Dark Birds over Water - Matrix or Spot Metering

[Brownbear]

Mike,

Thank you for taking the time to explain in greater detail. My incorrect assumption of the exposure compensation being applied to just the bird, although incorrect was based on my experiments/test with metering. I can darken backgrounds if I spot meter off a white bird, and lighten backgrounds if I spot meter off a dark bird, and it seems that a lot of professional bird photographers recommend spot metering over matrix metering.

Somehow I managed to run amok in my quest to figure out how to photograph light and dark birds. I will do your suggested experiment and I will also try to duplicate my spot metered bird exposures using matrix metering.

Thank you.


My sincere apologies to everyone for any confusion I have created on spot vs matrix metering

PS Chauncey, Please do not apologize and please know that I am truly appreciative that you take the time to share your your knowledge and expertise.

Sorry, Christina, but your theory isn't at all valid and I'll try to explain why. The following assumption in your theory is incorrect: spot metering on the dark bird measures the light on just on the bird, and the exposure adjustments whether + or - are seen primarily in the bird

The exposure compensation is applied to the entire scene, not just the bird.

So, let's start over so you understand exposure in the camera. You already know that the three factors are the aperture, shutter and ISO settings. Notice that all of those settings affect the entire scene (not just the bird). Similarly, when you use exposure compensation, it also affects the entire scene. That's because it compensates by adjusting either the ISO, the aperture or the shutter (depending on the various camera settings and the available light), which as already mentioned affect the entire scene.

To clarify all of that, assume you have disabled Auto ISO and enabled ISO 100 and spot metering that results in a shutter speed of 1/60 and an aperture of f/8. If you are using aperture-priority mode and you dial in +1 EV exposure compensation, the shutter will be changed from 1/60 to 1/30.

Now assume that instead of using spot metering, you used matrix metering resulting in the same shutter speed of 1/60 and the same aperture of f/8. Dialing in the same +1 EV exposure compensation will again result in the shutter being changed from 1/60 to 1/30.

As you can see, the entire scene is being affected when you use exposure compensation and it is being affected in the exact same way regardless of how your starting combination of ISO 100, aperture at f/8 and shutter speed of 1/60 were initially determined.

Make sense?

By the way, it's fine to make theories but you should test them. Doing so will help you understand how your camera works. In this case, you could have placed your camera on a tripod and spot metered a particular object in the scene. You could have noted the three settings that affect exposure, dialed in the exposure compensation, and then noted the changed settings. You could have then used matrix metering and done the same thing. You would have noted that, unlike your theory, the exposure settings would have changed exactly the same way once you dialed in the exposure compensation regardless of the metering method being used at the time.

The reason I like the idea of creating situations that test your theories is that you might gain command of your camera more readily than any amount of reading will provide. That's because reading about your camera is not the same as using your camera and noting the changes. This isn't to say that reading about your camera isn't helpful, but in the end using the camera and noting what happens might help everything sink in.

[Mike Buckley]

I can darken backgrounds if I spot meter off a white bird, and lighten backgrounds if I spot meter off a dark bird

True. However, that doesn't test your theory that exposure compensation only affects the bird.

When you conduct your tests, be sure to note the changes in the exposure settings. They are more important than viewing the changes made to the image insofar as understanding exactly what your camera does when you change a setting.

[Mike Buckley]

a lot of professional bird photographers recommend spot metering over matrix metering.

Keep in mind that I'm not a pro and that I don't photograph birds or use spot metering. Despite that, I would be willing to make the assumption that most pro bird photographers that use spot metering first gained a command of all metering methods using static scenes in which the subject is a comparatively larger portion of the scene than when photographing birds.

I have always wondered if you wouldn't be well advised to gain command of your camera's operations by photographing easier subjects before moving on to perhaps the most difficult subjects of all -- birds in flight. Considering that I have never photographed birds in flight and that I have rarely photographed birds not in flight, I'm not qualified to make a recommendation about that. But I do recommend that you ask people who avidly photograph them for their advice about that.

