Give your photos a professional finish with sharpening in Photoshop. Learn to enhance details, create contrast, and prepare your images for print, web, and social media.
Any suggestions on how to make an accurate grey card?
regards
PeterH
In my search on the internet I came across this article – any thoughts?
Also, if I make a homemade grey card what is the best way to test that it is correct form my 300D camera?
I presume one way is to get the camera histogram to spike at the midpoint.
"My View" <no spam > wrote in message
Any suggestions on how to make an accurate grey card?
regards
PeterH
Any suggestions on how to make an accurate grey card?
this is the article http://www.bythom.com/graycards.htmI bought an 18% grey card from my local photography shop and when I photograph it, my histogram on my 350D shows a nice spike right smack dab in the middle. Card cost me about $5 I think….
"My View" <no spam > wrote in message
In my search on the internet I came across this article – any thoughts?
Also, if I make a homemade grey card what is the best way to test that it is correct form my 300D camera?
I presume one way is to get the camera histogram to spike at the midpoint.
"My View" <no spam > wrote in message
Any suggestions on how to make an accurate grey card?
regards
PeterH
Any suggestions on how to make an accurate grey card?
regards
PeterH
Any suggestions on how to make an accurate grey card?If you have Photoshop or some other graphics editor,create a page in Grayscale mode and set the RG&B values at 128, 128, 128. Have the printer print it as a B/W image ( to avoid any color bias from printing in RGB.) Take that page to a paint store and have them mix up a quart of flat or satin paint to match that exact color.
regards
PeterH
BTW, The 18% gray card only means that its REFLECTIVITY is 18%. It generally doet necessarily mean that it has NO color bias. I have compared commercial gray cards from different manufacturers and they were not the same hue, but they both gave the same (correct) reading on the camera’s light meter.
Bob Williams
My View wrote:
Any suggestions on how to make an accurate grey card?If you have Photoshop or some other graphics editor,create a page in Grayscale mode and set the RG&B values at 128, 128, 128. Have the printer print it as a B/W image ( to avoid any color bias from printing in RGB.) Take that page to a paint store and have them mix up a quart of flat or satin paint to match that exact color. Cut up some masonite into 3×5 rectangles and paint them gray. Give them as presents to your photo buddies. They will appreciate it for years to come. If you REALLy want to impress them, paint one side gray and the other side pure white.
regards
PeterH
I did that and was amazed how appreciative everyone was. The 3×5 card fits snugly in a shirt pocket so it is always handy. BTW, The 18% gray card only means that its REFLECTIVITY is 18%. It generally doet necessarily mean that it has NO color bias. I have compared commercial gray cards from different manufacturers and they were not the same hue, but they both gave the same (correct) reading on the camera’s light meter.
Bob Williams
Bob Williams wrote:
My View wrote:
Any suggestions on how to make an accurate grey card?
regards
PeterH
If you have Photoshop or some other graphics editor,create a page in Grayscale mode and set the RG&B values at 128, 128, 128. Have the printer print it as a B/W image ( to avoid any color bias from printing in RGB.) Take that page to a paint store and have them mix up a quart of flat or satin paint to match that exact color. Cut up some masonite into 3×5 rectangles and paint them gray. Give them as presents to your photo buddies. They will appreciate it for years to come. If you REALLy want to impress them, paint one side gray and the other side pure white.
I did that and was amazed how appreciative everyone was. The 3×5 card fits snugly in a shirt pocket so it is always handy. BTW, The 18% gray card only means that its REFLECTIVITY is 18%. It generally doet necessarily mean that it has NO color bias. I have compared commercial gray cards from different manufacturers and they were not the same hue, but they both gave the same (correct) reading on the camera’s light meter.
Bob Williams
All well and good…but also a might fine argument for paying your $17.95 @ B&H for three pre-made Kodak cards!! 🙂 -What a hassle!
Mark
It’s a hassle if you don’t like to experiment.
But it is FUN if you like to try new ways to custom make stuff. BTW if you get dirt or grease on a Kodak gray card ( which is paper based), you cannot easily clean it without seriously altering the reflectivity and hue. That is probably why they come in sets of three.
There is no such thing as accurate grey card and you don’t need a grey card when you measure the incident light for exposure. Grey card is for reflected light exposure measurement like the one in your 300D.
Any suggestions on how to make an accurate grey card?
Any suggestions on how to make an accurate grey card?
regards
PeterH
Any suggestions on how to make an accurate grey card?
regards
PeterH
My View wrote:
Any suggestions on how to make an accurate grey card?If you have Photoshop or some other graphics editor,create a page in Grayscale mode and set the RG&B values at 128, 128, 128. Have the printer print it as a B/W image ( to avoid any color bias from printing in RGB.) Take that page to a paint store and have them mix up a quart of flat or satin paint to match that exact color.
regards
PeterH
Cut up some masonite into 3×5 rectangles and paint them gray. Give them as presents to your photo buddies. They will appreciate it for years to come. If you REALLy want to impress them, paint one side gray and the other side pure white.
I did that and was amazed how appreciative everyone was. The 3×5 card fits snugly in a shirt pocket so it is always handy. BTW, The 18% gray card only means that its REFLECTIVITY is 18%. It generally doet necessarily mean that it has NO color bias. I have compared commercial gray cards from different manufacturers and they were not the same hue, but they both gave the same (correct) reading on the camera’s light meter.
