Incident

Bill Pierce

Well-known
Local time
2:31 PM
Joined
Sep 26, 2007
Messages
1,407
Digital images are a bit like color slide film. Overexpose them and the highlights are reduced to detailless digital cellophane. Consequently, some TTL meters on digital cameras are set to slightly underexpose many scenes and guarantee that highlight detail is preserved. This is a good thing, sort of. It preserves highlight detail but reduces shadow detail and the overall brightness range captured in the file. Of course, if you are a landscape or still life photographer, you can check the histogram or, even simpler, make a series of bracketed exposures. But for street photography, news or just plain family snaps that’s not a realistic option.

As of late, I’ve been going back to a technique that news photographers used when quickly editable color slide film was the choice for journalism. Bracketing exposure simply wasn’t possible in many situations where a single moment told the story. The answer was to use a handheld incident meter because it favored correct exposure for the highlights and, at least where the meter was held, didn’t blow out those highlights.

Guess what, it works for digital color, too. The ISO you set your meter at may not be the same as the one you set your digital camera to, but, with experimentation, you can come up with a setting that gives you the fullest exposure (the most shadow detail, the least noise) that still preserves highlight detail, all-in-all, an exposure that gives you the greatest tonal range. (And, since that is a consistent, unchanging exposure, unlike automatic exposure, processing a burst of shots to a consistent appearance is often easier.)

So, stupid as it sounds, I am often using an old fashioned, hand held, incident meter with my modern, auto exposure capable digital camera. Any thoughts outside of the one that I may be an idiot?
 
Don't you just get to know what your camera will do in a given situation and compensate? I shoot an Epson R-D1 predominantly in AV and they are famous for under exposing and I've just learned to up the EV when needed. It's on the dial on top of the camera and easily accessed. Kinda like shooting full manual only better. Most of the upper end mirrorless have the same dial. DSLRs have a fairly easy adjustment too.
 
I would rely on my histogram before introducing a handheld meter into the equation, but if you get good results from a handheld meter, it's hard to argue.
 
"I would rely on my histogram before introducing a handheld meter into the equation."

I think that makes perfect sense except in those cases where the histogram is based on a jpg and not the raw file and the two are significantly different or when the eye level finder doesn't have a histogram or you feel that the histogram and other overlayed info interfere with your composing the image.

I'm often in situations where doing anything besides paying attention to the subject and pushing the button at the right moment isn't too wise. So having a manual exposure preset or relying on auto exposure is the best route. When it's possible, setting that optimum exposure manually gives me the best results.
 
I'm a big fan of incident meters, both for ambient and flash conditions. On anything important, I use a color checker card and an incident meter. It's also great for easily setting strobe lighting ratios, and for precise metering for video capture. Guessing that we are of the same era, since you mention shooting lots of transparencies.
 
With film I would choose an exposure, manually set the camera, and shoot until the light changes. With the current digital cameras one eventually learns to compensate based on the situation. ISO-less-ish cameras / sensors might make underexposure a better choice than trying to nail the exact exposure.

I have gone so far as to measure the color temperature or used other techniques to pre-set the white balance.

Even when using a spot meter when shooting digital, you might want to calibrate and/or know your equipment.

The other day I was blasted by some friends for using NDGrad filters rather than stacking...

One should take advantage of technology... returning to an incident meter is a good way to validate our exposure decisions. Sometimes the results of a 'rethink' are eye opening ; )
 
you can come up with a setting that gives you the fullest exposure (the most shadow detail, the least noise) that still preserves highlight detail, all-in-all, an exposure that gives you the greatest tonal range. (And, since that is a consistent, unchanging exposure, unlike automatic exposure, processing a burst of shots to a consistent appearance is often easier.)

So, stupid as it sounds, I am often using an old fashioned, hand held, incident meter with my modern, auto exposure capable digital camera. Any thoughts outside of the one that I may be an idiot?