[Brownbear]

Mike, for sure I will do more research, ask around and test it out on my camera by viewing the exposure on the histogram.

While I was in Mexico, trying to capture the beauty of all the tropical birds, especially their beauty in flight, stimulated my interest and a new found passion for photography. It just happened that way.

I do intend to learn to do other types of photography, learn more about my camera, the other modes available, study lighting, and whatever I need to learn, including post processing.

Thank you for your help.

[Brownbear]

Hi Larry,

Yesterday afternoon I visited the horse races for another try at photographing them. This time I tried substitute metering as suggested by Richard and you. For the first try I metered off of the grass (midtone) and I think those exposures turned out a little dark using matrix metering after spot metering the grass.

For the next set I spot metered off of the white railings so I could avoid clipping in the jockey's white pants and the white fence post, setting my exposure around -1.3. The horses at the starting gate turned out a little dark, but I had clipping in the whites in the shots at the finish line (where white filled most of the scene, whereas at the starting gate white filled only a small part of the scene)

So I ended up having to adjust my exposure compensation as quickly as I could by spinning the dial as a best guess because there was no time to adjust and focus on the horses.

Is it possible to set the exposure compensation correctly for when white is a small part of the scene and when it is a large part of the scene?

Joe,

I know that you use matrix metering on your birds with great results. With white birds in flight do you set your exposure compensation as a happy medium for shots where they will fill part of the scene and the entire scene?

Do you spot meter off of the white bird (or black bird) first and set your exposure compensation based on that reading and change to matrix metering? If not, how exactly do you use matrix metering?

Thank you.

I completely agree. Spot metering the black of the bird has led the camera to overexpose and wash out the black birds. Which is why I suggested the opposite: spot meter the brightest element in the scene and, to avoid grays again, add +2 ev. You can probably do the same thing with the blacks adding -2 ev but I rarely go that way. Since the sky is often the brightest element in a scene and I value getting the sky exposed well (actually overexposed as Colin suggested), that is what I spot meter. It is pretty big, too. Hard to miss. Exposing for the sky is espoused by Bryan Peterson in the popular tome Understanding Exposure. That can be the subject of some disagreement and one needs to follow rules only as long as they are working for you. I just do not get such washed out blacks when I use my technique. The real issue for me with spot metering the blacks is that you are going to lose the sky which you did in some of these pics. I find if I prioritize the sky, the rest seems to fall in to place. Unless it doesn't. Happy there are histograms.

[jprzybyla]

Joe,

I know that you use matrix metering on your birds with great results. With white birds in flight do you set your exposure compensation as a happy medium for shots where they will fill part of the scene and the entire scene?

Do you spot meter off of the white bird (or black bird) first and set your exposure compensation based on that reading and change to matrix metering? If not, how exactly do you use matrix metering?

Thank you.

Hi Christina, when I see a white bird flying or otherwise I use the +/- EV and set -1 EV. After having shot thousands of exposure of white birds in Florida I know in bright light that is my go to setting. On dark colored birds and birds with no white I use the exposure that Nikon's Matrix Metering determines. But... I always shoot RAW and tone map in post processing for the correct exposures of the background and bird. One thing to keep in mind, there are two types of highlights, specular and non-specular. Specular highlights are reflections such as the sun on water or on a chrome car bumper, these can clip because there is no detail to preserve. The other highlight is non-specular, think feathers on a white bird. Those have detail and should be prevented from clipping.

[Brownbear]

Hi Joe,

Thank you for sharing. Truly appreciated. I hope that I too will know how to adjust my exposures by instinct. And I will be revisiting matrix metering but still studying spot metering so I can understand how and why without thinking about it too, hard.