Bob Williams
Bob Williams wrote:
My View wrote:
Any suggestions on how to make an accurate grey card?If you have Photoshop or some other graphics editor,create a page in Grayscale mode and set the RG&B values at 128, 128, 128. Have the printer print it as a B/W image ( to avoid any color bias from printing in RGB.) Take that page to a paint store and have them mix up a quart of flat or satin paint to match that exact color. Cut up some masonite into 3×5 rectangles and paint them gray. Give them as presents to your photo buddies. They will appreciate it for years to come. If you REALLy want to impress them, paint one side gray and the other side pure white.
regards
PeterH
I did that and was amazed how appreciative everyone was. The 3×5 card fits snugly in a shirt pocket so it is always handy. BTW, The 18% gray card only means that its REFLECTIVITY is 18%. It generally doet necessarily mean that it has NO color bias. I have compared commercial gray cards from different manufacturers and they were not the same hue, but they both gave the same (correct) reading on the camera’s light meter.
Bob Williams
All well and good…but also a might fine argument for paying your $17.95 @ B&H for three pre-made Kodak cards!! 🙂 -What a hassle!
There is no such thing as accurate grey card and you don’t need a grey card when you measure the incident light for exposure. Grey card is for reflected light exposure measurement like the one in your 300D. This meter will see light as a mixing "glow" of light reflected from the scene. So subject against dark background and subject against light background will give different reading due to different "average" reflectance. So a standard reference in between is created and this is the grey. 12% or 14% or 18% doesn’t matter as long as it is the mid average because usually picture taken belong to this kind of scene. Therefore if you think that the scene is too much in light tones (snow scene), compensate a "+" with exposure compensation function when using auto exposure mode or vice versa. If your scene match the reference grey after averaging, exposure usually is good. If you are not sure about it, use a grey card that match the reflectance of the meter calibration. I say good exposure because there is no perfect exposure unless you are experience enough to fine tune one. That is why professionals prefer the incident light meter for exposure reading but use it properly and correctly or my advice is to use reflected light meter.
"My View" <no spam > wrote in message
If you are
using digital you may as well buy a grey card for colour correction, take a picture with the grey card in the scene, then carry on taking pictures in the same conditions. If lighting conditions change, take another picture of the card. When you get home you can use the grey card to correct the colour by clicking on the grey card with the eye dropper tool.
"kctan" wrote:
There is no such thing as accurate grey card and you don’t need a grey card
when you measure the incident light for exposure. Grey card is for reflected
light exposure measurement like the one in your 300D.
That is complete nonsense.
A grey card allows you to make a reading of the incident light falling on a scene using a reflected light meter such as the one in an SLR camera.
There is no such thing as accurate grey card and you don’t need a grey card
when you measure the incident light for exposure. Grey card is for reflected
light exposure measurement like the one in your 300D. This meter will see light as a mixing "glow" of light reflected from the scene. So subject against dark background and subject against light background will give different reading due to different "average" reflectance. So a standard reference in between is created and this is the grey. 12% or 14% or 18% doesn’t matter as long as it is the mid average because usually picture taken belong to this kind of scene. Therefore if you think that the scene is
too much in light tones (snow scene), compensate a "+" with exposure compensation function when using auto exposure mode or vice versa. If your
scene match the reference grey after averaging, exposure usually is good. If
you are not sure about it, use a grey card that match the reflectance of the
meter calibration. I say good exposure because there is no perfect exposure
unless you are experience enough to fine tune one. That is why professionals
prefer the incident light meter for exposure reading but use it properly and
correctly or my advice is to use reflected light meter.
"My View" <no spam > wrote in message
That isn’t how you make a grey card.
I suggest you go to a shop and ask to see a grey card, measure the light reflecting from it (it should be in the middle of the meter) then take a reading from the palm of you hand in the same lighting and check the difference in the reading (probably about 2/3rds of a stop). Leave the card
in the shop and use your hand as your calibrated piece of equipment, by making sure it is producing the same reading as it did compared to the grey
card when you take the new reading (about 2/3rds of a stop). If you are using digital you may as well buy a grey card for colour correction, take a
picture with the grey card in the scene, then carry on taking pictures in the same conditions. If lighting conditions change, take another picture of
the card. When you get home you can use the grey card to correct the colour
by clicking on the grey card with the eye dropper tool.
My View wrote:
Any suggestions on how to make an accurate grey card?
I believe it is 50% black and 50% white.
Just buy a sudio roll of seamless grey paper (backdrop) cut out a section and paste it to a board. Any size you want and you get a nice backdrop roll too which, as it’s used up, can be converted to grey cards for friends, family, profit and fun.
Cheers,
Alan
My View wrote:
Any suggestions on how to make an accurate grey card?
regards
PeterH
How about picking up a free one? Go to your local Sears or Sears Hardware store. In the paint department pick up the paint chip called "Pewter". Its number is KK331.
That chip measures LAB:
L = 68.28
a = -0.20
b = 0.06
That is VERY neutral in hue. The L means that it is a bit lighter than "middle" gray. I use Curvemeister to quickly and easily pin this neutral gray chip in Photoshop. It is a very fast way to do color correction.
I often hold this small chip in the picture and crop it out after color correction. Sometimes I take a picture of the chip in the light of the picture in a separate shot. In Photoshop I pin the gray and save the curve. Then I load that curve in the rest of the shots from that scene. It makes color correction quick and easy.
The beauty of this neutral gray is that it is FREE! Just go pick one up. Hey, pick up a bunch of them. Now you don’t care if it gets wet, damaged, or lost – there are a bunch more available for the same price.
————————–
I have found another neutral gray that gives a lot of flexibility. Golden Neutral Gray N5 acrylic paint.
http://www.goldenpaints.com/products/color/heavybody/colors/ 1445infopg.php
I went to my local art supply store and bought a 2 oz. tube of this paint. (Not all carry Golden, but there are plenty of Web places to buy it too.) Now I can paint anything I want to be my gray card. Actually the one I use the most is the lid of a metal AOL case that I painted the inside of the cover with this paint.
