View Full Version : Difference between using a gray card and a light meter
colormaniac
10th of September 2009 (Thu), 11:32
Hi, I've been having difficulty understanding the purposes of using a gray card and those of using a light meter (take L-358 as an example). I appreciate any enlightenment.
Is it correct to think that a light meter cannot replace a gray card completely? The light meter doesn't say anything about white balance, which is where a gray card can help. In L-358's manual, at the end it still suggests users to buy a gray card. However, in the histogram of a picture of a gray card, you can tell whether the gray color is in the middle and adjust your exposure accordingly (with some trials and errors). Or the other way round, if you get a good exposure (of any picture) with your light meter, your histogram should be pretty good, or you don't care about what your historgram says. Then why do you need a gray card?
I'm afraid I do not articulate my problem properly, partly because I'm quite new to all these concepts and gears. I hope it still makes some sense to you. I have done some search within my limited time and didn't find very clear answer.
Thanks!
canonnoob
10th of September 2009 (Thu), 11:33
grey card is for white balance
light meter is for exposure.
colormaniac
10th of September 2009 (Thu), 11:38
When we shoot a gray card and we see that the spike is too much to the left in the histogram, we would increase the exposure so that the spike would be located in the middle. For brighter pictures, people would like the spike to be closer to the right. Is that correct? Wouldn't that be a way to get a proper exposure without a light meter?
PacAce
10th of September 2009 (Thu), 11:39
When using a light meter in reflected metering mode, a gray card is very useful because it'll provide an accurate reference for the meter (the light that's metered will be the one bounced off the gray card).
When using the light meter in incident reading mode, a gray card will not be necessary because the meter is going to read the light source directly.
A gray card used for metering does not necessarily make it a useful tool for setting WB. You have to get a gray card that is truly neutral in color in order to use it for setting WB.
canonnoob
10th of September 2009 (Thu), 11:41
When we shoot a gray card and we see that the spike is too much to the left in the histogram, we would increase the exposure so that the spike would be located in the middle. For brighter pictures, people would like the spike to be closer to the right. Is that correct? Wouldn't that be a way to get a proper exposure without a light meter?
spike? i take it you are refering to a histogram... let me put it this way.. when looking at a histogram you want the exposure to be as spread out as possible. you dont want it bunched to the right or left (unless there is a purpose for it). the higher the "spike" that means that there is a higher number of pixels at that expsure.
make sense?
gjl711
10th of September 2009 (Thu), 11:42
A gray card is used along with your light meter to set exposure. The card gives the meter something standard to adjust one. It's very useful in tricky lighting conditions.
canonnoob
10th of September 2009 (Thu), 11:42
A gray card used for metering does not necessarily make it a useful tool for setting WB. You have to get a gray card that is truly neutral in color in order to use it for setting WB.
true i forgot to mention that he would need a true 18% grey card.
krb
10th of September 2009 (Thu), 11:44
Wouldn't that be a way to get a proper exposure without a light meter?
Obviously there are plenty of different ways that you can use the camera's built-in reflective light meter to get the correct exposure. The difference is that with a handheld incident meter you will get the numbers with a simply push of a button instead of having to set an exposure, take a test shot, view the histogram, and repeat until it looks right.
The much bigger advantage of a handheld meter is that most modern designs are also flash meters that can be used to get the correct flash level from manually controlled lighting. That can be done in camera with test shots too but is much slower process.
krb
10th of September 2009 (Thu), 12:00
spike? i take it you are refering to a histogram... let me put it this way.. when looking at a histogram you want the exposure to be as spread out as possible. you dont want it bunched to the right or left (unless there is a purpose for it). the higher the "spike" that means that there is a higher number of pixels at that expsure.
make sense?
No, that's not right at all. He specifically said "when shooting a gray card" and anytime you are shooting a monochrome subject you will have a spike in the histogram. You want to make sure that there are no spikes at either end of the graph, but a spike in the middle is going to be the correct result. Just like how if you are shooting against a chromakey green background you'll see a great big spike in the green channel. If the card is truly color balanced and you have the white balance set correctly then the spike in each of RGB histograms should be at about the same position.
gonzogolf
10th of September 2009 (Thu), 12:28
grey card is for white balance
light meter is for exposure.
Wrong!! A grey card, a true grey card reflects light at 18%, if you meter off of it with your camera you should get a proper exposure. I've used them long ago when white balance was a function of your film choice. They are kind of irrelevant for that usage now that we have a histogram to work with but their primary purpose was exposure.
