SLR Meters vs. Hand Held Meters

[jake]

Obviously the two are very different. I recently acquired a Canon EOS 1 and a Nikon F4, probably a dual act of insanity brought on by an intense fear of death. Or of my mother. Something Freudian I am sure. Anyway, so far I have shot quite a number of rolls of Ektachrome 100Gx and am interested that both cameras seem to routinely underexpose, not through anything faulty with their meters, but due to being (seemingly) easily fooled by tricky lighting - bright sky, dim foreground, 80% shade with 20% sun, etc. Even with the spot metering on, a subject standing in the shade will be under exposed if there is a sunny area nearby, unless I meter off a 100% shady spot, lock exposure, and then recompose and shoot.

Or if I do everything manually with a hand-held meter. The hand-held meter seems remarkably flawless. I use an incident meter - a Sekonic L308 - and it seems, even with the Ektachrome, to get it right (enough) every time without much fooling around on my part. Recently, I shot 10 rolls on my family's week long summer vacation. I have maybe two or three shots that are not exposed properly, using a plain-prism Nikon F and the hand-held meter. The other thing that gets it right is my Leica M7. The meter in that camera must be very good, because it seems much less susceptible to stray beacons of illumination than either of the zippy SLRs.

I am wondering why.

A couple answers I can think of on my own are that: 1.) the Canon EOS 1 and Nikon F4 are old technology and meters in cameras have gotten better; 2.) I am more used to working with the M7 as I can shoot with it in my sleep; 3.) the relative complexity of the cameras is creating a situation where I am actually not paying attention to everything and thus am missing situations where the camera is going to make an error.

Are the new digital SLRs better at metering or is it merely that one has the immediate evidence of the meter's judgment?

[jmi]

Well, FWIW I find the same to be true... IME, a handheld incident meter is almost totally foolproof in a lot of situations provided it is used correctly, whereas any in-camera meter involves some guesswork.

I have tried dSLR meters and don't find them to be much better than centre-weighted for my purposes - they get it right more often, but the failures are still there, and importantly, are not very predictable - matrix meter readings can fluctuate wildly with compositional changes.  The centre-weighted meter is a simple beast and its flaws can be learned, given enough time...

Anyway, my favourite in-camera metering method is good old 60/40 or better still 80/20 (Nikon F3), using the usual AE-lock-recompose trick.

I believe the M7 has a similar metering pattern to the F3 - very centre-weighted - although somebody will probably correct me on that

[melek]

The best thing about incident meters is that it can't "see." It can't see that dark red brick wall or that white garage. It only detects light falling on a white sphere or white cover and makes its reading.

In general, an incident meter isn't fooled.

With all meters, your experience and judgment can help.

I've also read that most digitals tend to underexpose in order to preserve highlights.

And that's about the limits of my knowledge.

[Alan Gage]

I've never been very happy with the metering of my DLSRs and don't trust them in most situations. I used to be much more satisfied with the meters in my old film cameras, but I think that's mostly because I shot print film/B&W back then and wasn't so darned picky.

I almost always shoot my DSLRs in manual mode using either sunny 16 or my Sekonic L-308 in incident mode. Results make me very happy.

I usually don't pay much attention to what the camera meter says as I'm dialing in the correct exposure but sometimes it's right and sometimes it's not. Sometimes it's WAY off, but that's just the way meters work.

I suppose if I used the camera in AV mode nearly all the time and got to know how it metered and what fooled it I could easily dial in exposure compensation to take care of it. I don't shoot enough for that though and I'm happy the way I'm doing it now. Who knows, maybe I'd be surprised how often it did get it right if I gave it the chance.

So I guess no answers for you, other than don't expect it to be much different when you get a DSLR.

Alan

[Raid Amin]

I have never used an incidence meter, and I got used to the hand held digital spotmeter. This does not let me get wrong readings when I use the Canon T90 built-in spotmeter. My start was with slow slide film, which required very careful metering to get what I wanted. I read several book on tricky exposures to refine my technique in taking a meter reading. Now, I feel safe with a spotmeter and no other type of meter.

[jake]

Yeah, my friend Jerry shoots a ton with his 5D and never uses the meter. He always uses a handheld incident/spot meter. And his photos look terrific, so along with everyone's own experience here, I think I am on the right track keeping the Sekonic around my neck.

[Mike Kovacs]

I love my incident meter.  It always goes with me but remember there are some situations where you will have to compensate with it, e.g. slide film with a very white or black scene.

[bruce]

The meter in a slr is just a reflective meter, no different than a handheld. I guess some of the slrs have pretty sophisticated algorythms for calculating the exposure, but they are still reflective meters, with all that entails.
 The only modern camera I own, a Contax G2, has a very good meter. It took me quite a while to realize I really could trust it. It does nothing fancy, but seems to do it well. I still have to compensate for backlight and such, but that is the nature of the beast.
 A reflective meter is just a different tool than an incident meter, and each used within their limits will work.
 The downside of an incident meter is that you don't get to choose where your tones fall. They will be in their "proper" relations to each other, but that may not be where you want them.
 I find that using an incident meter, along with a divided developer, gives me the highest percentage of usable negatives on rolls with varying contrast and lighting conditions, which if I am banging away with 35mm, is what I want.
 If, on the other hand, I have the luxury of having consistant lighting for the whole roll, or shooting single negatives, it makes much more sense to use a reflective meter to place the tones where I want them and to be able to adjust my development to control the contrast.
 Different tools for different things.

