Hi. I've heard that the zone system is often combined with spot metering. When using center weighted metering, is the zone system also recommended? Or is it not necessary in this case?
Thanks.
Hi. I've heard that the zone system is often combined with spot metering. When using center weighted metering, is the zone system also recommended? Or is it not necessary in this case?
Thanks.
My understanding of the zone system is that a specific area of the scene is exposed such that it will fall in Zone V (typically; it could be another Zone). A spot meter will be able to meter this area whereas a center weighted meter will pick up a much larger area which could include areas of other tonal values. That would seem to preclude its use, at least in the way it was originally envisioned to be used.
For the point of discussion, I will be refering to image values within the 0 to 255 range. When working with an expanded bit depth, scale accordingly.
Actually, the premise of the Zone System is to select an exposure that preserve details within the important areas of a photograph while maximizing the available dynamic range. Regardless of shooting with color or monochrome, film or digital, or how you divide up the range from Black to White, you will find that the detail falls off at the extremes of the exposure range. Please note that I'm not referring to situations where the exposure "hits the wall" at either extreme, but situations where every pixel lies somewhere within the usable 0 to 255 range.
When metering your exposure, you need to understand that every reflected light meter, which includes all in camera meters, have a single goal, to select an exposure combination of ISO, Shutter, and Aperture that will achieve a midtone value of 128. It doesn't matter if you meter off an object that's Black, White, or 18% Gray, the recorded image value, when using the meter chosen settings, will be 128. I put this fact to use in how I apply the Zone System.
For my own use, I have adapted Ansel Adams' Zone System from the original 11 zones to contain 15 zones. Adams's Zone V, midtone, is what I reference as the "as metered" 0 EV with a image value of 128. Each zone on my scale is then plus or minus 1 EV from it's neighbors. Unlike Adams' Roman numeral Zone labels, I find a scale centered on 0 EV with 1 EV increments easier to apply. Since the intent of the Zone System is detail preservation, I have determined that details are lost if an area is at or beyond 4 EV from my "as metered" 0 EV. This does not imply that I work with a low contrast narrow exposure range, But when a subject or feature contained within an image falls beyond 4 EV from the "as metered" 0 EV, I know that the level of detail is below what I feel is acceptable.
The concept behind a spot meter is to individually measure each object in a scene and determine how it would be rendered by the over all exposure, or to set the photograph exposure to place each object at a particular zone. Either way, it still amounts to a lot of work. I feel a spot meter is not as important as ensuring that all features of a photograph preserve the appropriate level of detail.
Regardless of your camera's metering mode, just remember that, for the area it measures, it will select an exposure matching my 0 EV Zone. Now, this is where the histogram becomes important. I have my software set to warn me when pixel values fall out side the -4 EV to +4 EV range, which corresponds to 8 and 247. This does not mean areas containing pixel values below 8 or above 247 are unacceptable, but understood as lacking detail. When I'm post processing, I always maximize the histogram for the main subject to span the 8 through to 247 range yielding the best possible contrast level.
This link will take you to the Wikipedia page: Zone System
Or if you read Michael Freeman, "The fundamental question for using it (the Zone System) in digital photography is not how but why. ... frankly it has little value in its original intended way", p166 (The Complete Guide to Black & White Digital Photography, ILEX, 2009)
Spot / partial / centre-weighted / Evaluative are just ways you can tell the camera what portion of the scene to consider in it's exposure calculations; in theory you can use the zone system using any metering mode, but in practice it's a bit like trying to do a paint-by-numbers paining with a 3mm / 10mm / 30mm / 50mm brush in that the "bigger the brush the harder it is to isolate just the area you want".
The zone system has it's limitations when used with a digital sensor though, because the camera will typically capture up to around 12 stops of info, and we can only print around 4 and display around 6 ... but we can raise the value of the lower levels to reveal more of what's contained in the shadow / dark areas if we wish, but we can't do anywhere near as much with blown highlight detail - so with a digital capture, zone 5 isn't necessarily going to maximise the captured dynamic range (I say "not necessarily" because it's going to depend on the dynamic range of the scene).
So with a normal reflective scene you can use the zone system (best results with spot-metering) - or just let the camera figure it out for itself (normally with pretty good results). If you have a scene that's not normal from the point of being biased one way or the other (white bear in a snowstorm or black cat on a black rug) then you can use the zone system ... if you can find a zone 5 reference, or you can just use exposure conpensation with other modes for an OK result (albeit one that might require a tweak in post-production) (which is almost always needed anyway). For high dynamic range scenes though (eg scenes with backlighting) it's more important to avoid blown highlights in critical areas of the scene than it is to set zone 5 as the mid-point.
Probably too much information
Actually, lots of good information from all of you.
It seems that the zone system has the most relevance when spot metering. It seems that center weighted probably meters from too wide an angle for the zone system to be highly relevant, although the exposure may or may not need adjustment.