Golden nicely tells us what the LAB color is:
L = 52.02
a = -0.21
b = -0.28
That is about as neutral of a gray that you will find. It makes excellent gray cards. Of course, I have a pin for this in Curvemeister. You could do color correction with this gray and without Curvemeister.
Keep in mind that LAB 50% L is NOT the same as 18% reflectance. It isn’t even the same as RGB 128,128,128. I’m not going to go into the science of color in CIELAB. Read the book "Photoshop LAB Color" for all that. Actually, Golden N3 or N4 may be closer to an 18% reflectance card. You’ll have to experiment with that; I don’t care.
Nevertheless, you should be able to get neutral colors with either of these grays in your picture. You may have to adjust the overall brightness (L), but it should be consistent. Well, with one caveat…
Watch out how you hold these in your shot. These, or any other, can reflect some glare that will affect your correction. They are both a semi-matte and not a full matte. I find that 45 degrees from the source of the light works pretty well.
Have fun.
Thanks,
Clyde
Or you use a white foam hot coffee cup to get perfect white balance every time, and you can get a boatload of them at Sam’s Club for a couple of dollars.
Any suggestions on how to make an accurate grey card?
regards
PeterH
Clyde wrote:
My View wrote:
Any suggestions on how to make an accurate grey card?
regards
PeterH
How about picking up a free one? Go to your local Sears or Sears Hardware store. In the paint department pick up the paint chip called "Pewter". Its number is KK331.
That chip measures LAB:
L = 68.28
a = -0.20
b = 0.06
That is VERY neutral in hue. The L means that it is a bit lighter than "middle" gray. I use Curvemeister to quickly and easily pin this neutral gray chip in Photoshop. It is a very fast way to do color correction.
I often hold this small chip in the picture and crop it out after color correction. Sometimes I take a picture of the chip in the light of the picture in a separate shot. In Photoshop I pin the gray and save the curve. Then I load that curve in the rest of the shots from that scene. It makes color correction quick and easy.
The beauty of this neutral gray is that it is FREE! Just go pick one up. Hey, pick up a bunch of them. Now you don’t care if it gets wet, damaged, or lost – there are a bunch more available for the same price.
————————–
I have found another neutral gray that gives a lot of flexibility. Golden Neutral Gray N5 acrylic paint.
http://www.goldenpaints.com/products/color/heavybody/colors/ 1445infopg.php
I went to my local art supply store and bought a 2 oz. tube of this paint. (Not all carry Golden, but there are plenty of Web places to buy it too.) Now I can paint anything I want to be my gray card. Actually the one I use the most is the lid of a metal AOL case that I painted the inside of the cover with this paint.
Golden nicely tells us what the LAB color is:
L = 52.02
a = -0.21
b = -0.28
That is about as neutral of a gray that you will find. It makes excellent gray cards. Of course, I have a pin for this in Curvemeister. You could do color correction with this gray and without Curvemeister.
Keep in mind that LAB 50% L is NOT the same as 18% reflectance. It isn’t even the same as RGB 128,128,128. I’m not going to go into the science of color in CIELAB. Read the book "Photoshop LAB Color" for all that. Actually, Golden N3 or N4 may be closer to an 18% reflectance card. You’ll have to experiment with that; I don’t care.
Nevertheless, you should be able to get neutral colors with either of these grays in your picture. You may have to adjust the overall brightness (L), but it should be consistent. Well, with one caveat…
Watch out how you hold these in your shot. These, or any other, can reflect some glare that will affect your correction. They are both a semi-matte and not a full matte. I find that 45 degrees from the source of the light works pretty well.
Have fun.
Thanks,
Clyde
Do they tell you what illuminant that is under? You have to be careful that a gray reference used for color correction doesn’t suffer from metamerism. Kodak gray cards have a very flat spectral response and therefore exhibit virtually no metamerism.
There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.
There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.Next time you are in a car lot at night, have a look around at how poorly matched the panels on repaired cars look under street lighting. They are perfect by day. Pretty much all pigments are like that, including the pigments in paint. 😉
Bob wrote:http://www.goldenpaints.com/products/color/heavybody/colors/ 1445infopg.php
Clyde wrote:
My View wrote:
Any suggestions on how to make an accurate grey card?
regards
PeterH
How about picking up a free one? Go to your local Sears or Sears Hardware store. In the paint department pick up the paint chip called "Pewter". Its number is KK331.
That chip measures LAB:
L = 68.28
a = -0.20
b = 0.06
That is VERY neutral in hue. The L means that it is a bit lighter than "middle" gray. I use Curvemeister to quickly and easily pin this neutral gray chip in Photoshop. It is a very fast way to do color correction.
I often hold this small chip in the picture and crop it out after color correction. Sometimes I take a picture of the chip in the light of the picture in a separate shot. In Photoshop I pin the gray and save the curve. Then I load that curve in the rest of the shots from that scene. It makes color correction quick and easy.
The beauty of this neutral gray is that it is FREE! Just go pick one up. Hey, pick up a bunch of them. Now you don’t care if it gets wet, damaged, or lost – there are a bunch more available for the same price.
————————–
I have found another neutral gray that gives a lot of flexibility. Golden Neutral Gray N5 acrylic paint.
I went to my local art supply store and bought a 2 oz. tube of this paint. (Not all carry Golden, but there are plenty of Web places to buy it too.) Now I can paint anything I want to be my gray card. Actually the one I use the most is the lid of a metal AOL case that I painted the inside of the cover with this paint.
