PDA

View Full Version : Film Latitude, 5 over and 2 under.


05-07-2011, 08:16
Hi Roger,

I read an old 1970s article recently that some b&w films have a latitude between 7 stops (Plus-x) and 9 stops (Tr-x).

The article mentions that the film latitude tends to be bias towards overexposure and less towards underexposure. The number suggested for Plus-x was +5 stops for overexposure and -2 stops for underexposure.

I have 2 questions which I hope you can point me to the direction of answers:

a. is this still applicable today ?

b. does this mean that one possible method of calculating exposure is to measure the shadows which I want details and then stop down 1 or maximum 2 stops to ensure that it is within the film latitude for underexposure ?

thanks !

raytoei

Ronald M
05-07-2011, 08:53
Only with short tone scale subjects that are less than the tonal range of the film, you can over or under expose. Even then there are sacrifices to grain pattern etc.

Measuring a dark area you want rendered as a shadow with detail and then placing it 2 stops less than middle grey is considered a valid way to get exposure correct. You are not really under exposing it, but making a proper exposure. If you were to determine as above and reduce two further stops, that would be two stops under and there would be no shadow detail.

When you overexpose, all the bright tones get jammed at the bright end and they do not separate in a print, even if you print them down darker.

Dark tones will still separate around 3 or 4 stops under middle grey and bright tones 4 or 5 above. Meters tend to want to render all subjects middle grey, whether dark or light, it does not know. Therefore it wants to make a pile of coal grey instead of black and a mountain covered with snow was grey also. In the case of coal, the meter thinks it is a grey object with very little light on it thus recommending too much exposure. With snow, the meter thinks it is a grey object with lots of light on it thus recommending too short an exposure. It is up to the user to compensate.

The cure is an incident meter which reads the light falling on the subject and not the light reflected from it. Movies are done with them because they can not bracket and can not always accurately predict the true reflectance of the subjects to compensate for a reflected meter shortcomings.

Film can not be truly over or under exposed from optimal exposure to any degree and still make a quality print. Sure you can get a print, not an optimal one. The more you deviate from correct tone placement, the more heroic the print efforts become.
But i must repeat, placing darker tones at less than meter reading and brighter tones over is not considered over or under exposing, but what you are supposed to do. It is when you push them further from the meter reading than they are supposed to go, it is considered over or under exposure.

You will be well served if you do not count on film having any tolerance at all.

Read any book on Zone System. It will read like I said. It will further go into developing film longer or shorter times to compensate for short or long scale subjects.

Roger Hicks
05-07-2011, 10:34
Hi Roger,

I read an old 1970s article recently that some b&w films have a latitude between 7 stops (Plus-x) and 9 stops (Tr-x).

The article mentions that the film latitude tends to be bias towards overexposure and less towards underexposure. The number suggested for Plus-x was +5 stops for overexposure and -2 stops for underexposure.

I have 2 questions which I hope you can point me to the direction of answers:

a. is this still applicable today ?

b. does this mean that one possible method of calculating exposure is to measure the shadows which I want details and then stop down 1 or maximum 2 stops to ensure that it is within the film latitude for underexposure ?

thanks !

raytoei

Dear Ray,

Well, a lot depends on the film and on how you develop, but even more depends on the compromises you want to make. Basically, your summary is pretty much true, and basically, yes, if you take a shadow reading of the area in which you still want some density on the film, then stop down 2 stops from the mid-tone meter index, you'll have plenty of shadow detail unless your equipment (meter, lens, shutter...) is quite seriously off.

There is no such thing as a 'correct' exposure, but there is such a thing as a perfect exposure, defined as the exposure that gives you the effect you want. This can involve anything from very generous exposure and underdevelopment to significantly curtailed exposure and overdevelopment.

As you almost certainly know, ASA speeds originally incorporated a safety factor of about one stop, which was dropped in 1960. As you probably know, film speeds were based on psychophysical testing in about 1940, when people were shown a series of prints, from underexposed to overexposed, and asked to choose the 'first excellent print', i.e. the one that was not significantly worse than the next print with more exposure. As you may or may not know, these were 5x7 inch contact prints made from negatives exposed through uncoated lenses. And ISO standards for film contrast and exposure meters are based on 'average' subjects around Rochester, New York, where the original research was done. In the 1990s the Japanese members of the ISO Standards Committee lobbied unsuccessfully for a lower standard ISO contrast for development because in Japan, the light is (on average) contrastier.