[grizzz]

I have had fairly good luck with the Canon DSLR meters. To the point where I do not carry my L358 with them. With the DSLR's there is the instant feedback. When using film cameras both new and old I have found the same problems as listed here. I use the L358 and seem to get more properly exposed shots than if I were to trust the in camera meters. I have the CV meter 2 as well that fits in a hot shoe. It is funny to use it on a camera with a meter just to see the differences. Of course meters are built to meter differently so these differences are normal. For example my OM2N meters a full stop off of the CV meter. Sometimes the Oly is right and sometimes the CV is. The L358 usually is. I think the incident readings are better than reflective if you are able to get to the light the subject is getting. Of course I also think I need to learn how all metering modes work better on the newer cameras. Something I have not really explored all that much yet. I usually use the center weighted option in M mode and just adjust if I am not seeing what I want on the LCD .

[Andre Reinders]

I tend to rely on the meters in my modern (D)SLRs - I have an older Gossen incident meter that takes the 1.3 volt batteries, but I don't really use it. I need to convince my self it is working, and then use it with the oldies.
 
When I shoot a meterless camera, I either rely on the Sunny 16 rule - or meter with a (D)SLR.
 
Recently I was out shooting lily pads and flowers with a friend - me with my Canon 20D, and him with his Nikon D50 (or maybe D40).
 
He was asking advice about using his meter (he was shooting on Manual - and I was on Aperture priority and using exposure compensation) I told him how I looked at a scene, and if there was a lot of dark water I would put the EC at about -1 2/3 and then see what the histogram said, and if there was lots of green, I would use -2/3 and then 'chimp'.
 
He started to use my technique, and I noticed that we got very different exposures. Both cameras were at the same ISO, both on center weighted avg. metering. I had a polarizing filter on - he did not, but we set our cameras to Av and about -1 EC. My photos were about 1/3 underexposed, and his were about 1 2/3 undexposed - this from looking at the histogram. I could not believe the difference in the meters! I had heard that some complain that the Canons shoot about 1/2 overexposed - has anyone used both Nikon and Canon and noticed this type of difference?
 
(sorry a little off topic here...)

[jake]

Interesting. Not to keep quoting Jerry, but he knows a ton more about photography than I ever will - he says that he has come to mistrust the histogram on his 5D. He underexposes everything, sometimes as much or more than 2 stops. Since he is shooting RAW, he does this to keep the things he wants in the skies and highlights, and then he relies on RAW's ability to record things even in the deepest shadows & pulls out the details that he wants though adjustment in post-processing.

So I would imagine that his histograms would be way to the left, but it also sounds like he doesn't look at them or at the image on the LCD either.

[Alan Gage]

I always used to do the whole "keep the histogram to the right" that seems to be the mantra with digital but found I got very poor results that way. Well before the highlights would clip the lightest tones seemed to get very compressed so there was little to no separation in them.

I'm sometimes amazed how far my histogram is to the left when I know I got the exposure correct with my incident meter. And lo and behold when I look at it on the computer screen the exposure was correct.

I used to look at my histogram all the time but anymore I usually only do when I want to see if the highlights are clipped or not.

Sounds like Jerry's a good guy to know.

Alan

[jake]

He is. So is Sam. But both of them are equally frustrating because they have light meters in their heads, so quite a few times I find myself shaking my head in agreement without actually understanding anything said in the last ten or so minutes. Fortunately they will repeat themselves and use hand signals.

Heck, I can barely remember my phone number.

[mdcarma]

Heck, I can barely remember my phone number.

"Operator, could you trace this call and tell me where the hell I am. Looks like I'm on the corner of walk and don't walk, but that keeps changing."

[jmi]

"Operator, could you trace this call and tell me where the hell I am. Looks like I'm on the corner of walk and don't walk, but that keeps changing."

The worrying thing is - they probably can if you're using a cell phone.

Andre - my own experience says the same - every matrix meter seems to be different.  This is alluded to in many of the dSLR reviews.  I expect if you put them both on centre-weighted they would probably be much closer with the same composition.

I think this business about having to guess the exposure compensation is just what I was trying to get at - in the digital world it's not that difficult to learn since you can chimp, but with film it's going to take an awfully long time and a lot of waste (especially if you use velvia).  On the other hand old film cameras all have very similar metering patterns, so I find I can interchange them without much trouble.

I agree the incident doesn't make it too obvious where the tones will fall, except that an 18% grey subject in the same light as the meter ends up in zone V - I'm definitely not a zone system guy, for that one has to use a spot meter.

Jorn - I too aspire to not needing a meter at all on Tri-X but I need to get out and practice to make that true