Golden nicely tells us what the LAB color is:
L = 52.02
a = -0.21
b = -0.28
That is about as neutral of a gray that you will find. It makes excellent gray cards. Of course, I have a pin for this in Curvemeister. You could do color correction with this gray and without Curvemeister.
Keep in mind that LAB 50% L is NOT the same as 18% reflectance. It isn’t even the same as RGB 128,128,128. I’m not going to go into the science of color in CIELAB. Read the book "Photoshop LAB Color" for all that. Actually, Golden N3 or N4 may be closer to an 18% reflectance card. You’ll have to experiment with that; I don’t care.
Nevertheless, you should be able to get neutral colors with either of these grays in your picture. You may have to adjust the overall brightness (L), but it should be consistent. Well, with one caveat…
Watch out how you hold these in your shot. These, or any other, can reflect some glare that will affect your correction. They are both a semi-matte and not a full matte. I find that 45 degrees from the source of the light works pretty well.
Have fun.
Thanks,
Clyde
Do they tell you what illuminant that is under? You have to be careful that a gray reference used for color correction doesn’t suffer from metamerism. Kodak gray cards have a very flat spectral response and therefore exhibit virtually no metamerism.
There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.
I’m not sure how they make paint chips, but it isn’t the paint splashed on card stock.
I suspect that it is a very careful CMYK process. I doubt they could afford anything else. (OTOH, with the volumes they do, they might get spot color inks made up.) However, it is probably as accurate a CMYK printing as you are likely to find. The whole point of these is that they HAVE to be accurate colors.
They also have to be accurate colors IN ANY LIGHT. They don’t make separate paint chips for indoor colors vs. outdoor colors. The same paint chip has to work in any lighting situation or it’s useless.
Metamerism isn’t just the perception of different colors in different lighting. It is actual color shifts. This comes from minute variations in the color layer that was laid down. Those variations may be from different reflectance characteristics of the dots of ink or even their colors. This is a problem with inkjet output. However, I’ve never heard of it being a problem with quality CMYK printing. It can’t be a problem in spot printing.
Of course, the easy thing to do is to try using the Sears Pewter (KK331) paint chip and see if it works for you. I have been using it for months in just about every lighting situation you can think of. (I sell engineered interior and exterior lighting systems and take pictures of everything you can think of.) I have seen no metamerism at all with its use.
If you really want to be sure, use the Golden Neutral Gray N5 and paint whatever you want. You certainly don’t get metamerism with paint either.
Yes, you can get glare. You can get glare with just about anything less that velvet. If it’s not too bad, it only affects the saturation and brightness and NOT the hue. As long as you have the hue on a color standard, you can do digital color correction. Of course, you can usually see when you have glare and reshoot.
Oh, I have a gray card from my old film days. It is a bit more matte than these. However, it is NOT gray. It’s a brownish gray. It’s also bigger and harder to carry around than a nice little paint chip.
These two neutral grays work and work well. I don’t try to do the color correction in the camera. They do work very well for fast and easy color correction in Photoshop. For me that is critical to taking pictures of lighting so that the color of the lights is shown in the picture. Give it a try.
I’m not claiming these are the only true neutral grays that are available. For now, they are the only two that I know about AND have been tested to prove that. If anyone else has any other grays that have been accurately tested and shown to be neutral, I would certainly like to know. Please do tell.
Clyde
Clyde wrote:
There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.
The idea of the paint chips is that you take them to wherever is going to be painted and look at them in that light.
Andrew
In article , Clyde
writes
There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.Next time you are in a car lot at night, have a look around at how poorly matched the panels on repaired cars look under street lighting. They are perfect by day. Pretty much all pigments are like that, including the pigments in paint. 😉
Good logical argument. However, it doesn’t go far enough. If that happened, Sears wouldn’t have repeat paint customers. That would be a major blow to business. You would likely hear about an issue like that. Since there hasn’t been any news like that and millions of people buy Sears’ and another companies’ paints every day, we can logically presume that the paint and the chips don’t have enough metamerism to matter.There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.
How about if the paint itself has metamerism? If the paint does and the sample doesn’t, then you’re going to have those same pissed off customers.
Now you got me curious. I wonder how they do make those chips. I suppose they could have a nice big inkjet type printer that lays down paint instead of ink.I’m not sure how they make paint chips, but it isn’t the paint splashed on card stock.
Don’t know what they do now but at one time it was exactly that. More precisely large sheets were coated, cut up, and the cut pieces attached to the sample card.
I’m not sure what you are arguing about here. I think you are agreeing with me. Of course, that is why they make the paint chips – so you can bring that exact color to your lighting to match up with other colors on location. That’s the same reason you would take that known, standard color to photographic scenes.I suspect that it is a very careful CMYK process. I doubt they could afford anything else. (OTOH, with the volumes they do, they might get spot color inks made up.) However, it is probably as accurate a CMYK printing as you are likely to find. The whole point of these is that they HAVE to be accurate colors.
They also have to be accurate colors IN ANY LIGHT. They don’t make separate paint chips for indoor colors vs. outdoor colors. The same paint chip has to work in any lighting situation or it’s useless.
Yes, it does. But it has to match what the _paint_ does under that light.
Yeah, I’ve been corrected on this. However, it doesn’t seem to be a problem with this paint in the wide variety of lighting that I shoot. So, I’ll ignore it.Metamerism isn’t just the perception of different colors in different lighting. It is actual color shifts. This comes from minute variations in the color layer that was laid down. Those variations may be from different reflectance characteristics of the dots of ink or even their colors. This is a problem with inkjet output. However, I’ve never heard of it being a problem with quality CMYK printing. It can’t be a problem in spot printing.
Metamerism isn’t a "problem", it is just something that is.
