Question about exposing for highlights

[samueli]

Hi,

I'm trying to nail down the technical part of exposing for highlights in manual mode. I recently upgraded to a 50D and the biggest reason was for the center weighted metering option.

I can easily expose for highlights using the exposure lock, but what about manual mode? I can still meter on the highlight, but do I then re-meter with my real composition and compensate via shutter speed the difference between the two meter values? That's kind of what I'm doing now, but depending on the "oppourtunity" I may bracket as well.

[Mike]

I find the easiest way to do it is to meter any way you like, take a shot and then look at the histogram. If I'm needing more in the highlight range I know to drop the shutter speed/open the aperture/up the ISO a bit more.

Read here: http://www.luminous-landscape.com/tutorials/expose-right.shtml
and here: http://photography-on-the.net/forum/showthread.php?t=89123

[PhotosGuy]

:D Mike beat me to it! Thanks!

[yogestee]

Another way is to fill your viewfinder with the highlight or something equally as bright and meter off that..

BTW,,why do you want to expose for the highlights only??

[Mike]

:D Mike beat me to it! Thanks!

Haha! No problems Frank! :D

[samueli]

Thanks for all the replies! I thought there would a more mathematical way to do it in manual mode, without the ability to use exposure lock.

BTW,,why do you want to expose for the highlights only??

I don't really. I read this in some thread as one of the basic rules of thumb for photography - "expose for highlights". Sometimes I will expose both highlights and shadows and try a capture of each. I'm just trying to figure out a way to do this technically. Plus, I've only had spot/center weighted metering for a few weeks, with my new 50D.

I had read in another article that center weighted metering is really the only way to go, and have seen some creative uses of metering on a certain part of an image, locking the metering, then capturing the whole image where the metering would be different.

It's just that in manual mode, there is no exposure lock. So if you meter on a certain area, then recompose/focus, what is the math? In manual mode there is a missing variable to needed add/subtract stops. Or do you meter on several areas of the photo and figure out from there?

[Wilt]

Hi,

I'm trying to nail down the technical part of exposing for highlights in manual mode. I recently upgraded to a 50D and the biggest reason was for the center weighted metering option.

I can easily expose for highlights using the exposure lock, but what about manual mode? I can still meter on the highlight, but do I then re-meter with my real composition and compensate via shutter speed the difference between the two meter values? That's kind of what I'm doing now, but depending on the "oppourtunity" I may bracket as well.

BTW, your 50D has a mode which 'protects' highlights. The downside of using that is that ISO 100 is not available when the highlight protection is set.

[Mike]

It's just that in manual mode, there is no exposure lock. So if you meter on a certain area, then recompose/focus, what is the math?

There is no exposure lock because the camera is not controlling anything. Exposure lock only comes into play when you use Av/Tv/P etc as the camera is deciding at least 1 of the settings. In M you have full control. Therefore, if you meter for a particular spot and dial in the settings yourself for correct exposure it then doesn't matter where you look with the camera as the settings will remain as you dialled them in.

[tonylong]

First of all, I hope you have followed the pointers others have left you in this thread -- fundamentals and specifics about exposure have been worked through many times, and you can get more info by following earlier discussions than a single thread will produce. I'd also encourage you to look at the bottom of this page, in the "related threads" section, and read through the various threads to pick up other thoughts.

Thanks for all the replies! I thought there would a more mathematical way to do it in manual mode, without the ability to use exposure lock.

Well, I don't know how practical a "mathematical" approach is -- Ansel Adams, working with large format film and specific light meters developed the "Zone System" to identify specific "zones" of light in an image and incorporate them in his exposures, but he also did serious post-processing to fully develop his images. With our smaller/digital format we don't see the same range and you don't see much discussion about the Zone System in actual practice. The idea of metering for highlights (and/or shadows and/or midtones) is still very active, though.

A common "starting point" from the "old days" was the "Sunny 16 Rule", in which you took the idea that on a sunlit scene you could use f/16 and then a shutter speed that was 1/ISO and get a good exposure, but then you have to modify each of your settings according to the particular scene and your requirements as far as specific shutter speed needs, aperture/depth of field needs, and ISO/dynamic range or speed needs. You end up often with settings that don't seem to resemble the Sunny 16 idea much!