And a BIG thank you for letting me know about specular highlights... That metal railing at the horse races caused me all kinds of problems... Next time around I will ignore it, completely

Hi Christina, when I see a white bird flying or otherwise I use the +/- EV and set -1 EV. After having shot thousands of exposure of white birds in Florida I know in bright light that is my go to setting. On dark colored birds and birds with no white I use the exposure that Nikon's Matrix Metering determines. But... I always shoot RAW and tone map in post processing for the correct exposures of the background and bird. One thing to keep in mind, there are two types of highlights, specular and non-specular. Specular highlights are reflections such as the sun on water or on a chrome car bumper, these can clip because there is no detail to preserve. The other highlight is non-specular, think feathers on a white bird. Those have detail and should be prevented from clipping.

[Brev00]

Do not put too much emphasis on some clipping. While women cannot be just a little bit pregnant, a highlight can be a little bit clipped. When I shoot raw, much of the first wave of clipped highlights are easily recoverable as are much of the clipped blacks. Processing raw files gives one more leeway and, in fact, using the recovery and exposure sliders can have a net positive result on your images. If the original exposure fills up more of the dynamic range of your sensor, you will have less noise to reduce and less loss of detail. Blinkies are like a yellow light not a red. Proceed with caution. One thing I do when there is shifting intensity of brights is use manual exposure and auto iso. In bright conditions, the iso will stay low. But, if clouds cover or the angle of shot includes less sky, the auto iso will kick in. I do this with soccer but with conditions that mainly go from bright to dark. I am not sure I would spot meter a white railing in your example, but, to follow my earlier suggestion, I would apply + compensation. That would keep the blacks from being too dark. I apply - compensation in aperture priority and matrix to keep the blinkies from red light territory. Spot metering, I find a bright spot and apply + compensation. At a horse race, I might think about using Shutter Priority mode and let my other settings fall into place from my chosen shutter speed. Like aperture mode, using + or - ev depending. If you haven't spent a lot of time in Aperture mode using + and -ev, I might practice doing that before getting too much into spot metering. It will all come in time.

[Brownbear]

Thank you Larry,

Very helpful to know, especially that while women cannot be just a little bit pregnant, a highlight can be a little bit clipped.

I'd be nervous about trying aperture priority because of possible slow shutter speeds but perhaps I will try shutter speed priority the next time around.

I tried spot metering off the white railing only because I could see some clipping in it but I will ignore that next time around.

Thanks a million for your help.

Do not put too much emphasis on some clipping. While women cannot be just a little bit pregnant, a highlight can be a little bit clipped. When I shoot raw, much of the first wave of clipped highlights are easily recoverable as are much of the clipped blacks. Processing raw files gives one more leeway and, in fact, using the recovery and exposure sliders can have a net positive result on your images. If the original exposure fills up more of the dynamic range of your sensor, you will have less noise to reduce and less loss of detail. Blinkies are like a yellow light not a red. Proceed with caution. One thing I do when there is shifting intensity of brights is use manual exposure and auto iso. In bright conditions, the iso will stay low. But, if clouds cover or the angle of shot includes less sky, the auto iso will kick in. I do this with soccer but with conditions that mainly go from bright to dark. I am not sure I would spot meter a white railing in your example, but, to follow my earlier suggestion, I would apply + compensation. That would keep the blacks from being too dark. I apply - compensation in aperture priority and matrix to keep the blinkies from red light territory. Spot metering, I find a bright spot and apply + compensation. At a horse race, I might think about using Shutter Priority mode and let my other settings fall into place from my chosen shutter speed. Like aperture mode, using + or - ev depending. If you haven't spent a lot of time in Aperture mode using + and -ev, I might practice doing that before getting too much into spot metering. It will all come in time.

[Mike Buckley]

about specular highlights... That metal railing at the horse races caused me all kinds of problems... Next time around I will ignore it, completely

When the brightness is a distraction in the image, I often use the trick of selecting that area and changing the color to something more complementary or at least less distracting, perhaps changing it to off-white or light grey. Considering that the area was devoid of detail, no detail is lost because there was no detail there in the first place. In fact, it's possible to add detail by adding noise, grain and the like to that area to make it look a bit more natural. All of that has to be done judiciously, as over doing it makes the area stand out more than in the original.