In other words, anyone who claims to have a handle on objectively correct exposure, good for all conditions at all times, is spouting drivel. The original researchers, and everyone else before or since who has any idea of what they are talking about, knows that EIs, metering, etc., are based on rules of thumb that work pretty well most of the time. The most important thing you can say in favour of ISO standards is that they're better than leaving film speeds to the marketing department. And, today, a LOT better than leaving them to self-appointed internet experts.

Incident light metering works for reversal materials, because the exposure is 'keyed' to the highlights, i.e. you give the MAXIMUM exposure you dare without blowing the highlights to a featureless white. It is worthless for negative exposure unless the tonal range is sufficiently short that the shadows fall within the latitude of the film. They quite often do, but when they don't, an 'uninterpreted' (= non-fudged) incident reading will result in underexposure of the shadows.

This is why direct metering of the shadows is, as you suggest, the ONLY way to GUARANTEE full shadow exposure, by giving the MINIMUM exposure necessary to get detail in the shadows. Once you have hoisted that on board, it's all comparatively easy.

Minimum exposure will place the shadow values on the toe of the D/log E curve, which will give different tonality from generous exposure, where you are on the straight line portion of the curve. Which tonality you prefer is personal, but a lot of people prefer to get above the very beginning of the toe, i.e. to expose generously.

Metering the way you suggest will give remarkably consistent and (to most people) pleasing tonality, but 2 stops down from the main index should be plenty: IRE 1, widely used as a shadow index, is 2-2/3 stops down from the IRE (Institute of Radio Engineers) mid-point index. You might even get enough shadow detail, with most modern B+W films, at box speed, with normal development, at 3 stops down.

The disadvantages of generous exposure are (1) reduced sharpness (2) bigger grain (except with chromogenic films such as Ilford XP2 Super where the grain is finer with overexposure) and of course (3) bigger apertures or longer shutter speeds or both. This is why gross overexposure is seldom a good idea, at least with 35mm where such things are much clearer, and why many who now what they are doing use different compromises for 35mm and larger formats.

A further disadvantage of generous exposures, if you are scanning your negatives, is that cheap scanners often have a limited ability to penetrate high negative densities, so that negatives which would print perfectly well with a decent enlarger suffer from 'blown' highlights in the print.

Hope this helps,

Cheers,

R.

bwcolor
05-07-2011, 10:44
Thanks Roger.... That was the best, short summary that I have read.

05-07-2011, 16:28
Thanks Roger and ModelM,

Thanks for both of your detailed explanations! I am bookmarking it to re-read it again.

This method raises all sorts of uncomfortable questions in my head:

Like does this method work best in an outdoor day time scenario and whether does it apply to night or indoor settings ?

say i want to measure the exposure of this 2 scenes:

scene a: http://4.bp.blogspot.com/_KjCW0sc3Sz0/TMeS9GsO0uI/AAAAAAAAG1w/RHoAj17aegM/s400/h5.jpg

scene b: http://3.bp.blogspot.com/_KjCW0sc3Sz0/TMeTXCtBkpI/AAAAAAAAG2Q/1uyvUdmUEZ8/s1600/h.jpg

in Scene a, my thought process is this, please guide me if i am wrong.
I will spot meter the tuxedo, assuming it is the darkest shadow that I want some details. And then stop down 2 stops (may be even 1.5 stops).
Assuming the scenery tonal range is within the film latitude range, won't this method result in an underexposure ?

Perhaps in Scenario A, it would be better to spot meter the model's face, and perhaps use it as a grey 18% or even slightly open up by half a stop ?

In Scenario B, because the object is the model. I would spot meter the face, and then open up a stop. I wonder would this be very different if I just Incident meter from the face of the model ?

thanks

Steve M.
05-07-2011, 16:42
Just decide what are the most important tones in your scene, meter for it, and you're good to go. Reflective metering works fine, as you can always look around for something that matches your intended tones and meter off that. A good spot meter works great too, but you have to weigh it's potential against carrying something like that around w/ you. If nothing around you matches the tones you want to capture, meter off the palm of your hand in the same light as your subject. That will give you a good middle value as a base. Of course, w/ B&W film there's no need to be that fussy. The exposure latitude will give you plenty of opportunity to put things to your liking later, either at the development stage or the printing.