I don't really. I read this in some thread as one of the basic rules of thumb for photography - "expose for highlights". Sometimes I will expose both highlights and shadows and try a capture of each. I'm just trying to figure out a way to do this technically. Plus, I've only had spot/center weighted metering for a few weeks, with my new 50D.

I had read in another article that center weighted metering is really the only way to go, and have seen some creative uses of metering on a certain part of an image, locking the metering, then capturing the whole image where the metering would be different.

It's just that in manual mode, there is no exposure lock. So if you meter on a certain area, then recompose/focus, what is the math? In manual mode there is a missing variable to needed add/subtract stops. Or do you meter on several areas of the photo and figure out from there?

One of the conveniences of Manual mode is that you don't have to use exposure lock continuously -- you lock in your settings, period, and only change them when the lighting of the scene or your orientation to the lighting changes.

There are a few practical approaches to using manual that can arrive at the same place -- a good exposure that will take in highlight as well as shadow areas. This involves using your spot (or at least center weighted) meter and identifying an area of a consistent tone in the same lighting that illuminates your scene that you can reasonably identify and adjusting your settings accordingly -- telling the camera that this item is, say, medium or light or dark. You can use various things -- the palm of your hand, a gray card, a white sheet of paper, the sky, snow, as long as you know how the camera should interpret that object.

If left to automatic exposure, the camera tries to interpret everything as "medium" so that, for expample, if you let the camera expose for snow, it would turn the snow gray in the exposure. You have to know that, and, depending on the lighting of the scene, adjust your settings to interpret the snow as bright.

Once you have a balanced setting for the scene as a whole, you can shoot away unless, like I said, the scene lighting changes or your orientation to the light changes.

As far as exposing for the highlights (or in other cases shadows), you can do the above, but you are specifically metering an important highlight and using that to set your overall exposure. In the "snow" example, you want to expose the scene so that the snow will be white, but not "blown white" without any detail, so you set your camera so that the meter scele when metering the snow reads up to but not against your right edge.

Sometimes, when "exposing for highlights" you may find that other parts of the scene are darker than you would visualize, and when I'm against that type of high contrast dynamic, well, first of all RAW is a real necessity, then, I'd pay real good attention to the right side of your histogram, making sure that those highlights are really nudging the right side. Third, I would shoot that scene with the lowest ISO possible because you will get the cleanest results bringing up the shadows in your RAW converter with a low ISO.

As a whole, try to keep it simple so that you can think and act quickly. For handheld shooting, I like the statement "PhotosGuy" FrankC uses: First set your shutter speed and aperture to get the effect you want, then adjust your ISO to properly expose the scene. For shooting with a tripod with a static scene, I mix it a bit: first set the aperture and ISO you need, then adjust your shutter speed to expose the scene. But you want to keep it simple and practical!

Hope this helps a bit!

[tdodd]

My approach - Spot meter off the brightest thing in the scene for which you want to retain highlight detail - perhaps a bright white cloud, sunlit snow or a sunlit wedding dress. Set a manual exposure that gives a reading of +3 stops above the centre. Shoot in raw and fire away.

If your camera is anything like mine you should find you get about 3 1/3 stops of leeway above a centred meter before you see signs of clipping. By shooting raw you have a bit more headroom to recover highlight detail above that. Thus, by shooting with the brightest thing of interest at +3 you pretty much maximise your ETTR exposure while keeping a small reserve in case you got your metered point wrong or the light changes a fraction.

If your camera meter only shows between +/- 2 stops then set your manual exposure to +2 and then simply add a further three clicks of something or other to increase the exposure by one more stop. The meter will blink but you will know where you are.

[rdenney]

My approach - Spot meter off the brightest thing in the scene for which you want to retain highlight detail - perhaps a bright white cloud, sunlit snow or a sunlit wedding dress. Set a manual exposure that gives a reading of +3 stops above the centre. Shoot in raw and fire away.