[chauncey]

"twould seem that we've about beat this thread to death but...I got this link in my mail this morning, it's worth a read. http://digital-photography-school.co...=Yahoo%21+Mail

[Brownbear]

Thank you for sharing. I found the article informative and helpful..

I'm not quite ready for this thread to die.

1. Why is it that when one is photographing the moon that with matrix metering the moon is blown but that with spot metering (without changing exposure compensation) the detail in the moon comes out? In this case using spot metering seems simpler and easier.

2. Why is it that spot metering seems to be recommended for photographing bugs?

3. Why is it that center weighted metering is recommended for portraits?

4. If the same exposure can always be achieved with matrix metering and exposure compensation (except for back-lit photos?) why bother using center weighted or spot metering at all?

"twould seem that we've about beat this thread to death but...I got this link in my mail this morning, it's worth a read. http://digital-photography-school.co...=Yahoo%21+Mail

[Brev00]

Answering these particular questions would qualify me for a book deal. They seem pretty simple, but they really are at the heart of some of the articles and books I have read about exposure. So, while I would recommend one of the many excellent such books, I can try to offer a quick response to some of these:
1. Matrix evaluates the entire scene and, creating an average, tries to return the scene to a neutral gray. So, a very dark sky with a tiny bit of moon will average out as very dark. The camera will respond by adding exposure, maybe as much as 2-3 stops, so the very dark sky becomes gray. The moon, an innocent bystander, gets added exposure and is blown to bits.
Matrix does not care about 1-2% of a photo though you might.
2. I seldom shoot bugs, but I imagine bugs often live in the opposite situation from the moon. They are the darkest part of an otherwise bright scene. If one went by matrix, the average would be bright and the camera would subtract exposure making the bug too dark. This is a situation where one cares about a small percentage of a scene.
3. In portraits, the subject often lies in the center of the frame. The portrait photographer's entire goal is getting the person rendered perfectly. She does not care about the rest of the frame so much. Center-weighted metering will evaluate the proper exposure of just the subject. That seems to me to be pretty sensible for a portrait. I shoot portraits just slightly more often than I shoot bugs and almost never use center-weighted metering.
4. Spot metering is good for scenes that have extremes, like your moon. You can use +/- ev in matrix but, for me, that is kind of hit or miss with a scene with extremes. Spot metering is the only type that allows you to evaluate areas, like the moon, that occupy a small percentage of the frame. By examining different spots, one can decide on a best exposure for the scene with full knowledge. For example, you spot meter the moon and see you need f 8 and 1/400 at iso 200 to expose it well. You also know that will make the sky look a lot darker than it is. Matrix might tell you to shoot at f 8 and 1/10 at iso 200. You don't have enough ev to compensate (-6 ev is needed. My D90 gives me -5 at most). So, spot works well with extremes. I shoot starbursts a lot. At f 22. I want to keep the sky pretty blue and not too dark or light. It is pretty straightforward to spot meter a part of the sky to expose for that tone. I don't have to worry about the massive overexposure of the sun. It will be tamed to some extent (not completely by any means) by spot metering some nearby blue sky. Much of the foreground might be in shadow but that can be lifted later if I wish. By finding the right spot, I can get the balanced exposure I am after even with all the extremes.

[rpcrowe]

I forgot to mention Romy Ocon's great Philippine Bird Photography website...

http://www.pbase.com/liquidstone

Browse through his galleries. He provides technical information for many of his great photos and you might be able to do some comparisons of these photos.

[Mike Buckley]

In addition to Larry's very helpful ideas...

1. Why is it that when one is photographing the moon that with matrix metering the moon is blown but that with spot metering (without changing exposure compensation) the detail in the moon comes out? In this case using spot metering seems simpler and easier.

When asking about photographing any particular subject, you need to take into account and explain how much of the captured frame is being occupied by the subject. As an example, your question and answers would be very different when photographing the moon using a 100mm or 600mm focal length.

4. If the same exposure can always be achieved with matrix metering and exposure compensation (except for back-lit photos?) why bother using center weighted or spot metering at all?