I have never understood why someone would want an incident meter unless they were shooting in the studio. If my subject is across the street and in the shade, and I'm in the sun, good luck. Reflective meters are plenty good enough for any shooting situation I've ever encountered. If I shot portraits in a studio, then I might want an incident meter to read values in different areas of my sitter's head. But even then, w/ B&W film I don't see the need. For digital, probably so, as things can get blown out and you'll never see them again. Film will give you a lot more opportunities to get a REAL wiide exposure range. A lot of digital camera companies like to talk up their exposure range as equaling film, but it's just a numbers game. In real life, the high and low of that range isn't really usable.

Roger Hicks
05-08-2011, 01:26
Dear Ray,

Night photography, with extreme tonal ranges, is a bar steward. It's worse if (as in those pics) the model is wearing black. You'll often have to sacrifice something, and of course from the cheap screen on my internet computer, I can't tell what you've sacrificed already.They look pretty damn' good to me (nice pics, by the way) but I doubt that an extra stop would do any harm.

Paradoxically, in this case, I might just recommend an incident light meter, from the model position. An incident reading under (relatively) flat light (as seen at the model position only) will make sure the model is well exposed, and you can probably work with the rest in the darkroom.

@Steve M: Incident readings are ideal for slide and digi, for the reasons detailed in my previous post, but apart from that, as you say, they're really only any use for neg if you're shooting under controlled light or when you're concentrating on a subject that is reasonably consistently lit, and ignoring the background.

Cheers,

R.

Sparrow
05-08-2011, 03:04
.... just to be contrary, I often use an incident reading taken off the sky to get a datum for the general level of illumination ... then reason the actual exposure from there. That way I only need to do it the once, unless the light changes

Roger Hicks
05-08-2011, 04:10
.... just to be contrary, I often use an incident reading taken off the sky to get a datum for the general level of illumination ... then reason the actual exposure from there. That way I only need to do it the once, unless the light changes

Dear Stewart,

No contrarianism there. You're fudging, on the basis of intelligence and experience. From my post above:

Incident light metering works for reversal materials, because the exposure is 'keyed' to the highlights, i.e. you give the MAXIMUM exposure you dare without blowing the highlights to a featureless white. It is worthless for negative exposure unless the tonal range is sufficiently short that the shadows fall within the latitude of the film. They quite often do, but when they don't, an 'uninterpreted' (= non-fudged) incident reading will result in underexposure of the shadows.

After all, ANY metering system can be made to work, with a bit of intelligence and experience.

Cheers,

R.

charjohncarter
04-09-2013, 16:34
Dear Ray,

Well, a lot depends on the film and on how you develop, but even more depends on the compromises you want to make.
R.

Here is one I did with TriX (at 200) and developed in HC-110h. I think this has a long tonal range. If you read Roger's complete essay above you will get much knowledge on how to get long tonal range:

http://farm7.staticflickr.com/6150/5919238696_2ea95338fe.jpg

ChrisN
04-09-2013, 16:52
Wonderful example John!

Just to add a little to the discussion, Bruce Barnbaum published an article titled "Photographic Myths: There are ten zones in the Zone System" published in the Sept/Oct 2000 issue of PHOTO Techniques. In the article he explains how he regularly gets 13 stops of range recorded on film. Of course he was printing traditionally, and dodging and burning heavily to extract all the detail recorded on the negative. The article was reproduced in his publication "B&W Master Printing Class" in 2006.

A negative with that exposure range is beyond the dynamic range of most scanners, and will require several scans at different exposure levels, which can be blended in Photoshop (assuming the scanner can even see the detail in the most dense parts of the negative).

04-09-2013, 17:35
John, very nice and long tonal scale. can you care to share how did you meter this to capture details indoors as well as outdoors without the picture being too harsh ?

Rob-F
04-09-2013, 18:34
Dear Stewart,

No contrarianism there. You're fudging, on the basis of intelligence and experience. From my post above:

Incident light metering works for reversal materials, because the exposure is 'keyed' to the highlights, i.e. you give the MAXIMUM exposure you dare without blowing the highlights to a featureless white. It is worthless for negative exposure unless the tonal range is sufficiently short that the shadows fall within the latitude of the film. They quite often do, but when they don't, an 'uninterpreted' (= non-fudged) incident reading will result in underexposure of the shadows.