Of course, the easy thing to do is to try using the Sears Pewter (KK331) paint chip and see if it works for you. I have been using it for months in just about every lighting situation you can think of. (I sell engineered interior and exterior lighting systems and take pictures of everything you can think of.) I have seen no metamerism at all with its use.
If you really want to be sure, use the Golden Neutral Gray N5 and paint whatever you want. You certainly don’t get metamerism with paint either.
You don’t? You might want to google "paint metamerism" where you’ll find that metamerism of paint is a known problem in all industries in which matching the colors of paints is necessary–the match in the store turns out to not be a match in the field.
Horray for him or her. Since I am a semi-intelligent person, I would rather use my scissors on a free paint chip than an expensive, purchased gray card.Yes, you can get glare. You can get glare with just about anything less that velvet. If it’s not too bad, it only affects the saturation and brightness and NOT the hue. As long as you have the hue on a color standard, you can do digital color correction. Of course, you can usually see when you have glare and reshoot.
Oh, I have a gray card from my old film days. It is a bit more matte than these. However, it is NOT gray. It’s a brownish gray. It’s also bigger and harder to carry around than a nice little paint chip.
Some time prior to the Third Century BC, some enterprising fellow, possibly an Egyptian, invented a most remarkable mechanism called "scissors".
These two neutral grays work and work well. I don’t try to do the color correction in the camera. They do work very well for fast and easy color correction in Photoshop. For me that is critical to taking pictures of lighting so that the color of the lights is shown in the picture. Give it a try.
I’m not claiming these are the only true neutral grays that are available. For now, they are the only two that I know about AND have been tested to prove that. If anyone else has any other grays that have been accurately tested and shown to be neutral, I would certainly like to know. Please do tell.
Clyde
Kennedy McEwen wrote:But where are you going to get that digital definition of grey from? You don’t have any information about the spectral reflectivity of the painted surface, although you could measure it with the right equipment. So your painted grey could look more yellow under tungsten than a proper spectrally flat grey – which would mean you would be correcting everything else in the image.
In article , Clyde
writes
There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.Next time you are in a car lot at night, have a look around at how poorly matched the panels on repaired cars look under street lighting. They are perfect by day. Pretty much all pigments are like that, including the pigments in paint. 😉
I stand corrected on a fine technical point. However, it doesn’t change the use of a paint chip for color correction in digital photography. If you know exactly what the digital definition is of that gray, you will be able to use it to correct a picture taken in any light from that gray.
For photographic purposes, the Sears Pewter (KK331) paint chip works for digital color correction. So, does Golden’s Neutral Gray N5 acrylic paint. Try it, you’ll like it.You’re welcome, but I think I’ll stick with Kodak grey myself, thanks. At least it is supposed to conform to a spectral flat standard. 😉
Hey, I’ve just been trying to answer the original question of how to make your own gray card. I’m sorry I got embroiled in stupid newsgroup tangents and trivialities. I think I did give the only tested answer to the question.The problem is that your response wasn’t as correct as some of the others – get a Kodak grey card. Metamerism is just too easy to get burned by and sometimes it can be expensive and/or embarrassing. My memories of the details are hazy now, but I recall NASA made a similar mistake with the first Viking lander mission – and it was only after someone noticed that the colour of some cables in a few of the images weren’t quite right that they realised Mars has red skies at noon, not just sunset! 😉
However, it doesn’t change the use of a paint chip
for color correction in digital photography.
If you know exactly what the digital definition is of that gray,
you will be able to use it to correct a
picture taken in any light from that gray.
Clyde wrote:
However, it doesn’t change the use of a paint chip
for color correction in digital photography.
What ever surface that is spectrally decently flat will do as a graycard. Most often I use the normal Xerox copy-paper, it has the desired property and is readily available.
If you know exactly what the digital definition is of that gray,
In this context there is no single Lab (or RGB or CMYK) digital definition for any graycard. The digital code that the graycard is mapped to in your photo depends mainly 1) on how the scene was exposed and 2) on what is the most light/bright surface/object in that scene. So, imaging is relative.
Spectrally flat surface that was photographed under daylight (D65) should have R=G=B in any normal RGB working spaces. But the absolute R=G=B level varies according to the above.
In case the photo was taken e.g. under incandescent illumination (yellowish, 2850K) then there are mainly two choises for you:
1) do you want to conserve that information/appearance in your photo
-or-
2) do you want to edit the image in such way that it looks as if it was shot under daylight.
In case of (2) your aim is again the R=G=B. In case of (1) things get much more difficult as the current color science does not yet have an answer for this (so you need to do it by visual feedback).
you will be able to use it to correct a
picture taken in any light from that gray.
With a graycard you can always correct the gray-balance and _if_ the photo was taken under a light source that is not far away from the blackbody radiator (e.g. sun, incandescent light bulb, flash) then the colors usually get improved also, more or less.
But you really can not _correct_ the colors with just the graycard, you can only adjust/modify them with it. For accurate colors you need to have colorimetricly accurate workflow for that particular illumination situation where the photo was shot.
Also, when gray-balancing is done in a non-linear colors spaces like the sadRGB or AdobeRGB it will result large color errors.
Timo Autiokari
http://www.aim-dtp.net
In article , ClydeOne place you could get the correct definition of that gray is from my first message. I told you want it was: Lab(68.28,-0.20,0.06) I also told you what it was for Golden N5.
writes
Kennedy McEwen wrote:But where are you going to get that digital definition of grey from? You don’t have any information about the spectral reflectivity of the painted surface, although you could measure it with the right equipment. So your painted grey could look more yellow under tungsten than a proper spectrally flat grey – which would mean you would be correcting everything else in the image.