This is exactly correct. In Zone System parlance, you are measuring bits of the scene that you want to place on Zone VIII (highlights with texture). When you meter that, the camera wants to put those bits at Zone V (middle gray). So, you increase the exposure by three stops to expose it such that it will fall on Zone VIII.

To the OP: But don't misunderstand the principle you are quoting.

In the old days of shooting negatives, we worried about underexposure, because too much underexposure would leave a clear piece of plastic. And there's nothing you can do in the darkroom to print a clear piece of plastic as anything but solid black. So, we would always make sure we had enough exposure to provide texture in the shadows. If the highlights were then too bright for the film, we'd do something (filtration, pull development, etc.) to reduce the highlights. It was easier to pull the highlights than to make something out of that clear piece of plastic. We called this exposing for the shadows and developing for the highlights.

With slides, it was the opposite. If a slide is overexposed too much, it turns into a clear piece of plastic, and that's just a blank white screen when you project it. So, we would keep the exposure low enough to make sure the highlights didn't blow out, and just let the shadows go black if necessary. Black shadows were better than blown highlights.

Deciding which of the highlights not to blow was a decision (i.e., which highlights you place on Zone VIII--which has texture--versus Zone IX--which doesn't), and such decisions require experience and a clear visualization of your final presentation.

Digital is like slides. Once the highlights reach 255, they are gone for good. So, we make sure to keep the exposure low enough to avoid blowing out the highlights. Then, we bring up the shadows and middle tones as necessary in Photoshop. (That's why we shoot RAW--it gives you more ability to pull those dark bits up). If that approach isn't workable because the shadows are too dark in relation to the highlights, then we filter, use fill flash or reflectors, or choose a composition that doesn't show or require that particular highlight detail.

So, with digital, we expose for the highlights and post-process for the shadows. That's the same thing as pushing the exposure to the right.

The histogram is the best way to do it. Make an exposure, and check the histogram. If there is space between the histogram and the right end of the scale, dial in a little positive compensation. If you see the highlight warning on the display (you do have the highlight warning feature turned on, don't you?) and the histogram is jammed up against the right end of the scale, then dial in a little negative compensation. We used to do that with Polaroids.

Rick "Yogestee knows but forgot the bit about reducing the exposure three stops after metering on the highlight" Denney

[tzalman]

Rick "Yogestee knows but forgot the bit about reducing the exposure three stops after metering on the highlight" Denney
Typo :p

[PhotosGuy]

The histogram is the best way to do it. Make an exposure, and check the histogram. If there is space between the histogram and the right end of the scale, dial in a little positive compensation. If you see the highlight warning on the display (you do have the highlight warning feature turned on, don't you?) and the histogram is jammed up against the right end of the scale, then dial in a little negative compensation. Generally, that's all true & one way to work. Sooner or later though, you'll run up against something that has highlights that you don't care about & shadows that you do care about, like cars with chrome & black tires in the shadows of wheel wells, so Rick & I assume that a person will use their brain & decide what is important in the shot & what isn't? (Which is why there are so many different & creative ways of metering to get the same end result.)

So again, this is my "rule of hand" which works in most situations for me. I make exceptions as I think they're needed, & shooting RAW takes up the slack when I'm wrong.
Need an exposure crutch? (http://photography-on-the.net/forum/showthread.php?t=89123)

[rdenney]

Typo :p

Oops. Yes. That should be increasing the exposure. If you meter on the sky, it will try to turn that light sky into middle gray, which would be under-exposed. So, you increase the exposure by three stops.

Rick ":oops:" Denney

[rdenney]

Generally, that's all true & one way to work. Sooner or later though, you'll run up against something that has highlights that you don't care about & shadows that you do care about, like cars with chrome & black tires in the shadows of wheel wells, so Rick & I assume that a person will use their brain & decide what is important in the shot & what isn't? (Which is why there are so many different & creative ways of metering to get the same end result.)

So again, this is my "rule of hand" which works in most situations for me. I make exceptions as I think they're needed, & shooting RAW takes up the slack when I'm wrong.
Need an exposure crutch? (http://photography-on-the.net/forum/showthread.php?t=89123)

Yes; I said that choosing which highlights to protect (i.e., place on Zone VIII) and which highlights to let blow out is an aesthetic decision.