After all, ANY metering system can be made to work, with a bit of intelligence and experience.

Cheers,

R.


To think about and understand Roger's statement that incident metering is keyed to the highlights, I started with the idea of an 18% reflectance (what the incident meter is supposed to be calibrated to) as the "mid-point" exposure. Then I counted the stops from there to the shadows and highlights. It went like this:

18%: the midpoint
9%: 1 stop towards shadow
4.5%: 2 stops toward shadow
2.25%: 3 stops into shadow
1.125%: 4 stops, maybe the darkest shadow?

18%: the midpoint
36%: one stop above midpoint
72%: two stops above midpoint
100%: probably about 2.5 stops to the brightest highlight.

So it looks like the 18% midpoint isn't really the midpoint. It's less than 3 full stops from the highest highlight, but more like 4 stops (by my casual method) from the deepest shadow. So it appears this is why the incident reading is biased toward the highlights.

Does that make any sense?

Rob-F
04-09-2013, 18:40
Another point, about incident meters being best for reversal film. There is an exception. Hollywood Cinematographers use incident meters, but they are not shooting reversal film. It's a negative. So how do they get the exposure right? See Roger's "After all, ANY metering system can be made to work, with a bit of intelligence and experience," above!

Sarcophilus Harrisii
04-09-2013, 23:40
To think about and understand Roger's statement that incident metering is keyed to the highlights, I started with the idea of an 18% reflectance (what the incident meter is supposed to be calibrated to) as the "mid-point" exposure. Then I counted the stops from there to the shadows and highlights. It went like this:

18%: the midpoint
9%: 1 stop towards shadow
4.5%: 2 stops toward shadow
2.25%: 3 stops into shadow
1.125%: 4 stops, maybe the darkest shadow?

18%: the midpoint
36%: one stop above midpoint
72%: two stops above midpoint
100%: probably about 2.5 stops to the brightest highlight.

So it looks like the 18% midpoint isn't really the midpoint. It's less than 3 full stops from the highest highlight, but more like 4 stops (by my casual method) from the deepest shadow. So it appears this is why the incident reading is biased toward the highlights.

Does that make any sense?

I drafted the post below before I had to go out earlier because, like you, I found myself pondering Roger's comments concerning incident readings being more relevant for containing highlights than recording shadow detail. You've gone further with that than I did, and perhaps you've answered the question I pose, but I'll include the post I'd previously drafted below.
Cheers,
Brett

Dear Stewart,

No contrarianism there. You're fudging, on the basis of intelligence and experience. From my post above:

Incident light metering works for reversal materials, because the exposure is 'keyed' to the highlights, i.e. you give the MAXIMUM exposure you dare without blowing the highlights to a featureless white. It is worthless for negative exposure unless the tonal range is sufficiently short that the shadows fall within the latitude of the film. They quite often do, but when they don't, an 'uninterpreted' (= non-fudged) incident reading will result in underexposure of the shadows.

After all, ANY metering system can be made to work, with a bit of intelligence and experience.

Cheers,

R.
Perhaps Roger is a little too modest because I don't believe he has mentioned his and Frances's book "Perfect Exposure" which I've found to be one of the most thorough and practical modern texts on the topic I've read.

If you don't mind, Roger, I do have a query leading on from your (now very old) post above.

I use incident readings nearly all the time for transparency, B&W and colour negative, and in general I'm very happy with my results. Granted, I'm probably doing what you discuss with my readings anyway (fudging), by taking multiple readings from a scene, when I think I need to, or simply deciding I should, for instance, give a stop or so of extra exposure for the shadow detail. But I confess I was of the understanding that a good incident meter, correctly used, would provide accurate exposures "keyed" to a mid tone. You've suggested in fact it's keyed to maximum exposure with highlight detail--which is not quite the same thing--or is it? Incidentally (sorry!) I generally use an old but very reliable Minolta Auto Meter III, (and occasionally the optional accessory "spot" attachment for partial reflective readings).

Best,
Brett

Rob-F
04-10-2013, 00:18
I just realized that my incident readings usually call for a little more exposure than my reflected ones, both usually taken with my Gossen Luna-Pro digital F. That would seem to suggest that the incident reading favors the shadows. What could be going on there? I suppose one answer could be that most scenes I meter are really brighter than an 18% average.

sevo
04-10-2013, 00:26
So it looks like the 18% midpoint isn't really the midpoint.