In article , Clyde
writes
There is no metamerism with the Sears paint chip. You won’t get any metamerism with any paint chip. It would be useless as a paint sample if you did. Paint chips have to be pretty darn accurate or you are going to get a lot of pissed off customers after a paint job.Next time you are in a car lot at night, have a look around at how poorly matched the panels on repaired cars look under street lighting. They are perfect by day. Pretty much all pigments are like that, including the pigments in paint. 😉
I stand corrected on a fine technical point. However, it doesn’t change the use of a paint chip for color correction in digital photography. If you know exactly what the digital definition is of that gray, you will be able to use it to correct a picture taken in any light from that gray.
I’m sure the Kodak gray card would make a nice standard too. Actually any color should work just fine. As long as you can adjust a picture to a known, standard color in the picture, the rest of the picture’s colors should fall right into place. So, what is the LAB definition of the Kodak gray card?For photographic purposes, the Sears Pewter (KK331) paint chip works for digital color correction. So, does Golden’s Neutral Gray N5 acrylic paint. Try it, you’ll like it.You’re welcome, but I think I’ll stick with Kodak grey myself, thanks. At least it is supposed to conform to a spectral flat standard. 😉
Hey, I’ve just been trying to answer the original question of how to make your own gray card. I’m sorry I got embroiled in stupid newsgroup tangents and trivialities. I think I did give the only tested answer to the question.The problem is that your response wasn’t as correct as some of the others – get a Kodak grey card. Metamerism is just too easy to get burned by and sometimes it can be expensive and/or embarrassing. My memories of the details are hazy now, but I recall NASA made a similar mistake with the first Viking lander mission – and it was only after someone noticed that the colour of some cables in a few of the images weren’t quite right that they realised Mars has red skies at noon, not just sunset! 😉
Clyde wrote:
If you know exactly what the digital definition is of that gray,
In this context there is no single Lab (or RGB or CMYK) digital definition for any graycard. The digital code that the graycard is mapped to in your photo depends mainly 1) on how the scene was exposed and 2) on what is the most light/bright surface/object in that scene. So, imaging is relative.
Spectrally flat surface that was photographed under daylight (D65) should have R=G=B in any normal RGB working spaces. But the absolute R=G=B level varies according to the above.
In case the photo was taken e.g. under incandescent illumination (yellowish, 2850K) then there are mainly two choises for you:
1) do you want to conserve that information/appearance in your photo
-or-
2) do you want to edit the image in such way that it looks as if it was shot under daylight.
In case of (2) your aim is again the R=G=B. In case of (1) things get much more difficult as the current color science does not yet have an answer for this (so you need to do it by visual feedback).
you will be able to use it to correct a picture taken in any light from that gray.
With a graycard you can always correct the gray-balance and _if_ the photo was taken under a light source that is not far away from the blackbody radiator (e.g. sun, incandescent light bulb, flash) then the colors usually get improved also, more or less.
But you really can not _correct_ the colors with just the graycard, you can only adjust/modify them with it. For accurate colors you need to have colorimetricly accurate workflow for that particular illumination situation where the photo was shot.
Also, when gray-balancing is done in a non-linear colors spaces like the sadRGB or AdobeRGB it will result large color errors.
Timo Autiokari
http://www.aim-dtp.net
That’s a white card, not gray. It’s fine for adjusting the white balance, but not for setting the exposure.
One place you could get the correct definition of that gray is from my first message. I told you want it was: Lab(68.28,-0.20,0.06) I also told you what it was for Golden N5.
I got the Pewter by picking up a chip and mailing it to Mike Russell. He scanned it with his calibrated spectrophotometer and he e-mailed the Curvemeister pin file with that definition to me.
I got the Golden N5 definition by looking on Golden’s Web site where they told me what the LAB definition was. You aren’t going to get more accurate than that.
The problem with RGB is that is changes depending on the Color Space that is being used.
That is not an issue with LAB. LAB color definition
IS a standard defined by CIE. (Hence, the real name "CIELAB".)
Color science is a science and is NOT relative.
LAB has no color spaces or variation based on output. It is one single standard for defining color.
J. Clarke wrote:
That’s a white card, not gray. It’s fine for adjusting
the white balance, but not for setting the exposure.
My response was for the gray-balancing/white-balancing that was the issue under the current discussion.
But it is very easy to similarly use the white card for exposure setting, one only has to know the required exposure correction. This is easily tested using a digicam or you can calculate: The gray cards have about 18% reflectance and the Xerox paper has about 85% reflectance so you set the exposure correction to Log2(1/0.18)-Log2(1/0.85)=+2.24 stops. So +2 stops or +2 1/3 stops.
However it is by far the best approach to use the automatic exposure control of the camera. It takes the scene content into account.
If you use a graycard for exposure setting and there is nothing in the scene that is say over 30% reflectance then you’ll get a seriously underexposed shot.
And if you use a graycard for exposure setting and
there is something like white clouds in the scene then you’ll get a seriously overexposed shot. The white clouds are far more bright than a surface that has 100% reflectance so they will be clipped to pure white in your photo.
Clyde wrote:
The problem with RGB is that is changes depending on the Color Space that is being used.
Well, the numbers will change also when you convert your camera images from RGB to CIELAB, do they. Also they change when you convert your camera images from RGB to CMYK or to CIE XYZ or to CIE Yxy or to CIE LCH or to any other similar color definition. It is not a problem, they simply must change. And colors will change if you define them incorrectly, 100 units of money has different purchasing power when you assign a Dollar sign to it than when you assign an Euro sign to it.
That is not an issue with LAB. LAB color definition
IS a standard defined by CIE. (Hence, the real name "CIELAB".)