Wilt
10th of September 2009 (Thu), 12:28
A term like 'light meter' is a broad generalization for a class of devices. Just like 'vehicle' is a broad generalization, too...a bicycle is a vehicle, so is a Ferrari, so is a Mack truck, so is a glider, so is an Airbus.
An 'incident meter' can fully replace a gray card for exposure measurement, but not for white balance
A 'handheld reflected light meter' used with a gray card is equivalent to using an 'incident light meter' for ambient light measurement
A 'handheld light meter' may be able to do far more than any meter with a gray card...a spotmeter can determine the range of brightness zones in a scene
A gray card with meter is useless when you need to read the brightness of transmitted light, such as a gelled light shot thru a backdrop
PacAce
10th of September 2009 (Thu), 13:03
true i forgot to mention that he would need a true 18% grey card.
If the gray card is going to be used for setting WB, it doesn't have to be an 18% gray card. It can be any gray card as long as it is neutral in color. :)
usukshooter
10th of September 2009 (Thu), 13:12
grey card is for white balance
light meter is for exposure.
A grey card can absolutely be used for exposure. I had a entire class on using a grey card to get correct exposure. And this was with film so you just used the camera's light meter, no histogram. Just google "grey card exposure" and you'll get tons of articles about it.
steveathome
10th of September 2009 (Thu), 13:55
I do find an incident meter to be more consistent, as a slight movement of a grey card can alter a reading considerably.
Maybe grey cards should be in the shape of sphere's :p
gonzogolf
10th of September 2009 (Thu), 13:59
I do find an incident meter to be more consistent, as a slight movement of a grey card can alter a reading considerably.
Agreed, modern technology has pretty much made a grey card obsolete, except for reference use.
matonanjin
10th of September 2009 (Thu), 14:13
Wrong!! A grey card, a true grey card reflects light at 18%, if you meter off of it with your camera you should get a proper exposure. I've used them long ago when white balance was a function of your film choice. They are kind of irrelevant for that usage now that we have a histogram to work with but their primary purpose was exposure.
You are absolutely correct in that a true grey card provides proper exposure. But I disagree that histograms render grey cards irrelevant.
There are lighting situations that still fool built in meters. These situations include highly contrasty scenes, highly backlit ones, or scenes with significant preponderance of very light or very dark tones. In camera meters, while better than reflective meters used to be, are still reflective meters. In these examples a histogram may provide the user a very pretty graphical representation of an improper exposure.:rolleyes:
The question becomes is the histogram accurate in enough situations to not warrant carrying a grey card around. This, of course, is a matter of the importance of shooting session (important client shoot or fun vacation shot) and the experience of the photographer in handling tricky lighting, among other things. It also depends on the availability of alternative methods such as a hand held incident meter and comfort levels with such "tricks" as metering off skin tone or grass.
I rarely carry a grey card any more since I generally use an incident meter. But I certainly don't consider the grey card irrelevant.
gonzogolf
10th of September 2009 (Thu), 14:20
Matonajin,
I wouldn't for a second argue that histograms are perfect. In fact, after reading how many users on here deploy them i think they may cause as many problems as they cure. My point was with the ability to analyze and view your image on the back of the camera a competent photographer can easily do without one.
Wilt
10th of September 2009 (Thu), 14:35
Matonajin,
I wouldn't for a second argue that histograms are perfect. In fact, after reading how many users on here deploy them i think they may cause as many problems as they cure. My point was with the ability to analyze and view your image on the back of the camera a competent photographer can easily do without one.
...when it is not important to portray something at its inherent tonality -- which is often, but not always true -- a competant photographer can do a creditable yet inexact job without one.
SkipD
10th of September 2009 (Thu), 15:20
A term like 'light meter' is a broad generalization for a class of devices. Just like 'vehicle' is a broad generalization, too...a bicycle is a vehicle, so is a Ferrari, so is a Mack truck, so is a glider, so is an Airbus.
An 'incident meter' can fully replace a gray card for exposure measurement, but not for white balance
A 'handheld reflected light meter' used with a gray card is equivalent to using an 'incident light meter' for ambient light measurement
A 'handheld light meter' may be able to do far more than any meter with a gray card...a spotmeter can determine the range of brightness zones in a scene
A gray card with meter is useless when you need to read the brightness of transmitted light, such as a gelled light shot thru a backdropDitto the above 100%
If the gray card is going to be used for setting WB, it doesn't have to be an 18% gray card. It can be any gray card as long as it is neutral in color. :)Quite true, though determining which gray cards are truly neutral is not easy. I like the WhiBal card for a white balance reference.