There is and can be no midpoint - the scale of illumination density is infinitely long either side. Or rather, only limited by the size and age of the universe.

18% is not supposed to be the midpoint - it is a average (for scenery reflectance). Worse, it is a outdated average supposedly once determined for aerial reconnaissance photography, and positively off unless you are 10,000ft above a sunlit subject. The modern time use for 18% grey cards is purely for colour balancing, not for exposure determination.

If you have a average (and around 12-14% would be the common averages used for meter calibration throughout the last sixty years), you can place that at the midpoint of your film's useful working range. A more common application would be to place it at some offset, as depending on the film type shadow (negative film) or highlight (chromes) protection will be more critical.

brbo
04-10-2013, 01:42
Another point, about incident meters being best for reversal film. There is an exception. Hollywood Cinematographers use incident meters, but they are not shooting reversal film. It's a negative. So how do they get the exposure right? See Roger's "After all, ANY metering system can be made to work, with a bit of intelligence and experience," above!

This is not an exception as "incident meters work best for reversal film" is simply not true. The same as "reflective meters work best for negative film" is also not true.

Photo_Smith
04-10-2013, 01:43
Pretty much everything you need to know has been covered especially by post no 3 from Roger.
What I do is keep it simple; meter for emerging shadow detail and then stop down 2 stops.
What this does is place the 'emerging detail' at the point of the film curve where density rapidly builds; that is just after the toe but before the linear part.
This will give you good negatives that print or scan well with minimum effort.

Roger Hicks
04-10-2013, 02:45
. . . But I confess I was of the understanding that a good incident meter, correctly used, would provide accurate exposures "keyed" to a mid tone. You've suggested in fact it's keyed to maximum exposure with highlight detail--which is not quite the same thing--or is it? Incidentally (sorry!) I generally use an old but very reliable Minolta Auto Meter III, (and occasionally the optional accessory "spot" attachment for partial reflective readings).

Best,
Brett
Dear Brett,

Another name for incident light metering is 'artificial highlight' metering. What the meter does is to take a reading from an artificial highlight (the incident dome/Invercone) and then recommend an exposure about 2-1/3 stops higher. This gives good mid-tone exposure but without risk of 'blowing' the highlights in a transparency. Very dark shadows (more than about 5 to 6 stops down from the highlight) WILL however be underexposed.

This is what 'keyed to the highlights' means. This is essential with reversal film and digital, where overexposed highlights 'blow' to an irrecoverable white, and this is why I say that incident light readings are better for slide and digital. Slide film ISO speeds are based on highlights, i.e. the MAXIMUM exposure you can give before you 'blow' the highlights.

Negative films, though, can cope with a much longer brightness range. The highlights become increasingly dense, but with wet printing, or a good enough scanner, you can 'see into' very dense areas of the negative.This means that you can afford to 'key' your highlights to the darkest area in which you want texture and detail, and you'll still get highlight detail in areas that are 7, 8, 9 or even 10 stops brighter, at least with black and white, though you'll need to reduce development times with very long subject brightness ranges.

Negative film speeds are based on shadow detail, i.e. the MINIMUM exposure you can give before the shadows 'block up' to a solid black. With black and white you can compress or expand the subject brightness range to fit on the print (or screen or whatever) but with colour, expansions and contractions soon start to look unconvincing -- hence my comment 'at least with black and white', above. On the other hand, once you start dodging and burning, you can represent much bigger tonal ranges on the print or screen.

With small brightness ranges, either approach will do, but with big brightness ranges, the optimum exposure for negative (based on a MINIMUM exposure) will be greater than the optimum exposure for slide or digi (based on a MAXIMUM exposure) This is why limited-area reflected light metering (of the darkest area in which you want texture and detail) is more suitable for negative films.

NO film speed system is based on a mid-tone, but meters can (fairly) safely be based on the empirical fact that most outdoor scenes around Rochester, New York (where the research was done) reflect about 12-14% of the light falling on them. Despite its colour, the 18% grey card is a red herring: a Munsell mid-tone that looks, to the human eye, like the middle grey in a set of cards of varying reflectivities, from about 2% to about 90%, the realistic limit for reflectivity.