RGB spaces are similar standards, I believe the CIE has standardized only one RGB space but RGB spaces generally are standardized by some standardization bodies or companies.
Color science is a science and is NOT relative.
In this context (digital imaging) color (light) is relative. Namely relative to the most white surface in your scene. The automatic exposure metering of the camera (or you in case the camera is in manual) will scale the incoming light by adjusting the aperture and/or the exposure time in such way that you’ll get a full histogram without clipping, no matter what the absolute luminance of that most white surface in the scene happens to be. See, relative.
Color "science" itself also is fully relative, again relative to the most white surface in the field of the view of your vision. For example look at the CMFs (Color Matching Functions), what is the unit of the vertical axis? It has no unit since CMFs are relative. (CMFs are the very foundation stone of color specifications, they are the functions that are used when spectral data is converted to trichromatic data, most often used CMFs are called as the CIE 2 degree Standard Observer or just simply the CIE 1931 Observer).
What is not relative is the spectral reflectance of the surfaces, they all have that absolute property even if that absolute property is expressed as unitless or in percentage.
But take a very white and a very black surface into a completely dark room, they both appear exactly the same there. Light up a candle and you will see that they are not anymore similar. Bring them under the sunlight and they appear very much different from each other. So what you see is relative to the (strength of the) scene illumination.
LAB has no color spaces or variation based on output. It is one single standard for defining color.
Not so, please read the Lab specification. Lab is relative to the scene illumination (and therefore to the Reference White), you’ll find in the equations that the CIE XYZ (that is the real trichromatic definition of color) are first normalized by the scene white and only then converted to Lab values. The fact that the Photoshop Lab space is hard-wired to D50 white is a limitation in that software (but a well considered limitation).
Timo Autiokari
http://www.aim-dtp.net
That’s a white card, not gray. It’s fine for adjusting the white balance, but not for setting the exposure.
My response was for the gray-balancing/white-balancing that was the issue under the current discussion.
But it is very easy to similarly use the white card for exposure setting, one only has to know the required exposure correction. This is easily tested using a digicam or you can calculate: The gray cards have about 18% reflectance and the Xerox paper has about 85% reflectance so you set the exposure correction to Log2(1/0.18)-Log2(1/0.85)=+2.24 stops. So +2 stops or +2 1/3 stops.
However it is by far the best approach to use the automatic exposure control of the camera. It takes the scene content into account.
If you use a graycard for exposure setting and there is nothing in the scene that is say over 30% reflectance then you’ll get a seriously underexposed shot.
And if you use a graycard for exposure setting and there is something like white clouds in the scene then you’ll get a seriously overexposed shot. The white clouds are far more bright than a surface that has 100% reflectance so they will be clipped to pure white in your photo.
Confused here. True if you don’t use a grey card because 30% scene reflectance means brighter than the calibrated 18% so average exposure reduced. If you use a grey card, then the reflectance is always 18% regardless of the average scene reflectance; your exposure will be good regardless of subject’s reflectance.
I don’t understand you. Look like you are confused. Let me give you an example. You are going to shoot pictures of a white cat, a 18% grey cat and a black cat individually against a 18% grey background. Your TTL meter will try to expose all the cats toward grey.
Remember that the TTL meter or any reflected light meter can’t see the cats or backgound but an average "glow" of 18% grey reflected light or whatever calibrated mid grey reflectance.
The graycard only rarely (only in special situations) gives good exposure.
Do we agree that a good exposure is such that has full (or nearly full) histogram without unnecessary clipping at the light end (often something in the light end simply has to be clipped, like specular reflections and direct light sources)?
For normal/typical shooting situations TTL average metering will generally do much better than what can be accomplished with the graycard, since average metering does take the scene content into account, to some extent. That is the sole purpose of the average metering modes. Of course it will fail in extreme situations (there is nothing but snow in the scene or there is nothing but very dark surfaces in the scene) but choosing the exposure by simply measuring the graycard will generally fail much more, much often.
Now, what is always forgotten in the discussions of this feature is that while the exposure is adjusted according to the 18% average there is one very important, but not very obvious, assumption that is hardwired to that adjustment: The exposure metering assumes that every scene has some fixed maximum reflectance. But in the reality the most bright surfaces/objects in our scenes are not constant from shot to shot. E.g. white clouds and many glossy and semi-glossy surfaces are more bright than what 100% diffusively reflective surfaces are. This is the reason why average metering generally does much better job than the graycard method, it takes the scene content into account (the graycard method does not consider the scene content at all). Because it is doing so it will sometimes fail, but only in special situations, like taking closeups of the white cat or the black cat.
Timo Autiokari
http://www.aim-dtp.net
By using a grey card near to the subject recieving the same amount of light will give you better reading and hence better exposure.
This is the drawback of TTL meter and lead to the designs of center weighted meter mode, pattern meter mode…multiple points reading…etc
Grey card by itself is already an average metering tool for average metering.
By the way, what do you mean by average metering takes the scene content into account? Reflected exposure meter is just a semiconductor device or voltmeter without the human’s eyes visual sense…???
"Color science is a science and is NOT relative"
Color measurement of a reflective / absorptive object (picture) IS relative to the light source.
Color measurement (by the eyes-subjective) of a monitor is also relative to the light sources- RGB from the monitor and ambient light.
Say we have a scene where max surface reflectance is at 99.6%. And we have the 18% graycard there in the scene. We set the exposure according to the graycard and perhaps the camera puts the 99.6% surface to R=G=B=254 and the 18% surface to R=G=B=46 ((the actual values depends on the gamma of the RGB space in use, these are linear, easier to understand). So, everything is OK.