Possibly a better way of thinking about it might be along the lines of "the bigger the metering area, the bigger the zone 5 area is going to need to be". If you have a bit of food stuck between your teeth you could use a crowbar, but a tooth pick will be easier
To be honest, I seldom use the zone system. When I'm shooting high dynamic range scenes (ie most of my landscape) it's easier and more accurate to spot meter the brightest portion of the scene and then just up-shift the exposure around 2 stops so that the metered "midtone" gets shifted to the highlight that it is, and everything else just falls in behind.
Colin, that sounds different from my "rule of thumb." (Then again, I have funny-looking thumbs. )
I have considered the spot metering setting to require the Zone System. I have considered the center weighted metering to be doing the "zoning" for me. I have considered the matrix (or wide area) metering to be intended for those shots with fairly subdued contrast, lighting or such. Mostly indoors without much, if any, back lighting. It has worked fairly well for me, but I'm going to try your method, as I am having trouble with pictures having extreme contrast, as in my sunsets.
Pops
Hi Pops,
I'm not sure what you mean by "I have considered the spot metering setting to require the Zone System.", but I think we're still in general agreement. Basically with spot metering of a normal contrast scene you HAVE to spot meter a medium gray to get a correct exposure if you don't want to apply any exposure compensation. If what you spot meter isn't a Zone 5 / Medium then that's fine also - but - you then have to add or subtract the correct amount of EC. You don't have to be in spot metering to set exposure this way - but - if you're in a metering mode that looks at a bigger area then the zone 5 will need to be correspondingly bigger (in theory, but in practice I son't know whay anyone would want to do it this way).
With higher dynamic range scenes though the number 1 rule changes from "highlights showing as highlights and shadows showing as shadows) (ie a traditional normal exposure) to "don't blow the highlights". So when you meter the brightest portion of the scene - upshift it a couple of stops (give or take depending on the scene) you've then pushed the metered area up into the highlights. By moving 2 stops you usually have about 1 stop of safety margin (a little more with HTP on) - but you can loose highlight detail if you push it too far.
It's all "solid science" with a big dollop of "magic and luck" thrown in because the result depends on a number of things, with the main one being the DR of the scene -- often one has to use GND filters to compress the DR range -- and at other times one just has to let a portion of the scene blow because there's just no way to capture everything (eg naked sun in the shot); at least when one understands the techniques then one can use the tools to make the best decisions at the time.
Having said all of that, the simple rule is "if you have the blinkies then they'd better only be in areas that you don't care about, or you need to reduce the exposure" and if you're shooting a high DR scene then if you DON'T have a blinky or two then you're probably going to get shadow noise when you raise the shadows to reveal shadow detail.
Hope this helps
You and I are on the same page. It matters not that one of us is reading upside down, so long as the message is understood.
We are doing the same thing for the same reason, except that you meter the blinkies and adjust. I was evaluating the scene from metering on the mid tones and adjusting. I think your way might give more consistent results. I should be able to shoot some this weekend, if I'm able to move around by then (Tree trimming scheduled for tomorrow, from the bucket of a backhoe. )
Pops
Pops - you're too hard on yourself - I've never thought of you as being upside down!
The "problem" with the zone system is that it assumes your dealing with only reflected light - and in that situation it works just fine ...
... but of course the problem is that with backlighting you're then dealing with direct light sources AND reflected sources and that's a major game-changer; so the challenge becomes how to shift stuff around to get what you want.
Personally I just use a lightmeter (because I can meter multi-minute scenes and apply other compensations for filters), but even the built in meter can get one a long way. The other thing to remember though is that if you're shooting sunrises or sunsets then the light intensity is going to change considerably over a several minute exposure - so that also needs to be taken into account when upshifting from a spot-metered highlight (eg if it's a sunset you can increase the exposure by more than 2 stops, but if it's a sunrise then you need to be more conservative).
One more question about the zone system and metering. Yesterday I saw a heavy cluster of trees. Some part had direct sunlight and were quite bright. The other part was not in direct sunlight.
I guess it would all be in zone 5 and no compensating exposure would be needed (as would be the case if it was snow), regardless of the differences in brightness, since it was all green. Is that correct? Also, would you use center weighted metering in such a case focusing on the bright area? Or, would you use matrix metering?
Thanks
If some are in sunlight and others are not they are likely in different zones. Whichever portion you spot-meter, that portion will be placed on Zone V. But with the ability to see the photo on the LCD after you have taken it you can make adjustments to the exposure until you get a result you like.
Center-weighted metering uses the central portion, 8mm or so, of the view for most of the information to determine the exposure. The remainder of the view only contributes a little to determining the exposure. Matrix metering uses the entire view and takes into account several factors to compute the exposure. Try both to see the difference, if any.