SI Photographer Peter Read Miller teaches something similar. When we first get to a sporting event, we do a series of test shots, seeing which exposure gives us the best highlight detail, tonal range, and uncrushed shadows. Then we set Manual exposure on our wiz-bang DSLR's and leave it there (as long as the lighting doesn't change) for the whole game. Seems kind of weird, over-riding the Matrix metering on the latest and greatest DSLR, but it works.

Best,
-Tim
 
Far from being the village idiot, some times out of the mouths of old men......

First let me say I am an iPhone-ite when it comes to digital. Everything else has gone into the 2-B-Fixed drawer. So I can't prove you are right or wrong, but damn it sound right on target to me.

I grew up on Centerweighted and got used to figuring out where to meter so I get the exposure I want. When I moved to RF I found incident worked much better than reflected (even on my M6). I metered and remembered Bright-light and shadows and used those two settings for most, adjusting if it felt right (different light from those two).

When I think about built in meters in cameras over the years, the OM3/4 stood out to me for control, but like looking at a histogram it takes time. The idea of have a camera auto-bracket sounds like fun, but I'm a decisive moment type who looks around for reactions to what I just shot.

The technology is there but the focus is on focus, exposure is good enough. Or perhaps as good or better than many film cameras of years past. Remember automation is only as good as the software. While you can look to find the eyes in the frame and focus on them (hopefully only one pair) but exposure isn't so binary like focus is.

I suspect that looking at the raw images that result from your approach to shooting would yield for most people. I loved shooting chrome because the prints that came out of the one hour machines back in the 80's were average. I felt chrome kept me closer to my true exposure.

But I think there are people who are used to using auto-exposure (perhaps auto-everything) cameras that are equally as quick adjusting the over/under control to get the look they want. Then other folks can work the black-magic (not that there is anything wrong with it) in post (used to be called the dark room) that can bring the picture to what they want.

I think it's a better way of going, incident metering and perhaps confirm with a histogram. Maybe it's time to buy an old Nikon D60, pull out the sekonic and a couple of old manual lenses and see what I get.

B2 (;->
 
I find reflected light (camera) metering with "decisions" is the second best method for street and other kinds of fast photography... First one is "knowing" the light we're in...
Incident metering is best for studio work or light situations like those inside a studio: not the majority of situations... Outdoors there are many times the light I'm in isn't the same light the subject's in, and incident metering can't be used... So, in the end, I prefer (if I can't guess the metering) metering with my camera and telling if the scene is middle gray or +1, etc., and for me that's the fastest and most accurate way in those cases I'm not sure about metering because light starts to be low...
 
Sounds reasonable to me - up to a point. But no this does not make you an idiot in any event. Problems with correct exposure in digital camera is one of my pet peeves. I am sure I lose more shots from this than from any other problem. My problem mainly is with blown highlights and I habitually rely on deliberate under exposure (and problems with shadows) to deal with this. I prefer to lose shadow detail than losing highlight detail as the latter is far uglier and damages an image - period. So if a hand held meter helps you select a more appropriate exposure then, good for you.

But here is the thing as I see it. If the contrast range in a specific image is higher than your sensor can handle no matter how you meter and no matter which exposure you select you will still end up with either blocked shadows or blown highlights or both. The only alternative being to exposure stack and fix in post. (Which I sometimes try if the image is static as in a landscape etc). so your technique may help, I acknowledge. But only to a point.

How I long for the good old days when I put a roll of color print film in my Leica M3, set my hand held light meter for the speed of film being used, stepped outside, metered once with the incident meter for full sunlight. Then metered one more time for shadow (usually a stop was needed for moderate shadow perhaps two for deeper shadow) then committed this to memory. I then selected an aperture and shutter speed and stuck to it until I passed from sunlight into shadow or shadow into sunlight. The rest of the time I just ignored my exposure meter and relied on print film's flexibility. Sadly like you say sensors are more like slide film.
 
