M7 light metering

[michel_vandeput]

Dear LEICA friends,

On Erwin Put's LEICA M7 test report, about the light metering, you can read:

 

<p>

 

"The measuring spot is often described as (semi) spotmeter. It is however best described as a center-weighted integral metering pattern."

 

<p>

 

Until now, I always used, with success my old M6 light metering system as a full spot metering. I can't understand why, with the same system, we have a center weighted integral metering on M7. Does it mean that the light is measured on the entire surface of the photo and not only on the white dot?

Thanks and best regards from Belgium.

Michel

[andy_piper2]

The metering area is the same for the M6 and M7 - no change in the dot

size or meter pattern.

 

<p>

 

True hand-held 'spot' meters cover an area of from 1 degree to 5

degrees.

 

<p>

 

With a 50mm lens mounted the M6/7 meter 'sees' about 20 degrees. With a

135 it sees about 7 degrees, and with a 21 lens it sees about 50

degrees. So it's never been what I'd call a 'spot' meter.

 

<p>

 

That being said, I think Erwin goes a bit far in calling the M meter

'integral'. No doubt the 'black' part of the shutter curtian reflects

some light to the meter, since few things are perfectly black. But the

light from the white dot is still probably close to 90% of the reading.

 

<p>

 

In a real "centerweighted integral" meter the edges of the picture

contribute far more to the total reading: e.g. in the Nikon F3 the

ratio center/edges is 70/30 while in the FM/FE/FM3A cameras the ration

is 60/40 - instead of 90/10.

 

<p>

 

I think Erwin was just trying to gloss over the fact that for auto-

exposure a real 'spot' meter is a liability - it can be too easily

fooled by small areas of light and dark - and since the M7 has

autoexposure it better have a 'centerweighted' meter.

 

<p>

 

In any event I like the M6 pattern myself - I don't actually have a

metered M body, but the SL also has a 'wide spot' meter - which is

great for working fast but also keeping some control over what light

the meter sees - and not being influenced too much by dark/light

backgrounds.

[glenn_travis]

The M6 "Classic" meter covers approximately 23% of the picture area.

The M6 ttl/M7 covers approximately 13% of the picture area. This is

constant, and does not change with focal lenght of a lens. This

would, at best, be considered to be a "Partial Area Meter," or a 100%

center weighted meter, since it's close to the area that most center

weightered meters read. Having used a Pentax Digital Spot meter (1

degree) for a number of years, under no circumstance should the Leica-

M meter be considered a "Spot" meter. Having said that, it's still

one hell'va meter.

[erwin_puts]

The basic principle of the exposure meter is the measurement of the

light intensity of the measured area. As there is only one

measurement, the intensity of the illuminance is by definition an

average value (or an integrated one, which is saying the same). If

the area to be measured has one uniform brightness, the measurement

is simple. If the area to be measured consists of various different

brighness values (some black, some grey, some white), the meter will

produce one and only one value for the total brightness. This value

of course is a weighted average of the various brightness areas.

The area to be measured depends on the receptive angle of the meter,

a spotmeter having one or three or five degrees (depending on

definition) and the classical acceptance angle is 27 degrees. Even a

spotmeter will act as an averaging meter when the object that is

being targeted by the spot consists of patches of varying brightness

levels. The idea of a spotmeter is that you can select a

representative area of uniform brightness as a basis for the

determination of your exposure calculation. If I am very close to an

object and can use my normal wide angled exposure meter to meter an

area of uniform brightness, I am in effect doing a spotmetering.

To be specific now for the M-exposure metering. If I use my 135mm and

focus on a subject that fills the frame outlines and this object

conssts of a large rnage of brightness values, the M-meter will do an

averaging reading of a selective part of the object. This is not a

spotmetering technique.

If I use the 24mm lens and I am so close to the subject that I can

select an area of uniform brightness that fills the wide angle view,

I have in effect a spotmeter reading.

When I noted that the M-meter should be interpreted as an integral

meter, I was referring to these aspects.

Any meter will make a weighted average of the various brightness

values that are falling on its sensor.

What area a meter will measure depends on acceptance angle and

distance from meter to subject.

What the reading is, depends on the range of brightness levels in the

measured area.

The typical use of the M-camera does indicate that the 13mm patch of

the exposure meter will act as an integral meter.

Maybe the confusion is in the reference to the integral meter or

center-weighed metering methods as implemented in Slr�s where you

camn vary the sensitivity of the meter to different parts of the

screen. In a typical classical slr the screen is the substitute for

the area to be measured and when (as in the Nikon 60-40 or 75-25

patterns) the metering cells are calibrated such that they are more

sensitive to a central part of the measured area, the weighting of

the brightness areas may be different, but it still acts as an

averaging meter.

 

<p>

 

Erwin

[anon_terry]

Umm... what does the instruction booklet say?

[paul_chefurka1]

Glenn, the spots on the M6 Classic, M6 TTL and M7 are all the same

size. This means that they all meter the same proportion of the

frame.

 

<p>

 

Erwin, thanks for explaining your terminology. I must admit I'd

wondered about your definitions as well.

