Photo science geeks HELP!

[UncleDoug]

OK, I'm a geek.

Are you?
If you can help me out on this one, you just may qualify....:D

:idea: Situ: outdoor, clear skies.
After sunset, if you look at a grey card you will perceive it differently than you would in the noon day sun.

Excluding color from this, i.e. concentrating on lightness only, how would you go about determining what exposure would be necessary to take an exposure of the grey card that would mimmic what your eye saw at that moment?

I've been fighting with this one for a day or two and have learned allot! But probably need some fresh angles of attack on this one.

Some of the concepts going on here are; eye dark adaptation, contrast ratios, lightness constancy, Webber's law thresholds for optical systems, simultaneous lightness contrast and how they relate to photography.

Any help, musings or a good heckle:razz: would be greatly appreciated! :D

[PhotosGuy]

how would you go about determining what exposure would be necessary to take an exposure of the grey card that would mimmic what your eye saw at that moment? One question - many answers! ;) The eye has WAY more dynamic range, so the problem is, just what is it looking at? You'll need to intergrate the exposure for that with the surroundings elements that you also want to see. Maybe you'll have a graduated ND filter to use. Maybe you'll have to take 2-3 pics & blend them in PS.
A "Quick & dirty" solution, but it works.

Forget the gray card when you're shooting sunsets IF the subject is the sunset, not the light falling on you. If the sun is still in the sky, meter about 30 degrees away from it so it's not in the viewfinder. Don't custom WB or you'll neutralise some of the warm colors. Try the presets for Sunny, Shade & Overcast to see which you like best.

When the subject is not the sunset, your eye is quick to adapt so you can't really trust variations on the "Sunny 16" rule. For late evening car shots, I sometimes use Shade & finish the WB in RAW processing. Most times, I'll do a new Custom WB & exposure check just before I shoot 'cause the light changes FAST! Then I might have to warm up the pic a bit 'cause I like them that way.

[lostdoggy]

Paraphrasin the late Bob Ross, "its your world, it could be anything you want it to be"

[UncleDoug]

Thanks for the replys.

PhotosGuy,

Thanks for the suggestions, but the subject of my question is the grey card.
This is a rather esoteric thing.
I've shoved my head in prety deep on this one.

Here is another example.

You are outside at 12:00 noon on june 21. You look at a grey card under direct light. You wait 15 to make an "observation" so your eyes have fully adapted to the lighting contitions. The card appears neutral grey of a certain tone level. Now at 9:30 that evening, just after sunset around here, you perform the same experiment, look, wait, make the observation. The card will appear darker or have a darker tone to it as compared to the observation at noon.
What I want to know is how to determine an exposure at noon of the grey card so that when viewing it in print it would have the same tonality, under a standard viewing condition, as what I observed at noon that day. And as well for the situation at 9:30.

[J Rabin]

This is not a geek Q.

It appears more about obtaining a exposure FAITHFUL to the scene as you remember it, vs. technically ACCURATE exposure from metering. This is the Chris Weston approach in his Photographer's Inst. Press book on exposure.
Personally, from a film background, I go for faithful over accurate exposure, unless it's a shot dealing with scientific subject matter.
For example, deliberate under exposure to saturate earthy tones, deliberate over exposure for high key oriental look. These are all about being faithful to the scene, not accurate to the meter. Same with sunsets.

Another author addresses this as quote, "Getting the light right, versus getting in the right light" are two different challenges. Getting the light right is all about exposure. Getting in the right light is all about composition, mood, tone, etc.
I think this is a nice way of relating to the challenge.

There is no right answer, its about faithful exposure and taste of the shutter presser, unless your Q was about where to find a middle tone to meter in these off times......
Jack

[UncleDoug]

Another author addresses this as quote, "Getting the light right, versus getting in the right light" are two different challenges. Getting the light right is all about exposure. Getting in the right light is all about composition, mood, tone, etc.
I think this is a nice way of relating to the challenge.

I agree.