In terms of family candids or street work, I find the auto-aperture bracket method to be an effective compromise. I use the following strategy with raw files. This approach is inappropriate for in-camera JPEGs

  • I chose manual exposure and typically use 1/125 or faster shutter times. Even in extremely low light, I don't go below 1/60.

  • My camera has an 8 frame/sec. capture rate. The buffer holds 25 raw files. With a fast card, the image write-to-memory speed is ~1 raw file/sec. So I get at least 8 different compositions every 30 seconds.

  • The auto bracket sequence is 0, underexpose, overexpose. So the first shot is most likely to be an exposure keeper. Typically, the other two are insurance.

  • I usually use 1/3 stop exposure steps. In bright sunshine, I increase the steps to 1/2 or even 1 stop.

  • When spontaneity is a priority, I choose a conservative aperture and only use one of three ISOs – 200, 800 or 1600.[a] I only use the meter to avoid gross overexposure. Typically all three exposures end up being underexposed. This method often compromises shadow region S/N.


  • During post-production, I choose the raw file with the most appropriate highlight region exposure. This means the shadow regions' exposure is optimal. I render the global brightness until the highlight regions clip. This is essentially identical to setting ISO manually for in-camera JPEG files.[c] Then I selectively pull the highlight regions. Next, if needed, I selectively push the shadow regions. The other two raw files are deleted.

    a/ These ISOs are based on my camera's (X-Pro 2) data stream design. Different cameras have different designs. But they all have ISO ranges where the electronic noise levels are optimal.

    b/ Real-time image review is often useless. This is a minor disadvantage because in spontaneous situations, chimping wastes time.

    c/ My camera's electronic noise level increases by only ~10% from ISO 800 to IS0 3200. But the analog dynamic range decreases by ~2 1/2 stops over the same ISO range. Between 200 and 800, the DR decreases by less than 1 1/3 stops.
 
I have the histogram pop up with each image replay. Then I see where those 'cellophane' highlights are. I just about only use my digital cameras indoors for theater work. Believe it or not, these shots can be very contrasty. I have time to adjust so I shoot one, check the histogram then adjust the shutter to get it where I want (AV mode). Dragging the shadows up is sometime impossible. (I like RAWTHERAPEE for shadow recovery.)

I have my incident meter with me when I shoot digital theater promos outdoors, it is usually right on.
 
I have a couple of Sekonic incident meters I use for digital work, especially for the Leica Monochrom - it has very little tolerance for overexposed highlights, and unlike slide film you can pull a lot of detail out of the shadows. If the light is pretty constant, I just take a couple of readings, set manual exposure and forget about it.
 
I invariably use an incident meter with the film cameras. Except the M6. It has a very good metering system and seldom differs from my incident reading exposure determination. The M9 and Monochrom can't be used the same way as the M6 and an incident meter would work in the same way it does for you. But I just know the cameras and as Juan says, consider the particular scene. I think many of us using those cameras shoot manual, either from the outset, almost sunny 16, or having taken a test shot on auto and checked the histogram. That is not much different to what you propose. As you point out, the object is to be all set for the next shot at a half second's notice and with good exposure, preset manually.
 
What is quicker, three bracketed auto exposures or exercise with the meter and setting camera for it manualy?
Also. Talking about street and news. You metered it and at very next second cloud came over the sun. Now what? Yes, do it again. Measure and hope light it not going to change again.

Idiot or not, using same camera, same lens for street, news and family will sooner or later lead you to take it right once and in the right moment. Even with awful M-E TTL metering 🙂.
So, for the street I use S16 or measure it for the bright and dark part of the street once, then adjust momentarily for film or I use digital camera with shutter and speed set to the average light for this place and time and let camera set the ISO.

Highlights... All I care for is people faces. Events, street and family pictures. The rest is recoverable from the RAW file format most of the time.
 
Back
Top Bottom