Always remember that the same exposure can be achieved using all three metering methods in all situations, including back-lit photos. The only reason to use one method instead of the other two in any situation is simply because one might work more predictably and easily for you.

I always use matrix metering because I know how to use it to reliably and easily achieve my desired results. Whatever metering method works the best for you in a particular situation is the one you should use. You can determine the method that works best for you only with sufficient experience.

[Brownbear]

Thank you Larry, truly appreciated and very helpful.

#4 is especially informative and helpful. I will purchase Bryan Peterson's Understanding Exposure this week.

Answering these particular questions would qualify me for a book deal. They seem pretty simple, but they really are at the heart of some of the articles and books I have read about exposure. So, while I would recommend one of the many excellent such books, I can try to offer a quick response to some of these:
1. Matrix evaluates the entire scene and, creating an average, tries to return the scene to a neutral gray. So, a very dark sky with a tiny bit of moon will average out as very dark. The camera will respond by adding exposure, maybe as much as 2-3 stops, so the very dark sky becomes gray. The moon, an innocent bystander, gets added exposure and is blown to bits.
Matrix does not care about 1-2% of a photo though you might.
2. I seldom shoot bugs, but I imagine bugs often live in the opposite situation from the moon. They are the darkest part of an otherwise bright scene. If one went by matrix, the average would be bright and the camera would subtract exposure making the bug too dark. This is a situation where one cares about a small percentage of a scene.
3. In portraits, the subject often lies in the center of the frame. The portrait photographer's entire goal is getting the person rendered perfectly. She does not care about the rest of the frame so much. Center-weighted metering will evaluate the proper exposure of just the subject. That seems to me to be pretty sensible for a portrait. I shoot portraits just slightly more often than I shoot bugs and almost never use center-weighted metering.
4. Spot metering is good for scenes that have extremes, like your moon. You can use +/- ev in matrix but, for me, that is kind of hit or miss with a scene with extremes. Spot metering is the only type that allows you to evaluate areas, like the moon, that occupy a small percentage of the frame. By examining different spots, one can decide on a best exposure for the scene with full knowledge. For example, you spot meter the moon and see you need f 8 and 1/400 at iso 200 to expose it well. You also know that will make the sky look a lot darker than it is. Matrix might tell you to shoot at f 8 and 1/10 at iso 200. You don't have enough ev to compensate (-6 ev is needed. My D90 gives me -5 at most). So, spot works well with extremes. I shoot starbursts a lot. At f 22. I want to keep the sky pretty blue and not too dark or light. It is pretty straightforward to spot meter a part of the sky to expose for that tone. I don't have to worry about the massive overexposure of the sun. It will be tamed to some extent (not completely by any means) by spot metering some nearby blue sky. Much of the foreground might be in shadow but that can be lifted later if I wish. By finding the right spot, I can get the balanced exposure I am after even with all the extremes.

[Brownbear]

Richard,

Thank you for the link. Indeed he shares a lot of detail on his photos which is helpful.

Mike,

Yes, I just didn't wish to complicate things by including two questions in one, ie; small or large part of the frame.

Thank you for teaching me that. I plan on doing that the exercise you recommended first opportunity, when it stops raining, and it is nice to know that I will eventually figure out which method works best for me, as I gain more more experience.

Okay, time to let this thread die... Thank you to all... And I should also say to Colin that I found your detailed reply very informative.

[Colin Southern]

If I use spot metering on a black bird and adjust the compensation to expose for the bird properly the effect of over or underexposing the water should be less than seen with matrix metering because spot metering on the dark bird measures the light on just on the bird, and the exposure adjustments whether + or - are seen primarily in the bird (spot metering) whereas in matrix metering the exposure adjustments are seen in the entire scene so if I increase the exposure the water is blown and if I decrease the exposure the water is dark.

Does my theory valid or is it something I dreamed up in my head?

Thank you.