Hmmm, I guess I was wrong about the zone assumption. So, if you have say green leaves that are in direct sunlight and are bright as a result (but still very green), and if you spot meter on that patch, would you compensate with Exposure Compensation and overexpose by a stop?
Thanks.
I'll explain the process so you can better understand how and when to apply this.
With your camera set to matrix or area metering, look through the view finder and find a balanced area of the scene containing both the brightest and darkest areas within the meters field of view. As long as the dark and bright balance each other, it will choose the midpoint exposure for you. Now set your camera to this exposure and frame the scene how you want it.
Now, when you are dealing with greatly off balanced scenes, a dark room with small windows to daylight, a dark forest with some sun getting through, or a scene with snow, matrix or area metering will probably not be the way to go since the predominant area will dominate the exposure and skew everything else in the scene off in the opposite direction. Even your histogram will fail you in such situations. Another problem is if the darks and lights are not close enough for them to all be within the metering area, often a bit smaller than the frame that the camera images.
Here is an alternate, have your camera set to center weighted or spot metering, look through the view finder and target the darkest area where you would want to preserve detail, noting the exposure and Exposure Value.
If your meter doesn't provide a measure of Exposure Value, you'll need to do a bit of math and ensure your calculator can perform a natural or base 10 logarithm. Either one will work since I'll provide a formula that will give the right value regardless of which LOG function you choose to use.
Exposure Value = LOG( f/ number * f/ number / Shutter Speed ) / LOG( 2 )
Now, aim your camera to the brightest area where you want detail and again note the exposure and EV.
The actual exposure that you need to set the camera to will be between these two values. As long as the dynamic range is not so extreme as to exceed your camera's capabilities, nothing will be blown out and all your desired details will be captured.
This is why I adopted a zone system based on Zone V being the "as metered" 0 EV and each Zone being 1 EV from the next.
Lets say the dark was 1/15th at f/2.8,
Calculate the Exposure Value by
Exposure Value = LOG( 2.8 * 2.8 / 0.06667 ) / LOG( 2 )
Exposure Value = 7
and the bright was 1/1000th at f/4 yielding a
Exposure Value = LOG( 4 * 4 / 0.001 ) / LOG( 2 )
Exposure Value = 14
( It's Ok to round these off )
So far so good, the total span is just 7 EV, well within the 8 EV that I use. Now, just pick an exposure that corresponds to the mid point:
Midpoint Exposure Value = (( Bright EV - Dark EV ) / 2 ) + Dark EV
Midpoint Exposure Value = (( 14 - 7 ) / 2 ) + 7
Midpoint Exposure Value = 10.5
Now lets find the exposure, shutter and aperture combination, to match. If you start with a known shutter speed, you can get the f/ number by:
f/ number = Square Root( Anti-LOG( LOG( 2 ) * Exposure Value ) * Shutter Speed )
Lets pick 1/60th:
f/ number = Square Root( Anti-LOG( LOG( 2 ) * 10.5 ) * 1/60 )
f/ number = Square Root( 1448 * 0.016667 )
f/ number = Square Root( 24.135 )
f/ number = 5 (Rounded off)
With the camera set to 1/60th at f/5 when the shutter is pressed, you'll know right off how everything will turn out. The darks will be 3.5 EV / Zones below 0 EV and the brights 3.5 EV / Zones above. I've seen large format photographers take the time to meter each element of a scene and try to juggle the exposure of each, which can end up becoming quite involved. In my experience, you just need to be concerned about the end points, the darkest and lightest areas where detail is desired, the rest will be somewhere within this range which is where you want it anyway.
A point of note, I simplified these calculations for ISO 100 and you can find look up tables and the ISO Scalling formulas here:
Exposure Value on Wikipedia
Here is an article from View Camera magazine where Steve Simmons explains the process of spot metering a scene:
Metering the Large Format Photograph
Obviously, he likes making work for himself.
UPDATE: If your camera or hand held meter provides Exposure Values, then most of the math has been done for you and this becomes a very simple process.
Last edited by Steaphany; 18th June 2010 at 04:43 PM. Reason: Simplified formula & typo
Nice explanation
easiest might be to prepare a table (with exposure vs. aperture, giving the EVs) that you can carry in your camera bag.
Then first part is: look up the EV values from the select F/exposure pairs, second is: find the wanted EV value and pick one of the corresponding pairs.
The nice thing is that you don't have to worry about the iso values you use, as long as all measurements are done at the same iso setting (the EV values will all be off by a fixed amount, so you still end up with a proper combination of time and aperture).
Remco
Steaphany - Thank you for the explanation and all the work that went into that! It is quite involved and I've been reading/re-reading it, but it does seem like a lot of work to get an exposure correct. I am hoping to boil it down to something a bit more rule-of-thumb that one could do in one's head.