I used to use my old Kodak gray cards, but some folks say they are not truly neutral. My Kodak gray cards are now strictly used for exposure measurement if I use them.
A grey card can absolutely be used for exposure. I had a entire class on using a grey card to get correct exposure. And this was with film so you just used the camera's light meter, no histogram. Just google "grey card exposure" and you'll get tons of articles about it.A gray card (of the right density and reflective quality) definitely can be used for exposure measurement. My old Kodak cards are about 18% gray and not shiny at all. They are representative of what one would want in a gray card for exposure measurement (along with a meter, of course).
My WhiBal cards are much lighter than the Kodak cards and cannot be used for exposure measurement without some compensation to get the exposure values correct. Of course, their intended purpose is for use as a white balance reference.
colourblindphotographer
10th of September 2009 (Thu), 16:12
Can anyone post some sample histogram? Though its not the best way of explaining it but it would really help me (for some reason)
colourblindphotographer
10th of September 2009 (Thu), 16:14
No, that's not right at all. He specifically said "when shooting a gray card" and anytime you are shooting a monochrome subject you will have a spike in the histogram. You want to make sure that there are no spikes at either end of the graph, but a spike in the middle is going to be the correct result. Just like how if you are shooting against a chromakey green background you'll see a great big spike in the green channel. If the card is truly color balanced and you have the white balance set correctly then the spike in each of RGB histograms should be at about the same position.
Is this true with low light strobing shots??
krb
10th of September 2009 (Thu), 16:30
Is this true with low light strobing shots??
The histogram is simply representing the image that was captured. It doesn't matter how bright the ambient lighting was, whether a flash or strobe was used, or anything else. Take one picture in a brightly lit room using ambient only then turn off all the lights and use strobes to light the shot and there'll be nothing in the histogram that would let you know which is which. Set up a tripod and do a long exposure shot in the darkened room and you still won't be able to tell which is which from the histograms.
Wilt
10th of September 2009 (Thu), 17:15
My WhiBal cards are much lighter than the Kodak cards and cannot be used for exposure measurement without some compensation to get the exposure values correct. Of course, their intended purpose is for use as a white balance reference.
I have been wondering for a while why someone cannot make and sell an 18% gray card which is neutral in R-G-B balance!
TMR Design
10th of September 2009 (Thu), 17:44
I have been wondering for a while why someone cannot make and sell an 18% gray card which is neutral in R-G-B balance!
Sekonic's calibration target has a neutral 18% gray card on one side.
Wilt
10th of September 2009 (Thu), 18:01
Sekonic's calibration target has a neutral 18% gray card on one side.
nuh-uh, Robert, quit trying to get me to sell my Minolta meter! :)
Wilt
10th of September 2009 (Thu), 21:42
Can anyone post some sample histogram? Though its not the best way of explaining it but it would really help me (for some reason)
As an intro to a histogram, here is a shot with a target which you would expect to create three peaks, one for black and one for middle gray and one for white...
http://i69.photobucket.com/albums/i63/wiltonw/0EV.jpg
And here is a histogram of a MacBeth color checker, at the same exposure...
http://i69.photobucket.com/albums/i63/wiltonw/MacBeth.jpg
Both shots were made according to a reading taken with a Minolta incident meter, and confirmed with a Minolta spot meter reading from the gray patch of the Macbeth card. BTW these were not shot under perfect daylight, but using a CFL in an overhead fixture, and white balanced with Lightroom using a dropper in the gray zone, so results under daylight are not going to look exactly like this. This was merely to illustrate the histogram peaks.
krb
10th of September 2009 (Thu), 22:28
Can anyone post some sample histogram? Though its not the best way of explaining it but it would really help me (for some reason)
Shot 1:
Shot of "digital grey kard" with WB set to daylight but illuminated by tungsten lighting. http://home.comcast.net/~kennethberry/WhiteBalance_Before.jpg
Shot 2:
Same "digital grey kard" and all the same settings except that shot 1 was used to set a custom WB. Now that the WB is correct the spikes in each channel are lined up.
http://home.comcast.net/~kennethberry/WhiteBalance_After.jpg
CliveyBoy
11th of September 2009 (Fri), 00:10
I have been wondering for a while why someone cannot make and sell an 18% gray card which is neutral in R-G-B balance!
