Light meter - the mood killer

[Paul J]

Advertisement (gone after registration)

The day a camera can think for its self will mean that there is no need a photographer. So, at the moment, a photographer can only accept they need to learn how it works. We wouldn't be here answering if we didn't want to try and help - Why must we sugar coat everything?

 

Generally for a dim interior, if you are shooting with A, start by adjusting your exposure compensation to somewhere around -2/3 of a stop (you can program exposure comp to the thumb dial on the back of your M9 in your settings menu).

 

If you keep in your mind that a camera will make everything mid tone grey (i.e. it will make it brighter if it's dark and darker if it's bright) you are on the right foot.

[sblitz]

all meters essentially bias to mid-grey values at noon. if you want your photo to reflect the mood you see there are two options -- one, point the meter (lens) at the lighted object where you want it brightest and then speed the camera two stops faster (or close the lens two stops). two, take photo as you did and lower the exposure (or play with the curves) in the computer -- the easy option not available in the print and develop days for those who did not do their own printing.

[dbraid]

I guess some of you guys missed my point. There's no need for a light meter to know what I think, I never asked for it. I'm aware it can be tricky for the meter in case of extra light sources, but the elements in the samples I took are more or less evenly lit from the light comming through the window to the right.

 

I simply stated the camera should expose properly i.e. the way scene really looks like, the real light amount in the scene.

 

But the camera is not measuring the "real light" in the scene, it is measuring the reflected light from the wall and other objects and this depends on the incident light and what objects are reflecting the light. Imagine the exact same scene with a black wall. How will the camera know that the wall is black and how to compensate to make the wall appear black in the exposure? You might come closer to what you imagine by using an incident light meter to determine the light falling on the objects and to calculate an exposure that is independent of the reflectance of the objects in the scene. Yet it is even more complex than this due to the neurophysiology of human perception. Our minds process information and "see" very differently than a sensor or film sees. There are lots of good sources on this.

[TomB_tx]

This is certainly not unique to the meter in a Leica. In other forums on dSLRs people marvel how the low light performance makes every picture look the same as in daylight - and they think that's a good thing. That's why all good cameras have controls to allow the photographer to compensate the exposure to show the mood he wants.

I've always preferred the spot (or semi-spot) meters of the M5 & CL, Leicaflex SL, etc, as I like to meter off a part of the scene that I know should appear neutral density. However, as you gain experience with any meter type you learn how it will present different scenes, and you learn to predict and compensate to get the effect you want.

[tobey bilek]

Mood will just give a dark underexposed pic.

 

Go to manual and underexpose or use exposure compensation.

 

What will make these look nice is a small flash at a slow enough shutter to pick up some ambient light, 1/8 or 1/15, yet keep the subject properly exposed. .

[furcafe]

The problem w/your statement is that until recently, cameras had no way of knowing what the "scene really looks like." All you had was a meter that measured light, not the relationship between the different lighting & tones that make up a scene the way a human sees it.

 

The closest thing to jaapv's "Lightmeter of Delphi" is matrix metering in modern cameras. That technology is computer-driven &, if I understand it correctly, works by having the camera compare the lighting w/an internal database of different scenes, i.e., primitive photographic artificial intelligence. However, even when there's a match between a given scene & the database, the camera still can only guess & provide (or recommend) an exposure based on what the programmers thought most people would want a scene to look like, not what a particular photographer may see or want to achieve.

 

As you have discovered, Leicas are not the cameras you want to buy if you want the latest in computer-based photo technology.

 

I simply stated the camera should expose properly i.e. the way scene really looks like, the real light amount in the scene.

[aesop]

Advertisement (gone after registration)

...great thread - even if only for making the 'auto-mob' pause to think about exposure AND to realise that meter readings are merely suggestions that may need to be interpreted.

[pgk]

The closest thing to jaapv's "Lightmeter of Delphi" is matrix metering in modern cameras. That technology is computer-driven &, if I understand it correctly, works by having the camera compare the lighting w/an internal database of different scenes, i.e., primitive photographic artificial intelligence. However, even when there's a match between a given scene & the database, the camera still can only guess & provide (or recommend) an exposure based on what the programmers thought most people would want a scene to look like, not what a particular photographer may see or want to achieve.

And when all is said and done 'matrix' metering is simply a way of automating exposure assessment and as such isn't really a good thing to use if you want to understand and learn about exposure and metering.