Finally, @ RobF: Hollywood. No, they don't always rely on incident metering. A classic exposure technique is the 'key tone' method. Note the word 'key' again. The normal key tone would be the highlight on the cheek of the leading actor. Take a spot reading from this, and key your exposure to this tone, and the actor's skin tone won't change from shot to shot. This, of course, means spot metering, and explains why the SEI photometer with its 1/2 degree reading remained so popular for so long: see http://www.rogerandfrances.com/subscription/sei.html

Cheers,

R.

leicapixie
04-10-2013, 03:25
There are no fixed rules.
Everything varies, light, reflectance, shutter speeds, apertures and film speeds. Be consistent. Make some tests. Keep notes.
When all is needed for a special shot, bracket.
There is no one truth.

Rob-F
04-10-2013, 03:34
There is and can be no midpoint - the scale of illumination density is infinitely long either side. Or rather, only limited by the size and age of the universe.

18% is not supposed to be the midpoint - it is a average (for scenery reflectance). Worse, it is a outdated average supposedly once determined for aerial reconnaissance photography, and positively off unless you are 10,000ft above a sunlit subject. The modern time use for 18% grey cards is purely for colour balancing, not for exposure determination.

If you have a average (and around 12-14% would be the common averages used for meter calibration throughout the last sixty years), you can place that at the midpoint of your film's useful working range. A more common application would be to place it at some offset, as depending on the film type shadow (negative film) or highlight (chromes) protection will be more critical.

I see your point. The key point is average. The average (I should get out my statistics class notes) is not the midpoint. The midpoint is the median. I should get rid of the idea that an incident meter is calibrated to a midpoint.

Roger Hicks
04-10-2013, 03:55
There are no fixed rules.
Everything varies, light, reflectance, shutter speeds, apertures and film speeds. Be consistent. Make some tests. Keep notes.
When all is needed for a special shot, bracket.
There is no one truth.
Fortunately, though, there doesn't need to be. Exposure is, as you imply, a woolly, subjective matter, and (especially with neg/pos) can accommodate quite wide variations. And, as I said, "After all, ANY metering system can be made to work, with a bit of intelligence and experience." It's when you try to reduce it to a rigid set of rules, and worse still when you assume that there is any such thing as 'correct' exposure, that you get in trouble. Of course that's a general 'you' not 'you personally'.

I am not quite sure what you mean by 'be consistent', though, in the light of your earlier (and entirely correct) assertion that 'everything varies'.

And I rather disagree with 'Make some tests. Keep notes.' I'd say, "Take pictures of things that interest you, and rely on your memory rather than taking notes."

Finally, there actually are quite a few fixed rules, in the sense that if you do A, B will happen. Give more exposure and you'll get a lighter transparency or a darker negative. Use an unmodified incident light reading when the subject brightness range is more than 5-6 stops and you will lose shadow detail with negative film. And so forth. The trick lies in learning what these rules are; what assumptions are made by the meter manufacturer; and how to apply all this to get the exposure you want, which is what I define as the perfect exposure.

Cheers,

R.

c.poulton
04-10-2013, 04:14
Just come across this thread - and Roger's summary is one of the best explanations concerning exposure that I have read!

I used to rely on incident light readings for most of my exposures (coming from a slide film background) but I am now going to rely more on reflective readings as I use B&W film almost exclusively. (Luckily I have just recently bought a 'new' Weston Master V) :)

Roger Hicks
04-10-2013, 04:19
With reflective readings for negative film, using a broad-area meter such as the Weston V, it normally suffices to 'favour' the darker areas of your subject matter, i.e. to point the meter towards the darker broad areas rather than the lighter broad areas. See http://www.rogerandfrances.com/subscription/ps%20expo%20neg.html

Cheers,

R.

Sarcophilus Harrisii
04-10-2013, 06:06
Dear Brett,

Another name for incident light metering is 'artificial highlight' metering. What the meter does is to take a reading from an artificial highlight (the incident dome/Invercone) and then recommend an exposure about 2-1/3 stops higher. This gives good mid-tone exposure but without risk of 'blowing' the highlights in a transparency. Very dark shadows (more than about 5 to 6 stops down from the highlight) WILL however be underexposed.

This is what 'keyed to the highlights' means. This is essential with reversal film and digital, where overexposed highlights 'blow' to an irrecoverable white, and this is why I say that incident light readings are better for slide and digital. Slide film ISO speeds are based on highlights, i.e. the MAXIMUM exposure you can give before you 'blow' the highlights.