This has more to do with how the human eye, on average, reacts/adapts to different lighting conditions than to photography, directly. Trying to keep any artistry out of the equasion, this is a purely scientific and theoretical pursuit that may lend itself to some relay awesome creativity.;)

Thanks for the input.:D

[PhotosGuy]

What I want to know is how to determine an exposure at noon of the grey card so that when viewing it in print it would have the same tonality, under a standard viewing condition, as what I observed at noon that day. And as well for the situation at 9:30. What Jack said.
To answer your question, as written, is easy. The gray card will have the same tonality under all lighting conditions since it has only one "tone' to start with. Problem is that the tone on the card is ink which may not reflect the same wavelengths of light as the dye on the paper. Then there's that even getting that right when I shoot a car & the paint won't match the ink used in printing for the same reason. So you can get the card "right", but the paint won't be. Fun, huh?
It's a complex problem of calibration in the printing process which is a whole huge can of worms. Commercial printers use a process of Pantone inks exactly calibrated to high accuracy which you can't do. There are lots of threads here on calibration.
To start, look at these links:
Monitor calibration and gamma
Quick test:
http://epaperpress.com/monitorcal/index.html
Better info:
http://www.normankoren.com/makingfineprints1A.html
http://www.normankoren.com/makingfineprints1B.html
Dry Creek "Introduction to Digital Photo Lab Profiles"
http://www.drycreekphoto.com/Frontier/index.html

[Loki1117]

I agree.

This has more to do with how the human eye, on average, reacts/adapts to different lighting conditions than to photography, directly. Trying to keep any artistry out of the equasion, this is a purely scientific and theoretical pursuit that may lend itself to some relay awesome creativity.;)

Thanks for the input.:D

I think you are right when you say "This has more to do with how the human eye, on average, reacts/adapts to different lighting conditions," but not when you say "this is a purely scientific and theoretical pursuit." The problem is that you are looking at this, and then wanting to replicate what you "Remember" seeing. You as a person (regardless of how much you may try to be impartial) are biased. For this to be scientific, you would need to take yourself out of the equation and base your findings on fact rather than perception.

You need some type of control subject to compare. Maybe by shooting part of the scene with a controled light source and then shooting the same scene without the added light source. The problem you will face then is the limited latitude of light that a digital sensor can perceive.

You do bring up an interesting point though about what we perceive and how the scene really looked. If you come up with a way to actually validate a scenes "luminosity" I would be interested to hear.

[UncleDoug]

You do bring up an interesting point though about what we perceive and how the scene really looked. If you come up with a way to actually validate a scenes "luminosity" I would be interested to hear.

You are right about every person percieves things differently!:D

But by using the average human eye, I'm hoping to take my perception out of it.
Kind of like what the CIE & ICC use as a standard observer. They had to base their models on an "average" human eye, extrapolated from known scientific fact.

Working on another diatribe right now...

You guys really have me thinking on this one.

[Robert_Lay]

Here are a few imponderables.
1) Assume that the camera metering is perfect. If you let the camera metering determine the exposure, all you really get in the JPG virtual image is a number from 0 to 255. Consider that the number is of no significance other than as a calibration point for that image.

2) Consider what has been said in (1) above and apply it to a 2nd shot taken under different lighting conditions. All you have now is another calibration point that has no relationship whatsoever to the first calibration point other than the fact it is for the same camera.

3) In neither case above can you possibly produce a print or a screen display of either image that looks to your eye like the actual scene unless you employ another arbitrary calibration for each image and for the monitor used and the software and hardware that performs the print or display.

4) [We're close to the end now.] It is impossible to employ the two calibration points of a given image for any other image taken at any other time.

The basis for all of the above is that the system involved is not linear.

[Big_B]

I'm about to go out, so only time for a super quick thought. You said at the start that we should ignore colour, but I'm not sure that is correct. After all, grey is an amalgam of different wavelenghts of light and the brightness of the card will be determined by the quantity of photons of those wavelengths that are reflected.

My point about colour is that the wavelenght makeup of ambient light will vary throughout the day. Taking a hypothetical situation where we vary the wavelength balance, but keep the intensity of light the same, the 'brightness' of the grey card will change.

[Robert_Lay]

I'm about to go out, so only time for a super quick thought. You said at the start that we should ignore colour, but I'm not sure that is correct. After all, grey is an amalgam of different wavelenghts of light and the brightness of the card will be determined by the quantity of photons of those wavelengths that are reflected.

My point about colour is that the wavelenght makeup of ambient light will vary throughout the day. Taking a hypothetical situation where we vary the wavelength balance, but keep the intensity of light the same, the 'brightness' of the grey card will change.

In my opinion the brightness of the Gray card should NOT change, because an 18% Gray card is assumed to reflect 18% of the incident light, regardless of its spectral content. (yes, I know, that is probably idealistic).

[UncleDoug]

It's a complex problem of calibration in the printing process which is a whole huge can of worms.

Thanks for validating the complexity of this! :D
Got the color management theory & practice down, so that is out of the loop for the most part.

I think I'm having massive trouble clearly articulating what I'm investigating.


Here's a quick experiment.
Be cognicant of your perceptions throughout.