Hi Christina,

Just because you can't see detail in a black bird DOESN'T mean to say it's not exposed correctly; in fact - if anything - it's an indicator that it probably IS exposed correctly (with a correct exposure a black object will appear black and a white object will appear white). So if the bird appears black - highlights appear as highlights - and midtones appear as midtones - then you exposure is correct. It's just not revealing what you want (but it is still correct).

Any metering that exposes the black detail at a higher level is "incorrect" exposure in a technical sense; that's not so say that you shouldn't use it though, but the problem is that if you bias the exposure one way or the other then it affects the whole image - regardless of the metering mode used to calculate it. By default spot metering a dark bird will give you a middle gray bird, so if it's detail in the bird that you want to see then spot metering and an EC of -1 will probably be about as close to a "mantra" of exposure that you're ever going to get - but - although that'll lift the exposure of the bird by about 1 stop, it'll also lift the exposure of everything else by about 1 stop as well - so things like sky will be considerably more washed out.

As I mentioned earlier, what you have here is a tone mapping problem, not an exposure one. Very similar to wedding photography where we don't want to lose all detail in the grooms suit but at the same time we can't over-expose it or we risk losing detail in the bride's dress due to over-exposure. The solution is really dead easy; just expose normally and then reveal as much or as little shadow detail as one requires using the fill light slider in post-processing -- it'll raise the level of the shadow areas without touching the highlights. Takes less than a second

[AB26]

http://www.myphotocentral.com/articles/camera-metering/

http://www.whatdigitalcamera.com/tec...-metering.html

Hope it will help.

Regards.

[Brownbear]

Hi Colin,

Thank you for the extra details. I will read up on the zone systems and tones, too.

Yes, I have many photos of black crows that I spot metered with blown skies, but I also have a couple of photos of crows in blue skies that I managed with detail with spot and matrix metering where the sky is still fairly blue. But I can't seem to do it consistently. I practice photographing crows hoping that I will figure it out so I can apply the same principles to capture a black Cormorant in flight.

Here is a photo of a black Cormorant on the ocean with better detail then most of my Cormorants which I spot metered off of the head. Manual (with auto iso = 3200 in bright sunlight) SS 2500 F7... Exposure comp +.67

All the tones on the histogram seem to be represented well. The primary difference in this photo is that I took it at 10am in the morning in bright sunlight instead early morning. Is it that taking the photo of a black bird (or say a black tuxedo) in bright sunlight lessens the usual tone mapping problems? (I am assuming that by tone mapping you mean capturing the different shades of black in the lower zone along with the blue water or sky)




PS Andre

Thank you for the links and graph. The 3D graph is a helpful visual, and more informative than flat graphs.

Hi Christina,

Just because you can't see detail in a black bird DOESN'T mean to say it's not exposed correctly; in fact - if anything - it's an indicator that it probably IS exposed correctly (with a correct exposure a black object will appear black and a white object will appear white). So if the bird appears black - highlights appear as highlights - and midtones appear as midtones - then you exposure is correct. It's just not revealing what you want (but it is still correct).

Any metering that exposes the black detail at a higher level is "incorrect" exposure in a technical sense; that's not so say that you shouldn't use it though, but the problem is that if you bias the exposure one way or the other then it affects the whole image - regardless of the metering mode used to calculate it. By default spot metering a dark bird will give you a middle gray bird, so if it's detail in the bird that you want to see then spot metering and an EC of -1 will probably be about as close to a "mantra" of exposure that you're ever going to get - but - although that'll lift the exposure of the bird by about 1 stop, it'll also lift the exposure of everything else by about 1 stop as well - so things like sky will be considerably more washed out.

As I mentioned earlier, what you have here is a tone mapping problem, not an exposure one. Very similar to wedding photography where we don't want to lose all detail in the grooms suit but at the same time we can't over-expose it or we risk losing detail in the bride's dress due to over-exposure. The solution is really dead easy; just expose normally and then reveal as much or as little shadow detail as one requires using the fill light slider in post-processing -- it'll raise the level of the shadow areas without touching the highlights. Takes less than a second