Negative films, though, can cope with a much longer brightness range. The highlights become increasingly dense, but with wet printing, or a good enough scanner, you can 'see into' very dense areas of the negative.This means that you can afford to 'key' your highlights to the darkest area in which you want texture and detail, and you'll still get highlight detail in areas that are 7, 8, 9 or even 10 stops brighter, at least with black and white, though you'll need to reduce development times with very long subject brightness ranges.

Negative film speeds are based on shadow detail, i.e. the MINIMUM exposure you can give before the shadows 'block up' to a solid black. With black and white you can compress or expand the subject brightness range to fit on the print (or screen or whatever) but with colour, expansions and contractions soon start to look unconvincing -- hence my comment 'at least with black and white', above. On the other hand, once you start dodging and burning, you can represent much bigger tonal ranges on the print or screen.

With small brightness ranges, either approach will do, but with big brightness ranges, the optimum exposure for negative (based on a MINIMUM exposure) will be greater than the optimum exposure for slide or digi (based on a MAXIMUM exposure) This is why limited-area reflected light metering (of the darkest area in which you want texture and detail) is more suitable for negative films.

NO film speed system is based on a mid-tone, but meters can (fairly) safely be based on the empirical fact that most outdoor scenes around Rochester, New York (where the research was done) reflect about 12-14% of the light falling on them. Despite its colour, the 18% grey card is a red herring: a Munsell mid-tone that looks, to the human eye, like the middle grey in a set of cards of varying reflectivities, from about 2% to about 90%, the realistic limit for reflectivity.

Finally, @ RobF: Hollywood. No, they don't always rely on incident metering. A classic exposure technique is the 'key tone' method. Note the word 'key' again. The normal key tone would be the highlight on the cheek of the leading actor. Take a spot reading from this, and key your exposure to this tone, and the actor's skin tone won't change from shot to shot. This, of course, means spot metering, and explains why the SEI photometer with its 1/2 degree reading remained so popular for so long: see http://www.rogerandfrances.com/subscription/sei.html

Cheers,

R.
Hi Roger,
thanks so much for expanding on this. It's given me some things to think about. As I mentioned, I'm generally fairly happy with my ability to get the exposure I want using an incident meter but as you described it, I do "interpret" the meter readings. I did not know the basis for calculating the "box" speeds is different for pos and neg, but it makes sense—thanks for pointing this out.

There are some good lessons to take out of your comments. I may have to try doing some reflective readings off the shadows in future and perhaps slightly reducing the development time (for black and white) due to the increased exposure. Highlights aren't usually a problem for me, and shadows are OK, but I now find myself wondering if I might get even more shadow detail this way. Much appreciated and thanks to all who have contributed to this interesting discussion.
Regards,
Brett

charjohncarter
04-10-2013, 09:21
Wonderful example John!

Just to add a little to the discussion, Bruce Barnbaum published an article titled "Photographic Myths: There are ten zones in the Zone System" published in the Sept/Oct 2000 issue of PHOTO Techniques. In the article he explains how he regularly gets 13 stops of range recorded on film. Of course he was printing traditionally, and dodging and burning heavily to extract all the detail recorded on the negative. The article was reproduced in his publication "B&W Master Printing Class" in 2006.

A negative with that exposure range is beyond the dynamic range of most scanners, and will require several scans at different exposure levels, which can be blended in Photoshop (assuming the scanner can even see the detail in the most dense parts of the negative).

AND raytoei,

I used one of Roger's suggestions (from his excellent book co-authored with his wife: Perfect Exposure) which was to do a reflective and an incident reading and go in between. Not very scientific but it worked here. Also, you might take time to read this pdf.

http://johnsexton.com/images/Compensating_Development.pdf

I do a variation of this development, but then I'm not generally trying to get those astronomical tonal ranges.

Roger Hicks
04-10-2013, 10:45
Hi Roger,
thanks so much for expanding on this. It's given me some things to think about. As I mentioned, I'm generally fairly happy with my ability to get the exposure I want using an incident meter but as you described it, I do "interpret" the meter readings. I did not know the basis for calculating the "box" speeds is different for pos and neg, but it makes sense—thanks for pointing this out.