The range of light in the scene is a variable. We can meter on the average of it, hoping that the extremes stay within the boundaries of the sensor. We can meter the shadows and hope the highlights don't blow out. We can meter the highlights and hope we having something left of the shadows. We decide which of the highlights can blow out, and which of the shadows don't need texture. That decision is one of the most important decisions we make as a photographer (still well behind where we point the camera and when we push the button, but important), and most of us leave it to the camera.

Metering of the hand is the same as metering on the average, except that we are providing a calibrated average instead of just hoping the scene adds up to a useful middle value.

The more one thinks about it, the more one understands why there was (and still is) a market for 1-degree spot meters. But the ability to study the histogram overcomes that need.

For shots of dynamic subjects, though, we want to get it right every time. So we instinctively learn where to point the camera when we make an exposure reading, and we learn where that AE Lock button is so that we don't lose that reading when we move the camera to capture the best composition. With averaging or center-weighted meters of old, I learned just how much sky I needed in the frame when taking a meter reading.

Rick "thinking we are scaring those newbies, heh, heh" Denney

[BluewookieJim]

That information about the "zoning" and adding 3 stops to the metering of highlights you are interested in keeping is very interesting.

I must admit that I don't really know anything about this "zoning" system, but in concept it sounds very useful.

In my personal experience, I tend to expose for the highlights via chimping. I almost exclusively use the partial metering mode on my 20D. I don't know how accurate that really is, but I'd certainly like to try identifying a target highlight area, and adding 3 stops. I wonder how that would compare to chimping.

[rdenney]

That information about the "zoning" and adding 3 stops to the metering of highlights you are interested in keeping is very interesting.

Ansel Adams had taught himself to evaluate exposure on the basis of reading portions of the scene lighted in different ways, and making sure those portions would be accommodated properly by the characteristic response curve of the film. Then, he was asked to teach photography, and he realized that he had no way to present this information to his students. He also felt that this violated the principles of disciplined learning he had (finally) achieved in his very extensive musical training. So, he and another photographer developed the Zone System to provide a means of codifying his intuitive methods.

It has a reputation for being complicated and esoteric, but I believe that's mostly because of how it is presented. Here's a synopsis:

First we divide the scene into ten zones, numbered from I to X, which range from solid black to print-paper white. Zones II and IX have slight texture, and Zones III and VIII are what I would call textured shadow areas and texture highlights. Zone V is middle gray.

We correlate these scenery values to the sensor by making the (only slightly wrong) assumption that each zone is one stop more or less bright than the zone next to it. Thus, if we photographed a gray card, the exposure from the camera would be zone V. One stop less exposure would render the gray card at Zone IV, and one stop more exposure would render it at Zone IV. In proper application, we should calibrate our whole work flow so that the zones really are a stop apart all the way to solid black and paper white in our final print, but in my view we can usually achieve that in post-processing as we go.

Then, when we look at a scene, we decide which portions of the scene should be where on that scale. If we want the sky to be on Zone VIII, then we measure the sky. The camera's meter wants what it measures to come out on Zone V, so to get Zone VIII, we need to add three stops of exposure to brighten it up. That gives us an exposure that will place the sky on Zone VIII. Now, we check to see where other image elements fall. If we measure the shadow, and find that it measures two stops darker than the exposure that places the sky on Zone VIII, then we know the shadow will fall on Zone III--two stops less than Zone V where the camera meters it.

If the shadow falls on Zone I or less, then we know that to get texture in the sky, those shadows are going to have to go black. We can then employ techniques to bring those shadows up, such as fill flash, filtration that darkens the sky (such as a polarizer, to some extent), and so on. But with some scenes, we may just have to live with it, or recompose the scene so that it doesn't matter. Or maybe we give up on maintaining texture in the sky and let it blow out so that we have the tonal texture we want in the shadow.