 

IMHO the best way to learn with a digital camera is to switch it to manual (and centre-weighted or even spot metering if you can) and experiment, noting down the adjustments which produce the results you want - exposure is not always about being right, wrong or correct and to some extent will always be about what you are trying to achieve in the final image.

[pico]

A camera meter will only see, what is known as "18% grey". You need to compensate for it using the zone system.

 

I know what you are trying to say, but technically it is 12%.

 

.

[Paul J]

PGK is right. A Matrix meter will only prevent sever underexposure from something like a bright highlight in the frame. The M9 is terrible for this as you know being a classic centre weighted design. If it has a light or anything bright in the frame it will under exposure quite a lot to make that bright light mid tone grey.

 

A Matrix style metering system will is programmed to consider the whole frame and see a bright highlight and then average it with the rest of the scene. It will though work in the same way though, seeing only Zone V. So if you want a dark scene you will need to under expose it in the same way.

[pico]

Is there a way Leica could improve this, or this is something all cameras have, and I have to live with it?

 

The M240 has "Multi-field" metering in Live View if you wish it. It tries to predict exposure by reading 24 different locations rather than center-weighted (default).

[Paul J]

I know what you are trying to say, but technically it is 12%.

 

.

Hey Pico! There you are!

 

Technically neither is conclusively right. 18% for reflective readings and 12% for luminance. Ansel Adams actually thought that 14% was best, apparently.

 

I am still under the impression, though, the M9 like many cameras is 18%. I'm sure some one here can verify that though.

[bab]

Many times when I use the A mode, and trying to take a photo of someone in the room which is a bit dark or has the "mood", the meter is trying to compensate and all photos look as if it is a daylight. I even miss a shot because of too slow shutter speed due to this phenomena.

 

Not sure what ISO you had set, but taking a picture of a person in a room would'nt you want to meter off the persons face? That would most likely differ from the ambient room light in your scene depending on the position of the person. With an ISO of 640 and if your using a fast lens you should have the ability to meter your hand with the same light falling on it and quickly get to the right exposure. With the ISO set at 200 you would not have as much breathing room and struggling to find a comfortable shutter speed and F stop combination to nail a reasonable exposure. Controlling the ambient indoors and adding some fill light by reflector or strobe would be a very easy effective way to have proper exposure on you subject an then fine tune the room for the mood.

 

I just finished a shoot in Hong Kong, I was mainly shooting at 4:30 to 6:30 pm in the streets and shooting into the light, trying to capture that harsh direct sun. My experiment with the A setting is not conclusive yet as I just returned and still have another 400 images to go thru. So far without looking as to wether the M240 was set on A or if I was using Manual the images are comparable. I find with the meter set to center weighted to be very effective and predictable to judge.

 

What is evident is the way the chip handles the contrast of he scene in all but very few images the orangy haze almost noise looking is a challenge to correct. I'm planning on sorting the images by lens choice also to see which lens I prefer for that situation I'm leaning towards my 21 3.4 but the 35 or 50 Lux's may win out.

 

Temperatures outside were in the ninety s and everyone's face was shinny from perspiration as they were mostly facing me with the sun behind them there was quite a gap in exposures.

So far I have really gotten great exposures with the M240 and can't complain.

 

One trick I use is to meter the street and lock the exposure if I want the face on the money, or I might meter the sides of a building that isn't in direct sunlight both methods seem to produce great reaults. The worst case scenario would be to allow the center weighted meter to be pointed at the sun or the strongest direction of light that proves to be disaster.

 

Interestingly I questioning why when I use exposure lock by half pressing the shutter in continuous mode the lock doesn't hold for pictures 2-x. I haven't confirmed this yet but the numbers in he viewfinder when camera is set to A mode seem to reflect it?

 

Regards

[sblitz]

when you are in that kind of light situation you are exactly right -- meter a wall or the street (streets/sidewalks are usually grey as well). as for the lock afterwards, once you have that meter reading take the camera off of A and set the speed accordingly. one thing you learn using a manual camera without a built in meter is that light is more constant than you think. take a reading and understand where light change means 2 stops open or closed. and then just take pictures. it's liberating not to be adjusting and fiddling for each shot in the same spot. and if you are set at f8 or higher and using a 35mm or lower, you can even turn your very expensive camera into a very good point and shoot.