There are some good lessons to take out of your comments. I may have to try doing some reflective readings off the shadows in future and perhaps slightly reducing the development time (for black and white) due to the increased exposure. Highlights aren't usually a problem for me, and shadows are OK, but I now find myself wondering if I might get even more shadow detail this way. Much appreciated and thanks to all who have contributed to this interesting discussion.
Regards,
Brett
Dear Brett,

(Highlight): You shouldn't need to. Reducing development time will (fractionally) reduce film speed and therefore shadow detail, but what it mostly reduces is contrast, so you might well need a harder paper grade or a digital work-around to the same effect. With the kind of extra exposure you're talking about -- maybe a stop -- you're probably as well off without reducing dev times.

Cheers,

R.

Joakim Målare
04-10-2013, 10:59
"Expose for the shadows, develop for the highlights" - for me this is yet another very simple statement I have overlooked for far too long. For some reason, the shorter something is phrased, the less likely I am to pay attention.

Some thoughts:

1. I understand that a true spotmeter is the best choice. If I use a manual camera, I would prefer one of those. But since I have a combined incident/reflective meter (30-ish degrees I guess), that will have to do for now. If I want to measure a shadow area, will I get the same value by walking over to it and "filling up" those 30 degrees as I would get using a spotmeter from a distance? Does the distance to the surface make a difference to the intensity of the measured light? I could find out on my own, but for the sake of discussion...

2. I read about measuring ones hand in shadow. What's the problem with that method, provided I measure it in the same ambient lighting as the shadow area I want detail in and adjusting accordingly? It's a bit of a compromise, but also faster, no?

3. Similarly, what if I take an incident reading from the sky without incoming sunlight and basing my exposures on that, for shadow detail in the same ambient lighting? The fact that an incident reading is keyed to the highlights doesn't mean I can't use it for shadow detail, right? This is valid for open scenes - in darker scenes, where the skylight is more or less omitted, a broad reflective reading of shadows becomes easy anyway.

Note that I don't know what I'm talking about - just asking for someone to confirm or deny my thinking :-) I realise I'm looking for shortcuts where experience from proper practice would be of better use, but it's also interesting to hear if one actually understand anything of the subject...

Sarcophilus Harrisii
04-10-2013, 17:52
Dear Brett,

(Highlight): You shouldn't need to. Reducing development time will (fractionally) reduce film speed and therefore shadow detail, but what it mostly reduces is contrast, so you might well need a harder paper grade or a digital work-around to the same effect. With the kind of extra exposure you're talking about -- maybe a stop -- you're probably as well off without reducing dev times.

Cheers,

R.
Hi Roger,
thanks for that, good advice. Consistency in controlling the variables involved in producing an image are a good basic principle to work by, (as you and Frances mention in several different contexts, in various chapters of your books), whether that is for example: in the use of a particular reference thermometer as a master; how far you immerse it; your inversion regime, or; the temperature, and times, you generally use. So it's a much better idea, I think, to change one variable at a time—especially as I've been fortunate, over the three years or so I've been processing my own B&W (exclusively in ID-11), to hit the ground running with results that I like.

Here's one I scanned just a few days ago. Taken not far from my home late last year, I found the subject during a walk around the district here in Tassie. It was Pan F Plus @ EI 25 in ID-11 1:3. Processing temperature was closer to 21.5 degrees than my usual 20C when it was processed (measured with a spirit thermometer inside the centre core of the Paterson tank), because it was a 35C-plus day, and I was throwing ice into the water bath and sitting the tank into it between inversions, trying to keep it down near 20C. Camera was my ZI Contaflex Super BC & 50mm Tessar f/2.8 and scanned with Epson V700. I quite like the tonality and exposure overall, but there's not much shadow detail looking around what is left of the chassis and under the mudguards. Highlights are well contained, and experimenting with the highlight slider in Photoshop brings plenty of detail back on the roof for instance.
Cheers,
Brett

http://farm9.staticflickr.com/8546/8635930373_3ea93deb77_c.jpg

Joakim Målare
04-10-2013, 19:24
That looks great Brett!

I like how I can connect the highlight on the roof with the little bit of horizon visible at the top. I think the darkest areas balance the otherwise grey picture well. Not sure it would improve with more shadow detail!

Sarcophilus Harrisii
04-10-2013, 21:26
That looks great Brett!

I like how I can connect the highlight on the roof with the little bit of horizon visible at the top. I think the darkest areas balance the otherwise grey picture well. Not sure it would improve with more shadow detail!
Thanks for your kind words. :)
Cheers,
Brett