Jump to content

The Leica Forum uses cookies. Read the privacy statement for more info. To remove this message, please click the button to the right:    OK, understood.

Photo
- - - - -

M9 metering pattern


  • Please log in to reply
16 replies to this topic

#1 Lindolfi

Lindolfi

    Erfahrener Benutzer

  • Members
  • 870 posts
  • City / Ort:Groningen

Posted 05 March 2012 - 18:34

Advertisement (Gone after free registration)
Since I was curious how the M9 metering pattern was distributed over the film plane, I decided to measure it with 216 white blocks against a black background, together filling the focal plane. The blocks were arranged in an 18x12 matrix. Each time the meter reading was noted. Also a fully white and a fully black surface was measured. It appeared that the fully white surface had the same reading as the sum of the separate readings of all blocks together. The data were slightly smoothed and plotted in the following figure:

Posted Image

You can see that the pattern of gray shutter elements of the M9 comes through slightly, but the total result is very much like that of the Leica M6.
The values are scaled to the maximum reading in the center (1.0), while 0.0 is the lowest readings I got from the fully black image. The value is relative light intensity measured, not stops!

The scaling of the axes of the figure are in mm relative to the center of the focal plane.

For the readings the Macro Elmar 90/4 at f/8 was used, which gives minimal vignetting.

Edited by Lindolfi, 05 March 2012 - 19:12.

  • turner, jaapv, pico and 16 others said thank you
Bert
PhotoBlog

#2 Ecar

Ecar

    Erfahrener Benutzer

  • Members
  • 1,535 posts

Posted 05 March 2012 - 18:49

Interesting... It is somewhat larger and rounder than I thought.
Thanks!

#3 ho_co

ho_co

    Sponsoring Member

  • Premium Member
  • 7,761 posts

Posted 05 March 2012 - 19:22

Bert--Very nicely done!

A request: Consider trying the same thing with a wide-angle or ultra-wide as well, just as a comparison.

In the M8 introduction, my class was told that the camera "becomes more center-weighted average" with longer focal lengths, and "becomes more like a spot meter" with shorter focal lengths.

Just curious whether the M9 shares this tendency to any great degree.

It's good to see someone with your technical accuracy taking on the meter pattern! :)

Edited by ho_co, 05 March 2012 - 19:25.

  • mby, k-hawinkler and Lindolfi said thank you to this
Best,

Howard

#4 chris_tribble

chris_tribble

    Erfahrener Benutzer

  • Premium Member
  • 3,721 posts
  • City / Ort:London / Kendal

Posted 05 March 2012 - 21:28

My experience with lenses from 18 through to 135 is that this is very much the case. Meter and compose has to central to M exposure practice - especially when the light is complex.

#5 Lindolfi

Lindolfi

    Erfahrener Benutzer

  • Members
  • 870 posts
  • City / Ort:Groningen

Posted 05 March 2012 - 23:14

In the M8 introduction, my class was told that the camera "becomes more center-weighted average" with longer focal lengths, and "becomes more like a spot meter" with shorter focal lengths.

Just curious whether the M9 shares this tendency to any great degree.


That is not the case in the M9: I've tested a 21/2.8 at f/5.6 and it produced a similar result as with the 90/4 at f/8.

So that's nice: you only have to visualize one area during measuring, independent of focal length.

It is obvious that in absolute measurement angle, the longer focal length covers a smaller angle in the outside world than a wide angle lens. So if you want to use your M9 as a 5 degree spotmeter, mount a 135mm lens and nothing shorter.

Edited by Lindolfi, 05 March 2012 - 23:19.

  • ho_co said thank you to this
Bert
PhotoBlog

#6 ho_co

ho_co

    Sponsoring Member

  • Premium Member
  • 7,761 posts

Posted 06 March 2012 - 03:36

Sounds as if Stefan Daniel was right when he said the M9's metering pattern was similar to that of the M6!

Sounds like a noticeable improvement from the M8 as well. The M8 introduction class specified that the spot effect was stronger with wide angles, and reduced with longer focal lengths: just the opposite of your description with the M9. (Speaking of metering pattern, not lens field of view.)