2 pieces of white paper.
1 out in the open, the other in a box, with the side facing you removed so that you can see the piece of paper in the box. Both pieces are alligned the same relative to you. Both are in close proximity, next, to eachother.
Now view the scene so that both pieces of paper are visible. The piece in the box will appear darker or less white than the one outside.
Now take exposures of both, metering and exposing "correctly" for both and using auto white ballance, if using DSLR.
Besides a slight color cast the lightness of both images will be darn near the same, even though you perceive the one in the box to be darker.
If you take an exposure that has both pieces of paper in the scene; meter for the piece outside, the one in the box will be darker; meter for the one in the box the one outside will be blown out.; if you meter the ambient light the piece in the box will be darker than the one outside.
This is expected.

From this I'd like to know, which exposure will best reflect what YOUR EYE SAW IN THE BOX.

Taking this further, in a situation in the early morning "twilight" when you are fumbling around for your lense cap due to the lack of light, you place a piece of white paper on a rock. You want to take an exposure of the piece of paper to reflect how you eye would pervieve it at that point in time under those dim lighting conditions. What would you do?

More musings tomorrow, time for a soda! :mrgreen:

[PacAce]

What Jack said.

To answer your question, as written, is easy. The gray card will have the same tonality under all lighting conditions since it has only one "tone' to start with. Problem is that the tone on the card is ink which may not reflect the same wavelengths of light as the dye on the paper. Then there's that even getting that right when I shoot a car & the paint won't match the ink used in printing for the same reason. So you can get the card "right", but the paint won't be. Fun, huh?
It's a complex problem of calibration in the printing process which is a whole huge can of worms. Commercial printers use a process of Pantone inks exactly calibrated to high accuracy which you can't do. There are lots of threads here on calibration.
To start, look at these links:
Monitor calibration and gamma
Quick test:
http://epaperpress.com/monitorcal/index.html
Better info:
http://www.normankoren.com/makingfineprints1A.html
http://www.normankoren.com/makingfineprints1B.html
Dry Creek "Introduction to Digital Photo Lab Profiles"
http://www.drycreekphoto.com/Frontier/index.html
Although a gray card will have a fixed reflectiviy, such as 18%, that doesn't mean that the gray card will look the same under all lighting conditions. For example, is the amount of light falling on this card in a dimly lit room the same amount of light falling on it out in the open sunlight? No. So the light reflected back by the card should appear different, too, even though the card will be relecting 18% of the light back at the observer in both instances.

I think taking the EV of the light into consideration will give a better indications of the lightness or the darkness of an image since EV is absolute and not relative. All we need is for someone to come up with a mapping of the EV to what our eyes perceive (and then translated into a 0 to 255 gray-scale levels) since there obviously won't be a one for one relationship between the two. I'm no expert here so I'm just speculating but I would think that a gray card would look almost the same (after adaptation and ignoring color cast) to the eye no matter what the lighting is as long as the pupils are not fully dilated or closed down. But in very dim lighting or very bright lighting where are pupils are at their extremes, what the eye perceives would vary with the amount of light falling on the card.

[PhotosGuy]

Taking this further, in a situation in the early morning "twilight" when you are fumbling around for your lense cap due to the lack of light, you place a piece of white paper on a rock. You want to take an exposure of the piece of paper to reflect how you eye would pervieve it at that point in time under those dim lighting conditions. What would you do? I'd shoot RAW & underexpose it in processing.
I'm way out-geeked here. Time to watch the game! :D

[Robert_Lay]

... Now take exposures of both, metering and exposing "correctly" for both and using auto white ballance, if using DSLR.
Besides a slight color cast the lightness of both images will be darn near the same, even though you perceive the one in the box to be darker.



Right there is where the whole thing came off the tracks.

You have metered and exposed correctly for both, which I interpret as "correctly for each". Unless I misinterpret what you actually did, you exposed each piece of paper individually according to what your camera's internal metering called for.
Well, if that's what you did, why would they not come out the same in the virtual images? Everything that you expose "properly" is "supposed" to come out at the shade called middle gray. That's the nature of the metering philosophy and the very reason why the Zone System was invented.

[DocFrankenstein]

The answer depends on how much psychotrophic substances you have consumed before that sunset.

I'd shoot whatever you want me to shoot, making sure I get both shadow and highlight detail and the pull the thing in CS until it looks like I want it to look.

You can't afford to be nerdy, unless it's productive. ;)

[rdenney]

What I want to know is how to determine an exposure at noon of the grey card so that when viewing it in print it would have the same tonality, under a standard viewing condition, as what I observed at noon that day. And as well for the situation at 9:30.

