they can change this value in post processing so why care?
     
  4. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    Nice to hear from you again Chan.

    Let's make a deal. You and I will never stop asking 'dumb' question. By the way, I'm pretty sure this one doesn't belong into this group anyway. I suspect, a so-called-expert made a mistake, but I'm happy to be proven wrong. Just hang in there!
     
  5. John Jarosz

    John Jarosz Member

    Messages:
    149
    Joined:
    May 17, 2006
    Location:
    Chicago area
    Shooter:
    Large Format
    I'll just ask a question that may help.

    If you meter the card with a spotmeter and then use the digital camera, does the camera exposure agree with the spotmeter?

    John
     
  6. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    John

    That's a very question. Unfortunately, the difference between 0.66 and 0.75 density is a little more than 1/6 stop in analog photography, so I can't tell for sure (my meters differe more from one to the other). Anyway, let's assume they do. Still, I can develop my film to give me 0.75 in density, but my gamma-calibrated digital system insists on giving me a print with 0.66, which again, is too light for the gray card. It's not much, but why is it there?
     
  7. Chan Tran

    Chan Tran Member

    Messages:
    2,386
    Joined:
    May 10, 2006
    Location:
    Aurora, IL
    Shooter:
    35mm
    Do you know when we send out a digital file for printing, how are the tones be interpreted? In another word what is the RGB values in the file that will produce say a 0.75 density tone on the print?
     
  8. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    I do know that. It depends on the gamma setting of the attached profile, but most are set to gamma 2.2, which returnd about 53 - 54 % for 0.75 density. I've seen some go as high as 58%.
     
  9. L Gebhardt

    L Gebhardt Subscriber

    Messages:
    1,769
    Joined:
    Jun 27, 2003
    Location:
    NH
    Shooter:
    Large Format
    Is this a problem with all cameras or just one in particular? I imagine each manufacturer "tunes" the output to make what they consider a good image. I know that the JPG images I get out of my DSLR are much different looking than the JPG images out of the digital point and shoot. The RAW images as opened by default are also quite different.

    How are you determining density from the RGB values?

    Just curious, but why do you need this info to make digital negatives? Wouldn't you just make your curves match the screen output to the final print output?
     
    Last edited by a moderator: Aug 30, 2006
  10. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    It is a 'problem' with all digital cameras that I could get my hands on. Take a picture of the sky, gray card, white wall, all out-of-focus, it doesn't matter, always 50% in the jpg image (not the raw file, that's a different matter).

    I determined the density by measuring the lumination on the monitor with a densitometer, and this matches exactly the gamma 2.2 theory (at least from 30 to 70% the rest must be different for different reasons). Also, the calibrated lab print comes back with the same relative reflection density of 0.66. Print and screen match at 50%.
     
  11. Chan Tran

    Chan Tran Member

    Messages:
    2,386
    Joined:
    May 10, 2006
    Location:
    Aurora, IL
    Shooter:
    35mm
    Ralph can you direct me to some info on this matter. RGB to print density?
     
  12. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    Sorry, I didn't answer that. Yes, that's what I want to do, but

    1. It isn't actually possible over the whole range, since the monitor's brightness range is far greater than the print's (nevertheless, it can be done for the midtones).
    2. I want to do it objectively and not subjectively.
     
  13. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format

    Chan

    I'm only aware of my own research. I wasn't able to find any info, that's why I'm doing this. You'll find a DigitalStepTablet on my site in the 'Library' section. Attach your monitors color profile to it (mostly 2.2), give it to any digital print shop and have them print it. Then measure the densities and compare them with the screen values.
     
  14. Sponsored Ad
  15. RH Designs

    RH Designs Advertiser Advertiser

    Messages:
    657
    Joined:
    Apr 9, 2003
    Location:
    Yorkshire Da
    Shooter:
    Multi Format
    I think someone quite sensibly decided that "average grey" means 50%, being half way in appearance between "white" (100%) and "black" (0%).

    Taking perhaps an oversimplistic view, how does the camera (whether digital or otherwise) know what it's looking at? If you take a picture of any uniform surface with no detail (whether the eye thinks it's black, white or grey) surely the only sensible result is 50% since the camera has no other reference with which to compare it? This is analogous to a light meter being calibrated for 18% reflectance - 18% is just the figure which produces an apparent mid grey to the eye, rather than the 50% which might be expected.