[TonyField]

The light meter did exactly what it was supposed to do - generate an "18% grey" exposure for what ever the meter sensor saw. Since you are concerned about the "mood" (a subjective interpretation unknown to the camera), you really should chimp and possibly use EV compensation to correct the issue. Or use manual exposure teamed with chimping. Or point the lens at an "appropriate" subject tone content, press the shutter button half way to lock the exposure, then recompose for the real shot.

 

These are not "work arounds" - they are simply proper use of the light meter system.

 

The EV compensation mechanism is put in specifically to allow adjustment of the basic lightmeter functionality.

[jaques]

indeed the light meter is not the 'mood killer'- it is the 'mood equaliser' to get the desired mood you are after- use the 'mood compensation dial' (MCD) otherwise technically known as the EV compensation setting.... If you want a brighter mood- +EV if you want a darker mood -EV:rolleyes:

[JimKasson]

There seem to be several assumptions floating around in this thread, all of which I disagree with.

 

The best raw files look the best with no raw converter fiddling.

 

ETTR files, which produce the highest quality images, often look terrible -- too bright, washed out -- before raw converter adjustments are made.

 

The light meter is the primary exposure tool in the M9 and M240.

 

I think the histogram is the best tool in general. It will tell you if your highlights are blown, and the meter will not. If you've set up UniWB, it will do that with great accuracy. You can say that it's not as convenient as it should be in the M9, since you have to take a test picture to see it, but in the M240, you can see a live histogram in the lower right corner of the finder.

 

The mood of the image is determined at the moment of exposure.

 

I suppose this is a corollary of number one, but a file that captures all the important information in the scene and does that with the best possible signal-to-noise ratio will allow the photographer to get the best image that fits her conception of the mood by manipulating the raw file and its descendants, without regard to what the raw file looks like with no adjustments.

 

There are also people who have commented on this thread and look at things more the way I do. I do think it is useful to deal with these underlying assumptions directly.

 

Jim

[mark_j]

Jim

 

Agree 100%. I think there is a lot of confusion between exposure (fixed at the press of the shutter) and brightness (exposure as adjusted by gain, positive or negative, either in camera or in the raw converter).

 

Get the optimum exposure for the DNG files first then worry about mood (brightness) later.

[AlanG]

...What I stated at the begining was the light meter was a mood killer, and it was, word.

I simply asked whether this was the case with all build in meters in cameras, and this was where the story should have ended.

 

 

Pretty much true for all meters. (Although advanced systems sometimes try to make educated guesses at this.) Creating or preserving a mood via lighting and exposure is not what any kind of meter is designed for. After all, sometimes you might like the mood to be brighter and sometimes you might want the mood to be darker. Some of the exposure "scene" modes in many cameras try to do this providing you select the appropriate setting.

 

This type of photography requires human interpretation. EVF or LCD live view or chimping is a good solution that I often rely on. (Especially exacting is to shoot tethered.) Your other alternative is to shoot raw and hope you have enough headroom in the files to be able to adjust to get the mood you want when converting. When I shot film I often bracketed exposure on static subjects in order to have more chances of getting the mood I wanted. I still bracket with digital pretty often to be totally covered. And it costs nothing to do this.

 

I'd also put forth that your own visual system does not do a very good job of objectively registering the mood since it has an ability to automatically adjust to a contrasty scene or a flat scene much differently than the camera can.

 

As others have pointed out, the initial capture is just the starting point. I try to think of my captured files as I thought of a negative... they often need a lot of work to create the final product.

[Advertisement]

Hallo,

Look here: Leica M

[Your Old Dog]

I guess some of you guys missed my point. There's no need for a light meter to know what I think, I never asked for it. I'm aware it can be tricky for the meter in case of extra light sources, but the elements in the samples I took are more or less evenly lit from the light comming through the window to the right.

 

I simply stated the camera should expose properly i.e. the way scene really looks like, the real light amount in the scene.

 

Try using an "incident" light reading which is the light falling on the subject "from the camera position/direction" The Spectra Combi 500 is/was an excellent example of the "incident" light reader. The older Sekonic L28 was also a great inexpensive meter. Your Leica is giving you reflected readings.

 

I have read some articles elsewhere where people said it would be nice if camera manufactures would re-examine how light readings are taken for digital versus film cameras. The point was they are staying on an old path when a better path may exist.

 

An incident meter is normally placed at the subject but pointed back to the camera position. If outdoors in even lighting the readings can be taken at the camera position if you poiint the meter over your shoulder and not at the subject"