There are several things to remember. One is that seeing is one of five senses, and noon FEELS different than dusk even with your eyes closed. You cannot divorce your senses.

The card in noon sun also seems brighter because your pupil is much smaller to keep the scene from causing pain. Your brain knows the pupil is smaller, and it sends "BRIGHT!" signal to your thought processes.

Also, our entire experience is built on the notion that flat light is less bright than contrasty light. Thus, the flat light of dusk seems darker, because our eyes expose to put the histogram in the middle of our senses, rather than "pushing it to the right". At noon, our brain is gain-riding the pupil depending on where we look to manage the brightness of the scene, but it's having to do that because the histogram is always biased to the right.

Further, dark scenes must use more of the cones of the retina, which are monochromatic and which lack the same acutance of the rods (maybe I have those backwards, heh, heh), which are used in bright scenes.

And I don't think you can discount the effect of the color. Dusk light is far bluer than noon sunlight, and that limits the chromatic range of the scene. It's another signal for our brain to use to come to the conclusion, "DUSK!".

If we want to get the FEEL of dusk in a photograph, then we have to expose it the way we perceive it, not the way we see it. That means leaving the contrast flat, the shadows inky, and the highlights a bit down. Even correcting the white balance from blue to neutral will make a dusk shot look like a heavy overcast mid-day shot, so if we want it to look post-sunset, we need to leave it blue.

In this image, I left the image blue, and there are no highlights that reach Zone 10 to speak of. It's a real dusk shot, and I wanted it to look like it.

http://www.rickdenney.com/images/tree_against_sunset_lores.jpg

Here's one that was also taken in deep dusk, but I pulled the blue out and turned up the contrast to make it look like mid-day overcast rather than dusk.

http://www.rickdenney.com/images/fort-niagra-thru-trees-lore.jpg

In this image, I biased it blue and pushed the histogram more to the left. It's much darker than it was in real life, giving a dusk effect even though the sun was still up. There's no yellow in this shot and the only contrast comes from the specular highlights.

http://www.rickdenney.com/images/Seattle-sailboat-lores.jpg

(First image was taken with an Arsat 30mm fisheye on an Exakta 66 with Fuji Reala, second image was with the 10D, Sigma 12-24, 1/15 at f/5.6 and ISO 200. The third image was a 10D, 70-200/4L, 1/2000 at f/16 and ISO 400.)

Thus, if you adjust that gray card to be blue, it will look darker than the neutral gray card even at the same density. But to get the effect of dusk as perceived by the brain, it needs to be down a stop or so, too, along with the surrounding scene being reduced in contrast and pushed to the left in the histogram.

Rick "who thinks it impossible to isolate one sense from the others" Denney

[Nidz]

http://www.rickdenney.com/images/tree_against_sunset_lores.jpg

I love that pic!

[UncleDoug]

I think taking the EV of the light into consideration will give a better indications of the lightness or the darkness of an image since EV is absolute and not relative. All we need is for someone to come up with a mapping of the EV to what our eyes perceive (and then translated into a 0 to 255 gray-scale levels) since there obviously won't be a one for one relationship between the two. I'm no expert here so I'm just speculating but I would think that a gray card would look almost the same (after adaptation and ignoring color cast) to the eye no matter what the lighting is as long as the pupils are not fully dilated or closed down. But in very dim lighting or very bright lighting where are pupils are at their extremes, what the eye perceives would vary with the amount of light falling on the card.

This is what I had in mind.

Working on another example and a possible solution....

[UncleDoug]

PacAce,

Thanks for the directional change.
Getting somewhere on this....
All about EV's.:D

[Robert_Lay]

PacAce,

Thanks for the directional change.
Getting somewhere on this....
All about EV's.:D

Then you might be interested in my PhotoTool program, available at
http://zaffora.f2o.org/W9DMK/Programs
Look for PhotoTool.zip.
It has an extensive Help file which provides a thorough treatment of EV and how to combine EV, Aperture, Shutter Speed, ISO, etc into a meaningful relationship. Best of all, the program displays the resulting digital exposure as an index on a gray scale along with the value as a number from 0 - 255. The transfer curve used can be a hypothetical relationship or a calibration curve made from tests with your own camera.