So, the Zone System is a way for us to measure and evaluate the brightness of different elements of the scene, find an exposure that places key values where we want them, and predict where other values will fall. As such, it is really a visualization tool to connect the scene to what we visualize as a final image or print.

If you want to go further than that, there is no better source than the original--Ansel Adams's book The Negative. The S-shaped characteristic curve and the testing and calibration stuff are probably more specific to film than digital, but the principles of visualization are just as important in digital as in film photography.

Chimping is not an alternative to this process, but rather a different tool for evaluating where those highlights and shadows actually fall in the range of the sensor. Instead of using a meter to find an exposure that places the sky on Zone VIII, I can use the histogram to check and see if the sky falls on Zone VIII and adjust as necessary. That means I need some experience to know where Zone VIII is on the histogram, but again I find that I can usually make those adjustments as easily using the tone curve in post-processing.

Rick "who can no longer think about exposure outside the Zone System context" Denney

[Wilt]

Rick,
Great explanation for the Zone System, to newbies and particularly to anyone who thought that the Zone System applies only to black-and-white film!

While the Zone System is truly the full combination of the Zone System of exposure determination + the Zone System of exposing + the Zone System of adjusted film processing , one can indeed apply the Zone System of exposure determination to color film and to digital. It (Zone System exposure determination) is, in fact, an essential element of understanding which provides insight about why reflected light meters are typically 'fooled' in many situations...It is actually the photographer not understanding what a meter is doing when it reads a surface!

[PhotosGuy]

Rick "thinking we are scaring those newbies, heh, heh" Denney There's a lot to be said for considering the audience & embracing the KISS principle. ;)

[samueli]

Hehe... Thanks for all the tips and explanations! For a minute there I thought I might have already been a good photographer!

Most of my photography up until this point has been concerned with the sharpest possible focus. Exposure would follow and mistakes would be obvious since most of my subjects weren't that dynamic. And if they where, I was shooting under very controlled conditions with a computer and large monitor available to check every shot as I took them.

Now that I'm trying to capture different subjects, it seems I've ignored a good portion of photography. One thing that really woke me up was a couple test shots I took of table and chair set in very mixed light. Whatever auto metering would give me a really flat image with no range. Next I tried center weighted on a dark area, lock the exposure and capture a way too bright image. Then I metered on the white fabric of the chairs, locked it in, and all of a sudden I had the sweetest image of a stupid dining set I've ever seen - with depth and separation. So now I want all my images to come out that way!

So let me recap my understanding:

I'll use the wedding dress example. White wedding dress, green lawn, large looming tree in the composition. Meter off the lightest spot with visible texture on the dress. Call that VIII. ISO 100, f8, shutter speed 1/100 puts exposure meter right on 0. As a starting point, I could compensate -3 stops by increasing shutter speed by 3 stops (1/300?), or reducing aperture size by 3 stops (what would 3 stops smaller be from f8?).

Or, I could meter off the dark shadow that the looming tree creates behind the wedding dress but makes the grass there look really cool. Call that III. ISO 100, f8, shutter speed 1/250 puts exposure meter right on 0. Although I could reduce the aperture size by 3 stops, I'd most likely decrease the shutter speed to gain +3 stops, which puts me around 1/30?.

I understand that placing something in III or VIII is subjective with no hard rules.

Now to get some practice and read more about the zone method! This is a very good start for me to get to the next level. I anticipate some well exposed outdoor shots this year.

Thanks again!

[tonylong]

Hang on, back up, and go back over things, Samuel!

Rather than think numbers, think "lighter" and "darker" for a moment, and compare it to "neutral". The bride's dress is "lighter", so tell the camera it is "lighter", that is, set the meter to "+": bright white, you set the meter reading to +2/+3, depending on the light.

In other words, you're getting the "exposure compensation" setting backwards. When the camera meters a white object, it interprets that as gray, and sets the meter in the middle, which would give an underexposure, so you dial in an exposure which would put the meter reading at a "bright" reading.