Thanks for your graphic effort, Bert. ;)

Seems to me that a chart like yours would be helpful either in the M9 manual or in the M9 FAQ here on the forum.

Edited by ho_co, 06 March 2012 - 04:03.

Best,

Howard

#7 sblitz

sblitz

    Erfahrener Benutzer

  • Premium Member
  • 3,090 posts
  • City / Ort:new york

Posted 06 March 2012 - 04:22

interesting how the sensitivity leaks to the bottom and not to the top but seems fairly even left to right. thanks for doing this, very interesting and helpful.

steve

#8 Lindolfi

Lindolfi

    Erfahrener Benutzer

  • Members
  • 870 posts
  • City / Ort:Groningen

Posted 06 March 2012 - 08:35

interesting how the sensitivity leaks to the bottom and not to the top but seems fairly even left to right.
steve


Imagine shining a cone-shaped beam of light on a wall at an angle upward. You will get a deformed ellipse that spreads out more above the center of the beam than below. Now if this cone-shaped beam is analogous to the cone-shaped sensitivity of the light meter in the bottom of the M9, it is understandable that if it is aimed at the middle of the focal plane, there is more sensitivity above the middle of the focal plane, than below it.
Since later on the image is rotated to be viewed by us, there is more sensitivity below the middle of the image, but it's extension is limited by the blackness of the top shutter element (bottom of the image).
  • pico said thank you to this
Bert
PhotoBlog

#9 lars_bergquist

lars_bergquist

    Erfahrener Benutzer

  • Members
  • 5,467 posts
  • City / Ort:Greater Stockholm

Posted 06 March 2012 - 08:55

A request: Consider trying the same thing with a wide-angle or ultra-wide as well, just as a comparison.

In the M8 introduction, my class was told that the camera "becomes more center-weighted average" with longer focal lengths, and "becomes more like a spot meter" with shorter focal lengths.

Just curious whether the M9 shares this tendency to any great degree.


People – even people who should know better – automatically think SLR though this is not applicable to any M camera. A M, like any well-regulated English household, has an Upstairs and a Downstairs, and only the most formal connections between them. And metering is done in the Servants' Quarters.

In his Leica M Compendium, Handbook of the Leica M system (Hove, 1994) Jonathan Eastland, who surely knows better when he stops to think, declares (p. 130) that

"The diameter of the measuring field is equal to 2/3 of the short side of the bright-line frame in use. – This is a useful piece of information because in practice it means that the photographer can accurately measure very small areas of the object to be photographed using the preselector lever. – For example, when using wide-angle lenses, the measuring field of the meter appears larger. [ ... ] By flicking the preselector lever to bring up the 90mm bright-line frame, particular areas of the scene can be measured and interpolated."

So the otherwise estimable Mr Eastland thinks that by changing the frames upstairs, he can change the metering angle downstairs, without changing lenses! Good try.

And I do suspect that this kind of thinking lies behind the statement that metering angles are different relative to the picture angle with wide and long lenses. Mr Eastland I presume cut his teeth on SLR cameras with TTL metering (he mentions Nikon). Today's amateur photographers, and future pros, cut their teeth on cell phones, holding their humonguous DSLR cameras with long zooms in front of them on outstretched arms. I wonder what other strange behaviours this will lead to.

Anyone who launches a DSLR that you can make phone calls from, has a winner.

LB

#10 jaapv

jaapv

    Moderator

  • Moderators
  • 38,921 posts
  • LocationBehind a Red Dot

Posted 06 March 2012 - 09:51

Thanks, Bert, I have added this to the M9 FAQ
  • Lindolfi and dwbell said thank you to this
Jaap

WWW.JAAPVPHOTOGRAPHY.EU

Posts in blue bold font are made as a moderator

Leica user since 1976

#11 Lindolfi

Lindolfi

    Erfahrener Benutzer

  • Members
  • 870 posts
  • City / Ort:Groningen

Posted 06 March 2012 - 16:06

Indeed Lars, Jonathan Eastland was wrong there. The manual of the Leica M6 does show the measurement area correctly as scaling with the angle of view of the lens mounted and being constant in size relative to the correct framelines (not the one chosen with the preview lever), and thus keeping the coverage on the filmplane ("downstairs") the same.