    There are no absolute whites or blacks when looking at a reproduced image whether on a print or a monitor screen so to me it makes sense that a "50% grey" or "mid tone" or "zone V" or whatever you like to call it would be bang in the middle of the range of tones that the reproduction process is capable of. I was always under the impression (and I may be wrong, I'm no expert on this subject) that the Kodak grey card's reflectance acts so as to appear to the eye to be half way between 100% reflectance (white) and zero reflectance (black) of the light falling upon it. Therefore I would expect a digital system to represent it as 50% on a scale of 0 (black) to 100 (white).

    When I did some experiments on my own ICC profiled printer, I discovered that a 50% grey was reproduced as such on the printer; verified by scanning the print (which included 0% and 100% points for the scanner to use as a reference). There is a formula for relating printed dot density to log.D, and if you measure the log.D of 100% black relative to 0% (i.e the substrate base whiteness) and measure the log.D of the 50% patch, the formula produces the same result providing you account for the system gamma which is the device invented to make 50% grey appear to the eye to be mid-grey.

    In my view, a 50% grey should be printed at such a density on a printer that it scans at 50% also, and this seems to be the case with a profiled system. Having set my system up to do this, scanning a silver print and printing it on my inkjet printer produces a more or less exact copy, and surely that's what we should be aiming for.

    In any reproduction system there is a range of "black" to "white" which is only a fraction of the theoretical brightness range, and 50% grey is half way between those limits of reproduction. I concluded a while ago that measuring the Kodak grey card with a densitometer and trying to relate that to anything else was a route to total confusion, and I abandoned the attempt :smile:

    Just my 2p worth, but it works for me!
     
  16. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    Richard, thanks for the reply.

    Quite understandable view, but let's not confuse the average gray of the Kodak Gray card with 50% on a monitor. They are not the same. IMHO, it makes more sense to represent a gray card at its true reflection value and not arbitarily place it on the convenient 50% mark. If you are correct that someone made this assumption, then we should be glad that Rochester is not close to the North Pole, or the Kodak Gray Card would look more like a Zone VII, and digital camera pictures would be very bright indeed. I hope this is not it.
     
  17. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    Sorry Richard, I overlooked this earlier.

    I don't think this is the source of the Kodak Gray Card. Kodak did numerous tests around Rochester, NY to find out what the average subject brightness ratio and the reflection of an average scene is. The result was a ratio of about 160:1 (a bit more than 7 stops) and 18% reflectance. The reason for the studies were mainly due to print automation, but were used for lightmeter calibration. The lightmeter doesn't know what it's looking at, so let's tell it at least what the average scene looks like and we take any deviation from there. Therefore, it has nothing to do with average grays and everything to do with average scenes. Consequently, I would expect a digital camera to read a Kodak Gray Card at this average scene value and not at an average gray value. Hence my example of the North Pole in the previous post, where the average scene would be quite different from Rochester.
     
  18. RH Designs

    RH Designs Advertiser Advertiser

    Messages:
    657
    Joined:
    Apr 9, 2003
    Location:
    Yorkshire Da
    Shooter:
    Multi Format
    Yes, as I understood it the grey card was intended to provide a guide to exposure for an average scene - and therefore not necessarily an average grey although it is often described as a "mid grey" card. Perhaps what we should do is to find an "average scene", place a grey card in it, photograph it with a digicam and then see what % grey it reproduces at. However, in a digital system of source>computer>printer I would expect 50% to reproduce as 50% and maybe digital exposures of a uniform tone are set to 50% for convenience rather than science, I don't know.

    In your original post you mention the grey card as having an "absolute density" which puzzles me as I thought density was always relative. The formula for printer dot density vs log.D relies on measurements of D.min and D.max for its results, and so as expected a 50% grey would reproduce as a lighter tone on a system with a lower D.max, therefore maintaining the tonal relationships in the original.

    I'm fairly convinced after trying to get my head around this for months that any attempt to relate "mid grey" to a density relative to some "standard" is doomed to failure. After all, a grey card half in shadow appears to have two different tones ... which one is right? And surely in conventional photography, the density of Zone V on a print will depend on the white and black points of the paper being used? Otherwise, if a paper with a low Dmax is used all the shadow tones will be compressed if Zone V is maintained at the same density as a paper with a high D.max?
     
  19. RH Designs

    RH Designs Advertiser Advertiser

    Messages:
    657
    Joined:
    Apr 9, 2003
    Location:
    Yorkshire Da
    Shooter:
    Multi Format
    PS - when I played with digital negatives I also set the curves up so that 50% grey in the digital file scanned in as 50% grey from the silver print, and that all seemed to work too - the remaining tones fell where I expected them to ...
     