I use a Douglas Grey Card for both WB and exposure, especially when flower colour or quilt colour must be precise.
Without both Exp. and WB being fairly precise, the colours can be well off.
As the spelling indicates, it is British but they ship worldwide. Mine came with a spectograph confirmation. If card does not have a true matching intensity in each of the three channels, then it's a waste of time, IMO.
http://www.photo-software.com/greycard.htm
Wilt
11th of September 2009 (Fri), 00:31
I use a Douglas Grey Card for both WB and exposure, especially when flower colour or quilt colour must be precise.
Without both Exp. and WB being fairly precise, the colours can be well off.
As the spelling indicates, it is British but they ship worldwide. Mine came with a spectograph confirmation. If card does not have a true matching intensity in each of the three channels, then it's a waste of time, IMO.
http://www.photo-software.com/greycard.htm
Thanks for info about this card. BTW, while I lamented about cards, I do have a flexible fold up from Photovision, which is both neutral and 18%
SkipD
11th of September 2009 (Fri), 02:35
I use a Douglas Grey Card for both WB and exposure, especially when flower colour or quilt colour must be precise.
Without both Exp. and WB being fairly precise, the colours can be well off.
As the spelling indicates, it is British but they ship worldwide. Mine came with a spectograph confirmation. If card does not have a true matching intensity in each of the three channels, then it's a waste of time, IMO.
http://www.photo-software.com/greycard.htmI should get one of those cards, if just for the name that's on it :rolleyes:. Seriously, it sounds like a good choice.
colormaniac
12th of September 2009 (Sat), 00:47
Is it correct to say that when the RGB spikes are all lined up, the white balance is right? I was just processing pictures in LR2. I had a picture there of a Photovision target. I took the gray color there as reference. But it looks like that I still need to adjust the tint such that the RGB spikes are closest to each other. Then the pictures look good.
Shot 1:
Shot of "digital grey kard" with WB set to daylight but illuminated by tungsten lighting. http://home.comcast.net/%7Ekennethberry/WhiteBalance_Before.jpg
Shot 2:
Same "digital grey kard" and all the same settings except that shot 1 was used to set a custom WB. Now that the WB is correct the spikes in each channel are lined up.
http://home.comcast.net/%7Ekennethberry/WhiteBalance_After.jpg
Wilt
12th of September 2009 (Sat), 09:09
Is it correct to say that when the RGB spikes are all lined up, the white balance is right? I was just processing pictures in LR2. I had a picture there of a Photovision target. I took the gray color there as reference. But it looks like that I still need to adjust the tint such that the RGB spikes are closest to each other. Then the pictures look good.
'Neutral' is the same numeric value of for each primary color in the R-G-B mix, so 'yes'. But you can only do that with nothing but a neutral tone in the full frame. Look again at my shot of the MacBeth card, to see what the presence of any other color does!
PacAce
12th of September 2009 (Sat), 09:34
Is it correct to say that when the RGB spikes are all lined up, the white balance is right? I was just processing pictures in LR2. I had a picture there of a Photovision target. I took the gray color there as reference. But it looks like that I still need to adjust the tint such that the RGB spikes are closest to each other. Then the pictures look good.
No, not necessarily. All it means when the spikes line up is that your white balance has been set so that whatever you're using as a reference will show up as gray in the pictures. You can take a non-neutral gray card and adjust the WB on the camera so that the spikes line up but that doesn't mean that your picture will look correctly color balanced. ;)
Methodical
15th of September 2009 (Tue), 10:31
Thanks for these explanations everyone as I am in the market for a grey (gray) card and this helps
Btw, I noticed everyone seems to use grey (gray) interchangably...makes for a difficult search at times.;)
Again thanks
TMR Design
15th of September 2009 (Tue), 10:33
Thanks for these explanations everyone as I am in the market for a grey (gray) card and this helps
Btw, I noticed everyone seems to use grey (gray) interchangably...makes for a difficult search at times.;)
Again thanks
Gray and Grey are effectively interchangeable. The American spelling is Gray and the British spelling is Grey. It's like Color and Colour.
k_strecker
15th of September 2009 (Tue), 11:17
I'm american, but Grey seems more accurate :)
Grey rhymes with sleigh
Gray rhymes with Day
I'm probably freakin nuts. I'm pretty sure i've used both spellings within the same paragraph at times . . .