I have superimposed the circle of the M6 from the manual onto my graph:

Posted Image

You can see the strong resemblance between the meters of the M6 and M9 as Stephan Daniel already pointed out.
My graph can easily be viewed in stops also: 0.5 is 1 stop lost, 0.25 two stops and 0.1 slightly more than three stops lost in sensitivity.

Edited by Lindolfi, 06 March 2012 - 16:09.

Bert
PhotoBlog

#12 Pix_N

Pix_N

    Neuer Benutzer

  • Members
  • Pip
  • 3 posts
  • City / Ort:Paris

Posted 09 March 2012 - 01:19

Correct me if i'm wrong, or if i missed something, but i think there is a misunderstanding of Eastland assertion.
He says "With a wide angle mounted", typically a 35 mm, you can preview the most active metering area for this lens by flicking the lever to show the 90 mm frame.
With any other lens, the metering area will vary in the same proportion : 2/3 of the height of the frame : so approximately for a 50 mm that would give an equivalent of a 135, and for a 28 something like a 50 mm frame.
But without touching any lever it's easy to imagine a center active metering spot of the correct size in the wiewfinder according to the lens mounted.

Anyhow you made a great job Indolfi

#13 jaapv

jaapv

    Moderator

  • Moderators
  • 38,921 posts
  • LocationBehind a Red Dot

Posted 09 March 2012 - 01:30

Correct me if i'm wrong, or if i missed something, but i think there is a misunderstanding of Eastland assertion.
He says "With a wide angle mounted", typically a 35 mm, you can preview the most active metering area for this lens by flicking the lever to show the 90 mm frame.
With any other lens, the metering area will vary in the same proportion : 2/3 of the height of the frame : so approximately for a 50 mm that would give an equivalent of a 135, and for a 28 something like a 50 mm frame.
But without touching any lever it's easy to imagine a center active metering spot of the correct size in the wiewfinder according to the lens mounted.

Anyhow you made a great job Indolfi

See post #9
Jaap

WWW.JAAPVPHOTOGRAPHY.EU

Posts in blue bold font are made as a moderator

Leica user since 1976

#14 TBL

TBL

    Benutzer

  • Members
  • PipPipPip
  • 41 posts

Posted 09 March 2012 - 01:39

Thanks Jaap.

It is however confirming my objection to the system: if you shoot in portrait mode, it will give different results if you turn the camera 90 degrees to the left or 90 degrees to the right...

#15 jaapv

jaapv

    Moderator

  • Moderators
  • 38,921 posts
  • LocationBehind a Red Dot

Posted 09 March 2012 - 02:50

Looking at the pattern I seriously doubt that difference would be perceptable in real life.
Jaap

WWW.JAAPVPHOTOGRAPHY.EU

Posts in blue bold font are made as a moderator

Leica user since 1976

#16 Lindolfi

Lindolfi

    Erfahrener Benutzer

  • Members
  • 870 posts
  • City / Ort:Groningen

Posted 09 March 2012 - 16:53

Thanks Jaap.

It is however confirming my objection to the system: if you shoot in portrait mode, it will give different results if you turn the camera 90 degrees to the left or 90 degrees to the right...


Just did a test: filling the left half of the image with black and the right half with white in portrait orientation and taking the same image after turning the camera 180 degrees, so that now black is right and white is left. (care was taken that the boundary between black and white was exactly through the middle of the image).

The black and white had a difference of 7 stops in brightness.

Result: same measurement of exposure time in the 1/3 stop accuracy the M9 offers in both portrait orientations.

So no reason to be consistent in how you turn the camera to portrait orientation: just use the meter as though it is rotation symmetric around the centre of the image.
  • jaapv said thank you to this
Bert
PhotoBlog

#17 zlatkob

zlatkob

    Erfahrener Benutzer

  • Members
  • 1,963 posts
  • City / Ort:New Jersey, USA

Posted 09 March 2012 - 18:33

Leica should publish the metering pattern.
  • M9reno said thank you to this




0 user(s) are reading this topic