Then say we have the very same scene but in addition there are white clouds that we do not want to clip to pure white (or could also be shiny surface of a car, reflexion from a window or water surface etc). The equivalent luminance of white clouds can be like 200% (or more) since they not only reflect but also diffusely transmit light. And we have the same 18% graycard there in the scene. Again we set the exposure according to the graycard …and surprisingly it will be quite the same exposure value as in the above situation. Again the camera puts the 99.6% surface to R=G=B=254 and the 18% surface to R=G=B=46 but clips the clouds to pure white. So the method fails. Where the averaging exposure measurement will take the clouds into account, mapping them to say R=G=B=254, then the 100% surface will be mapped to R=G=B=49.8 and the 18% surface to R=G=B=23.
"Timo Autiokari" wrote in message
Say we have a scene where max surface reflectance is at 99.6%. And we have the 18% graycard there in the scene. We set the exposure according to the graycard and perhaps the camera puts the 99.6% surface to R=G=B=254 and the 18% surface to R=G=B=46 ((the actual values depends on the gamma of the RGB space in use, these are linear, easier to understand). So, everything is OK.
18% grey reflectance should be 127 or 128 if the brightness range is within the recording medium ability and not 46 which is near black.
18% grey reflectance should be 127 or 128 if the brightness range is within the recording medium ability and not 46 which is near black.
The exposure is not right somewhat because mid grey becomes black (R=G=B=23).
This is what I meant when the scene brightness range (brightest area to darkest area) is wider than the recording medium ability to handle (dynamic range)
18 percent gray is very close to RGB(118,118,118) which is Lab(50,0,0).
Mike Russell wrote:
18 percent gray is very close to RGB(118,118,118) which is Lab(50,0,0).
Only in a RGB working space that have the gamma 2.2 transfer function. In the KodakProPhoto and AppleRGB color-space that have gamma 1.8 transfer function the 18% is the level 98. In linear RGB working-spaces the 18% is level 46.
But the above is just a simplification that assumes that 100% reflectance will be mapped to level 255 by the camera.
Cameras do not behave like that. If you e.g. do -2EV exposure compensation (or the averaging exposure metering of the camera does it automatically for you) then the 18% reflectance will be mapped 2 stops below, that will be level 62 in the gamma 2.2 working spaces. because of this compensation there will be 2 stops of headroom in the light end of the levels range for capturing surfaces that are more light/bright than what a 100% surface in the scene is.
I bought an 18% grey card from my local photography shop and when I photograph it, my histogram on my 350D shows a nice spike right smack dab in the middle. Card cost me about $5 I think….
Kennedy McEwen wrote:No, that tells you the colour under a standard illumination – it doesn’t give you any information about the spectral response, which is what will cause its effective colour to change under different illumination spectra.
But where are you going to get that digital definition of grey from? You don’t have any information about the spectral reflectivity of the painted surface, although you could measure it with the right equipment.One place you could get the correct definition of that gray is from my first message. I told you want it was: Lab(68.28,-0.20,0.06) I also told you what it was for Golden N5.
Not as correct? I’m not talking theoretical answers, I’m talking real world, experimental answers. I know what my gray "cards" will do. I know exactly what color of gray they are. I know they will let me correct the colors in ANY lighting situation. I know I don’t have a problem with metamerism. I know this because I have very throughly tested it and continue to do so almost every day.Correction – all you *know* is that they are good enough for you under the conditions that you have encountered so far. They may well be as good as a spectrally flat Kodak grey card, but you have done nothing to demonstrate that. Every time you encounter a new type of illumination you are pushing the envelope of your pigment to another untested region. Any errors may be small enough to be acceptable to you, for others it may not – and in some cases that can be critical, as NASA discovered about 30 years ago. 😉
18% grey reflectance should be 127 or 128 if the brightness range is within the recording medium ability and not 46 which is near black.
I wrote earlier: "the actual values depends on the gamma of the RGB space in use, these are linear, easier to understand."
The exposure is not right somewhat because mid grey becomes black (R=G=B=23).
Is very correct, just in linear quantities (that really are easier to understand in this kind of examples).
This is what I meant when the scene brightness range (brightest area to darkest area) is wider than the recording medium ability to handle (dynamic range)
No it is not, you just did not read my message or did not understand what the word ‘linear’ refers to. So below I repeat the example, now in gamma
2.2 space (that is the non-linearity that e.g. the AdobeRGB working-space
profile has):
Say we have a scene where max surface reflectance is at 99.6%. And we have the 18% graycard there in the scene. We set the exposure according to the graycard and perhaps the camera puts the 99.6% surface to R=G=B=255 so the 18% surface will be at R=G=B=117 (the RGB values are in gamma 2.2 space). So, everything is OK.
Then say we have the very same scene but in addition there are white clouds that we do not want to clip to pure white (or could also be shiny surface of a car, reflexion from a window or water surface etc). The equivalent luminance of white clouds can be like 200% (or more) since they not only reflect but also diffusely transmit light. And we have the same 18% graycard there in the scene. Again we set the exposure according to the graycard …and surprisingly it will be quite the same exposure value as in the above situation. Again the camera puts the 99.6% surface to R=G=B=255 and the 18% surface to R=G=B=117 but clips the clouds to pure white. So the method fails. Where the averaging exposure measurement will take the clouds into account, mapping them to say R=G=B=255, then the 100% surface will be mapped to R=G=B=186 and the 18% surface to R=G=B=85.
So, imaging (color) is relative and proper exposure is relative to scene white. Not to a 18% graycard.
Any suggestions on how to make an accurate grey card?
Give your photos a professional finish with sharpening in Photoshop. Learn to enhance details, create contrast, and prepare your images for print, web, and social media.
Nice and short text about related topics in discussion sections