Wilt
15th of September 2009 (Tue), 12:21
I'm american, but Grey seems more accurate :)
Grey rhymes with sleigh
Gray rhymes with Day
I'm probably freakin nuts. I'm pretty sure i've used both spellings within the same paragraph at times . . .
By dey I where my grey suit, especially when the skies are gray with overcast and I am not creating body heat because I am riding to work in my sleigh rather than walking. I should maybe stey home instead. ;)
dmward
15th of September 2009 (Tue), 13:04
Wilt,
The two images you embedded look seriously underexposed.
The Macbeth Color Checker values are way low. The white tile should read about 93% in LR develop module and the black tile should read about 14%.
When I loaded your jpg into Lightroom 2.5 it took +1 EV adjustment to get them close.
I suspect the same is true for the other.
Since this thread is asking about exposure, one option is to use something white, i.e. a paper towel, expose it until the over exposure blickys just start to show in the info panel on the camera LCD then back off about 1/3 stop.
The result is a spike just to the left of the right edge of the histogram. The white spike from the MacBeth chart, and the three section reference as well.
Here is a short tutorial page I put together on exposure and histograms: http://www.digifotografi.com/Digifotografi/Histograms_%26_Exposure.html
Wilt
15th of September 2009 (Tue), 13:27
Wilt,
The two images you embedded look seriously underexposed.
The Macbeth Color Checker values are way low. The white tile should read about 93% in LR develop module and the black tile should read about 14%.
I was merely showing three spikes on the histogram of a RAW file to show three differing tonalities, I wasn' trying to show 'proper exposure', since one post requested, "Can anyone post some sample histogram? ". Then I simply substituted a mixed tonality target (the MacBeth) to show distribution acrossed the histogram. I guess I should have put just a bit more effort into it, to show 'proper exposure' too.
dmward
15th of September 2009 (Tue), 13:38
Wilt,
What got my attention was the comment that you measured the exposure using an incident meter and then confirmed with a spot reading. Based on that, your meters are approximately a stop off. I found I had to adjust my incident meter about 2/3 stop to get a good exposure reading compared to a color checker reference shot.
When using the camera meter on a color checker filling the frame on my 5Ds shoot several exposures, each 1/3 click apart. Usually, the best on is 1/3 to 2/3 over exposed based on the camera meter.
As an aside, reading a color checker or other reference in Lightroom or similar software also depends on the color space the software is using to render the image. LR defaults to ProPhoto which is larger then either aRBG or sRGB which means different reading for the squares.
Wilt
15th of September 2009 (Tue), 13:42
Wilt,
What got my attention was the comment that you measured the exposure using an incident meter and then confirmed with a spot reading. Based on that, your meters are approximately a stop off. I found I had to adjust my incident meter about 2/3 stop to get a good exposure reading compared to a color checker reference shot.
When using the camera meter on a color checker filling the frame on my 5Ds shoot several exposures, each 1/3 click apart. Usually, the best on is 1/3 to 2/3 over exposed based on the camera meter.
As an aside, reading a color checker or other reference in Lightroom or similar software also depends on the color space the software is using to render the image. LR defaults to ProPhoto which is larger then either aRBG or sRGB which means different reading for the squares.
Interestingly, on a separate thread I used the Minolta spot meter and the Minolta incident meter and the Canon 40D all they were all within + 0.06EV of each other.
http://photography-on-the.net/forum/showpost.php?p=8605373&postcount=17
DBOi
19th of September 2009 (Sat), 14:05
sorry to be a total newb,
Are you not suppose to set WB as flash when using strobes?
krb
19th of September 2009 (Sat), 14:30
sorry to be a total newb,
Are you not suppose to set WB as flash when using strobes?
No problem, that's a good question. It depends on how accurate you need the color temp to be.
If the only source of light is the strobes and the strobes have the same WB as what the camera expects, then it'll be fine.
If you are using a Canon speedlight attached to teh camera then the speedlight will pass WB info to the camera.
But not all speedlights and strobes emit at the same color temp. And sometimes you have light coming from other sources such as a nearby window letting in some cool light from outdoors. Or maybe you are using the strobes for fill light in an outdoor setting. Or any of a thousand other situations.
In some situations it doesn't matter how accurate the WB is set and you have full creative control to get the look you want. But if you are doing product photography then an exact match is mandatory. If you are shooting at a wedding then the dress had better look white.
vBulletin® v3.6.12, Copyright ©2000-2012, Jelsoft Enterprises Ltd.