[tzalman]

So let me recap my understanding:

I'll use the wedding dress example. White wedding dress, green lawn, large looming tree in the composition. Meter off the lightest spot with visible texture on the dress. Call that VIII. ISO 100, f8, shutter speed 1/100 puts exposure meter right on 0. As a starting point, I could compensate -3 stops by increasing shutter speed by 3 stops (1/300?), or reducing aperture size by 3 stops (what would 3 stops smaller be from f8?).

Or, I could meter off the dark shadow that the looming tree creates behind the wedding dress but makes the grass there look really cool. Call that III. ISO 100, f8, shutter speed 1/250 puts exposure meter right on 0. Although I could reduce the aperture size by 3 stops, I'd most likely decrease the shutter speed to gain +3 stops, which puts me around 1/30?.


I think you need to start with a better understanding of your camera's controls. This would be a good starter:
http://photography-on-the.net/forum/showthread.php?t=414088
As an example let's use the one you gave above.
Meter off the lightest spot with visible texture on the dress. Call that VIII. ISO 100, f8, shutter speed 1/100 puts exposure meter right on 0
As Tony said, you want to INCREASE exposure in order to get that dress back to white from the grey the meter wants. You can do that by slowing the shutter speed because that will mean that the shutter stays open longer, thus letting in more light, or by opening the aperture, making it bigger and letting in more light. You want to add 3 stops of exposure, so what is a stop? A single stop is a doubling or halving of the amount of light getting to the sensor. So starting from a shutter speed of 1/100, one stop more light is 1/50, two stops = 1/25 and three stops = 1/12. Or you can open the aperture. Wider apertures are represented by lower f-numbers and the relation between them is x1.4 (which is the square root of 2 because we are dealing with doubling or halving the area of the opening.) So starting from f8, one stop opened more would be f5.6, two stops f4 and three stops gives you f2.8. Of course you can also increase exposure by using a combination of speed and aperture changes.

If you use the 'expose for highlights' method, that's it. Forget about Zone III shadows - Rick mentioned them just to give you a fuller theoretical understanding. Later in post-processing you can adjust the shadows to be Zone III (it will be easier if you do it during the RAW conversion).

[ralff]

Ansel Adams could see a histogram in his mind, histograms are one of the most valuable tools that came with digital cameras, learn to use it. Doesn't matter if you want to expose for shadows OR hightlights, the Histogram will tell you if you got it, if you learn how to read and use them.

[BluewookieJim]

...
Rick "who can no longer think about exposure outside the Zone System context" Denney

I've got to ask this, is it really reasonable to try to equate 1 zone to 1 f/stop? Our cameras just don't have a 10 stop range, it's more likely between 5 and 6 stops. For example, going back to the white dress example. I find that adding 1 1/3 to 2 stops brings me to that point where I'm right about to start seeing highlight clipping (partial metering mode on my 20D).

[Wilt]

One Zone to the next is one EV (one f/stop). The Zone system, based upon B&W photography, assumed wider latitude than our digital sensors or color slide film.

[yogestee]

Ansel Adams could see a histogram in his mind, histograms are one of the most valuable tools that came with digital cameras, learn to use it. Doesn't matter if you want to expose for shadows OR hightlights, the Histogram will tell you if you got it, if you learn how to read and use them.

Doubt it.. More like a Characteristic Curve..

The very term sends shivers down my spine.. When I was studying photography we had to measure and draw up Charactistic Curves of our 4x5 negs submitted for practical exams..

[samueli]

That's why I kinda reiterated. To make sure I was going the right way. If it wasn't practically the middle of the night when I read all this, I would have just tried it...

I do need to put this in practice though. I'm certainly not new to photography, but trying to accomplish what exposure lock does using manual mode is confusing me with out seeing it visually.

Like my table set example. I meter on the brightest area, which would be the chair pads. Lock that meter value and take the picture. Picture turns out with the white chair pads very bright and the rest of the scene is pleasantly darker. So in this case, where did it put that bright area? I'd call it at least VIII if not VIIII.

If I meter up on the bright chair pads or the white dress, I have to apply positive compensation. Metering off a dark area or shadow I apply negative compensation.