[PacAce]

Then you might be interested in my PhotoTool program, available at
http://zaffora.f2o.org/W9DMK/Programs
Look for PhotoTool.zip.
It has an extensive Help file which provides a thorough treatment of EV and how to combine EV, Aperture, Shutter Speed, ISO, etc into a meaningful relationship. Best of all, the program displays the resulting digital exposure as an index on a gray scale along with the value as a number from 0 - 255. The transfer curve used can be a hypothetical relationship or a calibration curve made from tests with your own camera.
Bob, you never cease to amaze me with all these neat software you've written. I'm really impressed. :)

[DocFrankenstein]

WOW

Rdenney - that is an amazing piece of info. I had a gut feeling that I was doing something wrong in terms of my shooting, since sometimes a day shot would turn out like dusk and vice-versa... but to explain it like that! You have many tricks in your sleeve!

I will reserve Robert's program for tomorrow when I get my melatonin levels in order.

[Robert_Lay]

Bob, you never cease to amaze me with all these neat software you've written. I'm really impressed. :)

Well, I made my living as a computer programmer for a lot of years, and it is so much fun in Visual Basic, because it's so much easier than the way we used to program computers.

[PacAce]

Well, I made my living as a computer programmer for a lot of years, and it is so much fun in Visual Basic, because it's so much easier than the way we used to program computers.Bob, I'm a computer programmer by profession myself (have been most of my life) and VB is my forte, too. But that's not what impresses me about you. It's the depth and detail of knowledge you have that's needed to start coding all your neat photgraphy related programs in the first place that really impresses me. I'd sure like to know what library you hang out in so that I can go there and start reading the same books you read. ;) :lol:

[Robert_Lay]

Bob, I'm a computer programmer by profession myself (have been most of my life) and VB is my forte, too. But that's not what impresses me about you. It's the depth and detail of knowledge you have that's needed to start coding all your neat photgraphy related programs in the first place that really impresses me. I'd sure like to know what library you hang out in so that I can go there and start reading the same books you read. ;) :lol:

That's good!
I use Border's Book Store as my library :lol:
We have one in Fredericksburg, VA (30 mi W) and one in Waldorf, MD (30 mi N). I like the one in FXB better, because you don't have to get a token from the Information Desk to use the Rest Room :eek: AND I don't have to spend $3.00 for the Potomac River Bridge.:mad:

[PacAce]

That's good!
I use Border's Book Store as my library :lol:
We have one in Fredericksburg, VA (30 mi W) and one in Waldorf, MD (30 mi N). I like the one in FXB better, because you don't have to get a token from the Information Desk to use the Rest Room :eek: AND I don't have to spend $3.00 for the Potomac River Bridge.:mad:
Holy Smoly! You drive 30 miles to get to a book store. :shock: Mine's practically 3 minutes away from my place! I should really learn to appreciate my local library/book store more. :)

[UncleDoug]

Then you might be interested in my PhotoTool program, available at
http://zaffora.f2o.org/W9DMK/Programs
Look for PhotoTool.zip.
It has an extensive Help file which provides a thorough treatment of EV and how to combine EV, Aperture, Shutter Speed, ISO, etc into a meaningful relationship. Best of all, the program displays the resulting digital exposure as an index on a gray scale along with the value as a number from 0 - 255. The transfer curve used can be a hypothetical relationship or a calibration curve made from tests with your own camera.

Sounds great!
No PC here though :( ...

[Robert_Lay]

Sounds great!
No PC here though :( ...

That was never a problem until I joined this forum - then I found that a large percentage of the members were Mac oriented, and I can see why, since graphics has been the kick-butt advantage of Macs.

Unfortunately, I have not learned how to program for the Mac, although I did look closely at a product called RealBasic, which I think is quite good, but I'm too cheap to pay their price, and they really aren't a form of Basic. The only thing in common between Real Basic and Visual Basic is that the Integrated Development environment is similar. Otherwise the syntax is more like C++.

My Help files for the PhotoTool are all HTML - so, if you would be interested in a separate package of just the ".htm files, you can pick that up here:
http://zaffora.f2o.org/W9DMK/HelpPkgPhotoTool.zip

[UncleDoug]

Right on!

Thanks for the help.
This whole thing has been a massive education for me. Hope it has been for others so far.:D

[PacAce]

Sounds great!
No PC here though :( ...
I have a Mac (PowerMac G5), too, and PhotoTool works perfectly fine on my machine. :cool: :cool: :cool:

Of course, I also have the MS Virtual PC software running on my Mac, too, and that's where PhotoTool is running. :mrgreen: :mrgreen: :mrgreen:

[UncleDoug]

I have a Mac (PowerMac G5), too, and PhotoTool works perfectly fine on my machine. :cool: :cool: :cool:

Of course, I also have the MS Virtual PC software running on my Mac, too, and that's where PhotoTool is running. :mrgreen: :mrgreen: :mrgreen:

Cheater!:D