[glenn_travis]

Sorry if I'm incorrect Paul, but this information was taken from

Leica's Literature. First the instruction Manual for my Leitz M6, and

then from the Leica 2001 Catalog

[nigel_bowley]

Yes, the meter spots on the shutter curtains of M6, M6TTL, and M7 are

all 12mm diameter. I've seen quoted in various books and forums

(fora ?) that this equates to 13% of the image area, or 23%, or 13 -

23%. What's this mean?

 

<p>

 

The meter spot IS 12mm diameter. The image size IS 24mm × 36mm.

This is 13.09%. It's always 13.09%, no matter what lens you fit

(well, not the Super Angulons etc which block the metering cell!),

and no matter what distance the lens is focussed to.

 

<p>

 

BUT, if the lens is focussed to infinity, the framelines show

significantly less than the image which will be captured on film.

The metered area is more than 13% OF THE AREA SHOWN BY THE FRAMELINES.

So, the metered area is always about 13% of the image which will be

captured on film. It varies between 13% and 23% of the area shown by

the framelines, depending on the distance which the lens is focussed

to.

 

<p>

 

BTW, according to my calculations, the meter angle is 31.89° for a

21mm lens, 13.69° for a 50mm, and 5.09° for a 135mm. So with a 135mm

lens the meter angle approaches the classical definition of a

spotmeter. If you mount a 135mm temporarily, you can take a 'spot'

reading for that shot you want to take with a 21mm. The only thing

is, unless you're carrying a 135mm anyway, you may find it more

convenient to carry an actual spot meter!

 

<p>

 

Hope this helps.

 

<p>

 

Nigel

[eliot]

I believe the correct number IS 13% of the image area, the 23% refers

to the percentage of the area inscribed by the framelines when the

lens is focussed at infinity (which does not contain the entire on

film image). The 13-23% may reflect the fact that when the lens is

focussed closer and closer, the focal length actually increases a

little, but the framelines are not corrected for this effect (only

for parallax(, so the % coverage changes.

[nigel_bowley]

I don't think that the focal length changes, but the angle of view

does change.

 

<p>

 

Nigel

[eliot]

Nigel. As I see it, the focal length in Leica M lenses would have to

increase slightly as the lens is focussed closer and closer (thus

effectively reducing the angle of view by a small amount), since the

optical center moves further from the film plane. The only way this

would not happen is with some kind of internal focussing mechanism,

which M lenses don't have.

 

<p>

 

Unless you define focal length as that distance when the lens is

focussed at infinity, it seems that the focal length would have to

increase, albeit not by very much, as the lens is close-focussed.

[nigel_bowley]

Eliot

 

<p>

 

Yes, I was thinking that the focal length is the lens to image

distance when the lens is focussed at infinity. That's right isn't

it?

 

<p>

 

Things seem to be made a bit more complicated by the fact that there

are several elements to a lens. But these elements don't move

relative to each other (unless it's a zoom lens, when the focal

length DOES change, of course).

 

<p>

 

It makes it simpler to think about if you consider the elements of a

lens to be equivalent to a simple single element lens. Such a lens

would have a focal length solely determined by it's curvature. An

object at infinity would be brought to a focus at a distance from the

lens equal to the focal length. A closer object would focus further

from the lens.

 

<p>

 

The formula which governs this is 1/f = 1/u + 1/v, where f = focal

length, u = distance of image from lens, v = distance of object from

lens.

 

<p>

 

For an object at infinity with a 50mm lens: 1/50 = 1/u + 1/infinity,

so u = 50.

 

<p>

 

For an object at 1m: 1/50 = 1/u + 1/1000, so 1/u = 0.02 - 0.001, u =

52.63. The distance of the lens from the film increases by 2.63mm

when focussing from infinity down to 1m.

 

<p>

 

The distance to focus down to the near focus limit of 0.7m is

3.85mm. We can repeat these calculations for each focal length. A

135 lens moves by 13.35mm, but a 21 by only 0.65mm!

 

<p>

 

Now we know the image distance at each end of the focussing range we

can calculate the changes in angle of view and meter coverage. For

example: a 135 lens has an angle of view (diagonally across the film

plane) of 16.59° (2 x arctan (21.63/148.35) at the closest focussing

distance of 1.5m, and 18.21° at infinity. The meter angle is 4.63°

at 1.5m and 5.09° at infinity.

 

<p>

 

The metered area is 52.17% of the height of the framelines and 14.05%

of the area at 1.5m, and 57.33% of the height and 16.97% of the area

at infinity. It's always 1/2 the height and 13.09% of the area of

the film format.

 

<p>

 

The differences are much less with the wider lenses, and are least

with the 21.

 

<p>

 

I don't know where the figure of 23% comes from. It is the

proportion of the frameline area covered by a metered area which is

2/3 the height of the frameline, but for none of the lenses is the

figure quite this high. It is greatest with the 75mm, where the

metered area (at infinity) is 57.97% of the frameline height, and

17.34% of the area.

 

<p>

 

It is interesting to note that with the 75 (and the 90 and 135 aren't

much better) at infinity the framelines barely show 75% (by area) of

what will be recorded on film!

 

<p>

 

Nigel