I think the most confusing part is that in manual mode, there is no starting point. As suggested you could choose your ISO, then your aperture based on effect. But then what? Do you meter on the white dress and dial in to zero on your exposure meter, or just give it 3 positive stops. What if you meter on the white dress and your already at +3 stops (or however many stops will be required for a good exposure). That floating variable confuses me. There is no solid starting point to compensate from.

I can capture a good picture without all of this confusion, but the reason I upgraded my camera was for center weighted metering. But it doesn't help to have something I'm not using right.

I've probably asked too many questions, which is making me look like a total newb. I'm just trying to constantly, manually reproduce what aperture lock does. And hopefully to the point that I don't have to think about it, but becomes more of a reaction. There are some things I can never remember. Like in photoshop, flip vertical or horizontal. I get it wrong every time and have to undo and then choose the right one. I've been using photoshop for 10 years at least, but I still get that basic thing wrong everytime.

[tzalman]

I've got to ask this, is it really reasonable to try to equate 1 zone to 1 f/stop? Our cameras just don't have a 10 stop range, it's more likely between 5 and 6 stops. For example, going back to the white dress example. I find that adding 1 1/3 to 2 stops brings me to that point where I'm right about to start seeing highlight clipping (partial metering mode on my 20D).
Most current DSLRs will get about 2.5 stops above medium grey in jpg and 3.5 in RAW. I think it was said in earlier posts that the advice to add 3 stops to a metered highlight was on condition of shooting RAW. Depending on the noise level of the RAW image, which determines the lower end of the dynamic range, total DR can range from 8 to 11 stops.

[BluewookieJim]

I do shoot RAW exclusively, but I don't feel like there are 2.5 to 3.5 stops above medium grey there. I personally find that anything over 2 stops and I get into the realm of unrecoverable highlights.

[tdodd]

My simple test to find out where your own camera's clipping point is to point the camera at a sheet of plain, white, A4 paper and fill the frame completely with the paper. The lighting source doesn't matter so long as it is even and constant throughout the test. Bright sunshine and clear skies will be fine, as will constant tungsten lighting. Then meter off the paper using spot or partial metering and set a manual exposure to give you a zeroed meter needle. Take a shot in raw+JPEG, just so you can establish all the "facts" for your camera. Now increment the exposure by 1/3 stop by reducing your shutter speed and repeat. Carry on incrementing and shooting until you are at +4 stops in total.

Now load those images into your favoured image processor. For me it's Lightroom (or DPP) but anything that will give you an accurate histogram. If it has highlight clipping indicators then so much the better. Now you can simply review the images and see where clipping started with your JPEG and where it started with your raw file. With the raw file you might see how realistic/easy it is to recover highlights as well.

For a better subject, with a bit of detail so you can really see when things are ruined or not, try using a white towel instead of A4 paper so that you can observe more easily when the detail of the texture vanishes completely and how readily it can be recovered with highlight recovery.

Note that when you shoot to JPEG (or raw + JPEG) it is essential that you set your WB correctly beforehand and also your picture style and any other parameters that will affect the levels of each colour channel. If your camera has HTP and/or ALO and LPIC I suggest you disable these features for testing purposes. Your parameters for sharpening, contrast, saturation etc. should all be correct as you would normally shoot.

I have performed these tests with my 30D and 40D so I know how those cameras fare. The 40D has a bit more headroom because, as reported by DPReview and others, the ISO settings on the 30D are a little bit off, with the indicated sensitivity of 100 ISO actually being more like 1250 ISO.

If we turn our attention to grey cards (are they 18% reflective or 12% or 13% reflective - who knows?) then consider this....

100% saturation of the sensor/pixels is the clipping point. That is the brightest light the camera can record for any given exposure setting. 1 stop below the clipping point is 50% less light. 1 more stop below that is half as much light - just 25% of the original brightness. 1 more stop below that is 12.5% of the original light. So, if you aim your camera at a 12.5% grey card and zero the meter against that you should have exactly 3 stops of headroom before clipping. If you use an 18% grey card, that reflects more light towards the camera and will thus appear brighter to the meter. The camera will therefore lower the exposure to compensate and you will end up with more than 3 stops of headroom compared to a 12.5% grey card.