  20. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format

    Richard

    This was a lot of good stuff to go through, thanks. Maybe, you, Chris and I should go through this over the phone. Nevertheless, there is a clear difference between absolute and relative density. Absolute density considers the density of the base material, relative density ignores it.

    Zone V has an absolute reflection density of 0.75 in he print, just as the Kodak Gray Card, but if the paper white has a density of 0.05 itself, the relative density of Zone V on that paper is 0.70. In other words, absolute density is material independant and relative density is not. I hope this solves your puzzle. Zone V is always 0.75 in absolute terms, because it reflects 18%. Its relative density depends on the material. Shifting absolute Zone V density with the material is a mistake in my opinion, and causes all sorts of issues.
     
  21. RalphLambrecht

    RalphLambrecht Subscriber

    Messages:
    8,252
    Joined:
    Sep 19, 2003
    Location:
    Florida
    Shooter:
    Multi Format
    Richard

    Your messages contain a lot of data. It takes me more than one try to respond.

    The one directly illuminated. It says so in the Kodak instructions.

    I'm not convinced of that. Zone V reflects 18%, and I want it to in the print too, no matter what paper I'm printing on (absolute vs relative).

    Well, may be they are. It's this absolute vs relative discussion again.
     
  22. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,646
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    Ralph,

    I think a stumbling point is that you are making some assumptions as to some absolutes that may not be so - middle gray being 18%, Zone V having a corresponding print value, the meter being calibrated to any reflectance, and reproducing the reflection value of any point of the subject in the print.
     
  23. Chan Tran

    Chan Tran Member

    Messages:
    2,386
    Joined:
    May 10, 2006
    Location:
    Aurora, IL
    Shooter:
    35mm
    Ralph,
    My symphathy.
     
  24. John Jarosz

    John Jarosz Member

    Messages:
    149
    Joined:
    May 17, 2006
    Location:
    Chicago area
    Shooter:
    Large Format
    Two other thoughts:
    1. Some meters are set up to return an exposure value different than 18% grey. The SEI meter comes to mind immediately.

    2. Back in the 50's (I think is was then) all meters were set to a different standard. (I forget this information exactly, but the standards have changed over the years) My Weston meter does not corelate with modern meters unless a fudge factor is used.

    My reason in pointing out those two items is that (if I remember this correctly) the standardization value was set arbitrarily. In the early early days there was no uniform film speed designation, each manufacturer had their own standard fopr filmspeed.

    I think the digital manufacturers just adopted a different standard than film. I don't think the chip has an S shaped H&D curve like film, it's a straight line, right?

    John
     
  25. JBrunner

    JBrunner Moderator Staff Member Moderator

    Messages:
    7,076
    Joined:
    Dec 14, 2005
    Location:
    Basin and Range Province
    Shooter:
    8x10 Format
    I think the place where things have gone wrong here is to assume the digital camera designers cared, or even knew anything about 18% grey cards, or the zone system.

    Chips and their philosophy spring from the video world, where everything is measured from 0-100% hence the 50% grey thing, which is a measurement that has nothing to do with anything except the bottom and the top of a waveform monitor.

    When you consider that the zone system would have to be modified considerably to work with the short latitude, and the non existent toe and foot of a digital image, it seems to me that you are sort of beating your head against a wall, trying to find something that isn't really there, and is of dubious reliability, even if the tortured truth is wrung out.

    Its like trying to covert dollars into widows mites.

    FWIW
     
    Last edited by a moderator: Aug 30, 2006
  26. RH Designs

    RH Designs Advertiser Advertiser

    Messages:
    657
    Joined:
    Apr 9, 2003
    Location:
    Yorkshire Da
    Shooter:
    Multi Format
    I think that's where we differ - I want the tonal relationships in the image to remain the same, and I think if you specify an absolute density (thanks for clearing that up BTW) in one area then you're going to run into trouble with compressing or expanding the tones in others. Taking an extreme example, what would happen if you were to print an image with a full range of tones on a material with a D.max corresponding to 18% reflectance? If Zone V is also 18%, then you will have zero separation of Zones I-V which would look a bit odd to say the least! I'm no scholar of the Zone System but did Adams specify absolute densities for print materials, and if so was he doing so based on a typical D.max of 2.0?