So, forget grey cards for a minute. Do the test I suggested earlier and you will find out how much headroom you have with your camera above the metered exposure with a centred needle.

I would be very surprised and disappointed if a camera shooting to JPEG could not hold highlight detail at +3 stops above a centred meter needle. As I only shoot raw I guess I'll never know the truth about JPEGs.

As for the total dynamic range of modern sensors and cameras, take a look at DPReview's data in their latest camera reviews. Selecting the 50D as a sensible mid-level example, according to DPReview the DR from this camera can be as high as 8.5 stops. Here are the figures....

http://www.dpreview.com/reviews/canoneos50d/page19.asp

The DR is from about -5 stops to +3.5 stops. This old figure of 5-6 stops for digital is a myth.

[rdenney]

Doubt it.. More like a Characteristic Curve..

The very term sends shivers down my spine.. When I was studying photography we had to measure and draw up Charactistic Curves of our 4x5 negs submitted for practical exams..

Heh, heh.

Rick "shivering" Denney

[rdenney]

I've got to ask this, is it really reasonable to try to equate 1 zone to 1 f/stop? Our cameras just don't have a 10 stop range, it's more likely between 5 and 6 stops. For example, going back to the white dress example. I find that adding 1 1/3 to 2 stops brings me to that point where I'm right about to start seeing highlight clipping (partial metering mode on my 20D).

Equating zones to stops was part of the testing and calibration process. If I placed a scenery value on Zone III, and the highlights fell on Zone VII, but I really wanted them on Zone VIII, I would need to add development time to increase contrast. How much to add was the result of testing and calibration. But I would be doing that to change the actual performance of the film to match my visualization of the scene.

So, we started with the definition that a zone equaled a stop, and then learned how to adjust our process to stay within that definition. Film has an S-shaped characteristic curve, so the relationship between zones and stops changed depending on where you were in its response. Digital is more linear, and therefore easier to manipulate.

I'm sorry your camera only provides five or six stops of range. Mine seems to provide about eight or nine. Velvia is about six or seven, and we used the Zone System with it, too. Polaroid Type 55 P/N was about five stops, and we used the Zone System there as well. In fact, Adams wrote a book about using Polaroid film and how to deal with its narrower dynamic range.

But using a film with narrower dynamic range means that we have to make tighter decisions. If we know that our sensors blow out at Zone VIII instead of Zone X, then we have to place our textured highlights at Zone VII. Just as with Velvia, that may mean more shadows go black, or that we have to choose a different composition to hide the values that we can't record. Even with digital, we still need to understand our equipment and how it responds. But the histogram is a tremendous help here, and we can perform that testing in minutes instead of weeks as it used to take.

There is a weakness to the histogram, however. It shows the expanse of data that represent the image. Small highlights may not be visible on the histogram, but they form a vital part of the image. That's part of what you are seeing.

Another part is that you are thinking that Evaluative Metering really exposes to Zone V, which it does not do. It tries to fit the average light levels in the various metering zones into a range of values around Zone V. That's quite different, and for any given image, it might be stops above or below a Zone V reading on a middle gray portion of the scene. If you are going to use the Zone System, you'll end up using a spot meter, or otherwise being very aware of what the camera is metering when you take a reading.

Frankly, I don't usually go to that trouble. I use evaluative and then adjust based on viewing the histogram, and that's a faster process than taking multiple readings. But for tricky metering situations, and a white, textured bridal gown is the ultimate example, I still do it. I meter on the dress, and brighten up the image by two or three stops. Then, I know I won't blow it out. I know from experience that that my big reflector will bring up the black tux, but even if it doesn't, nobody will care as long as the bridal gown shows evidence of the money Daddy spent on it. Chimping a lot at a wedding gig probably won't put a good impression across on the client, heh, heh.

Rick "whose last wedding gig was shot on film" Denney

[yogestee]

Without trying to over simplify things..

-2 1/2 stops black ---------- 0 mid grey --------- +2 1/2 stops white

Naturally there are many variables like processing inaccuracies, monitor calibration and lens differences..