M9 shutter speed off by 1/3rd

[photohc]

Advertisement (gone after registration)

I recently started to incorporate the use of a light meter (Sekonic L-358) into my shooting. My M9 is now 2 mos old and with the use of the light meter, I now notice the exposure off, from the light meter reading by about 1/3rd. Although I can compensate by adjusting the light meter to now increase the reading by 1/3rd to result in a proper exposure, I'm thinking it might be better to send the M9 in for recalibration. So my question is, is this just a variable of the shutter speed and expected or should I it to be spot on and therefore send in for adjustment? I checked my meter against my D700, and the D700 is spot on, so I'm very confident the meter is correct.

[marknorton]

I'm not sure (have no idea actually) why you think that just because the M9 and Sekonic are coming up with different exposures that it must be the M9 shutter speeds which are at fault.

 

How do you know that the two meters are looking at the same thing and have the same scene weighting? How do you know that Leica's ISO 160 is the same as Sekonics? How do you know that Sekonic's thinking as to what the correct exposure is is the same as Leica's?

[wparsonsgisnet]

Ditto. What are the chimp images showing you? You can actually see what the camera is seeing, when you can't evaluate the meter.

 

Have you ever tried to synchronize two watches?! I don't know if I can see a 1/3 stop effect in my images.

[jaapv]

Digital ISO is different from "real"ISO with Leica. 160=200, 640=400, 1250=1600, etc...The same with the M8. You must set your meter accordingly. (and manual flash!)

[andyb]

Digital ISO is different from "real"ISO with Leica. 160=200, 640=400, 1250=1600, etc...The same with the M8. You must set your meter accordingly. (and manual flash!)

Perhaps I missed something in your logic but when I double the ISO setting on my M9, the internal meter cuts the exposure time in half and that's how it should be. Following your logic, this shouldn't be true any more. In your numeric model, the exposure time should be cut in half if I switch from ISO 160 to ISO 640, but this simply doesn't seem correct to me and to my internal meter. My M9 seems to follow the more conventional relation between exposure time and ISO setting :-)

Please correct me if I got something wrong

 

Andreas

[wparsonsgisnet]

Andy, I think Jaap meant that the so-called ISO values in the M8 are too conservative.

 

Sean Reid says that in the M8, 160 is really equivalent to 200 on other digicams he uses, and so forth.

[andyb]

Advertisement (gone after registration)

Andy, I think Jaap meant that the so-called ISO values in the M8 are too conservative.

 

Sean Reid says that in the M8, 160 is really equivalent to 200 on other digicams he uses, and so forth.

 

OK, now I understand. I was thinking too much in terms of numbers

[photohc]

Ditto. What are the chimp images showing you? You can actually see what the camera is seeing, when you can't evaluate the meter.

 

Yes, I took both my M9 and D700 outside and set both to the same manual setting specified by meter, shooting the same image. Both set at the same ISO. After downloading to Lightroom, the M9 was clearly underexposed by 1/3-1/2 in exposure. The D700 was spot on. Not very difficult to compare. As a further check, I took a second light meter to compare against the Sekonic and both measured the same in all situations. This leads me to conclude that the shutter on the M9 is not firing at the speed specified on the dial (faster for underexposure). Having said that, if I place the Shutter speed to Auto, it appears to meter correctly in your average 18% scene.

[jaapv]

The point is that the ISO norm is formulated rather soft. It was adequate for film, but for digital it is open to interpretation. The methods used for determining the ISO speed of an emulsion are not adequate for use on a sensor. Leica follows the most rigourous version, which means that the Leica ISO 160 is equivalent to ISO 200 on your meter, your flash, the box of your roll of film, etc. So when you set the M9 to 160, you wil have to set your exposure meter to 200 etc.

[photohc]

I'm not sure I agree with that logic. The ISO is a standard for the industry and I would guess that Leica has measured their camera's against that standard. Otherwise why have one if you need to fudge. Besides, if the real ISO is 200, that would mean my M9 is even further off. If I left my M9 ISO at 160 and set the meter at 200, the meter will give me a shutter speed that is even faster, thus causing the M9 exposure to be further under exposed.

[jaapv]

Just read the ISO sheet...

[wparsonsgisnet]

Again, can anyone see a 1/3 stop difference?

 

Ansel, Ansel, where are you when we need you?

[scott kirkpatrick]

There was a mistype in Jaapv's original note. I don't think he meant to say that 640 on the M8 and M9 was equivalent to 400 on bog-standard Canon DSLR's, but that it was equivalent to 800, while 320 was equivalent to 400.

 

I think you have to learn how each camera interprets the appropriate exposure. After all most DSLRs have rather tricky matrix metering algorithms that try to place the whole histogram of light values reasonably, measuring the light at multiple places across the frame. The M8 and M9 take a center-weighted average reading, and add or subtract some thing to " protect the highlights." My sense has been that the M9 needs a little less highlight protection than does the M8, so I set the M8s to -0.3 exposure correction and my M9 to +0. But as a test, I just pointed both at the same light source with lenses at f/4, shutter on A and ISO 160, and got an exposure time of 1/3 second with each camera. So the exposure to set middle grey where it should be is apparently the same in each. (Sean Reid also thinks they need slightly different exposures, using real tungsten-lit scenes.)

 

scott

[zlatkob]

I'm not sure I agree with that logic. The ISO is a standard for the industry and I would guess that Leica has measured their camera's against that standard. Otherwise why have one if you need to fudge. Besides, if the real ISO is 200, that would mean my M9 is even further off. If I left my M9 ISO at 160 and set the meter at 200, the meter will give me a shutter speed that is even faster, thus causing the M9 exposure to be further under exposed.

 

It does appear that your camera is even further off. With all manual settings, the extra 1/3 stop sensitivity (160 really being 200) should cause the camera to overexpose by 1/3, but yours is underexposing by 1/3 to 1/2. However, it is strange that if fixes itself -- becomes accurate -- when you set the shutter to Auto.

 

Although ISO is a standard, I know that even among Canon cameras, some models are known to be slightly more sensitive than the standard while others are at the standard or slightly under.

[zlatkob]

Again, can anyone see a 1/3 stop difference?

Yes. It's not big, but it is big enough to be noticeable. Sometimes it is the difference between highlights being clipped or not.

[adan]

In film days, people shot Tri-X at 200, or at 400, or at 320, or at 500 - depending on how they wanted their negatives and prints to look, or how their meter measured light. The ISO/ASA/DIN rating was a guideline - but no one sent their Tri-X back to Kodak to be "recalibrated" if they found they preferred 320 or 500 to the nominal 400 speed printed on the box.

 

Instead they recalibrated their meter setting by the low-tech means of changing the ISO setting on the meter, relative to the ISO of the film or sensor. I've never assumed a meter would give me "correct" exposure out of the box - I've always shot tests to figure out which ISO to use on the meter, whether I was shooting Velvia or silicon.

 

What you've discovered is that you like 1/3rd stop more exposure with your meter and your Leica sensor than the defaults. In my case, I prefer 1/3rd LESS exposure than the meter normally gives.

 

Incidentally (pun intended) if you are using the L-358 with the incident-light dome - incident meters tend to favor highlights if the light is contrasty. It is just a function of how that 3D dome collects light relative to the flat surface of a gray card. Strong highlighting on one side of the dome will overpower the shaded area on the other side, giving a higher light reading (and thus recommending less exposure).

 

As to what the D700 does - sure, the Nikon and the M9 have different "films" - anyone can tell that just by looking at the noise levels and sharpness. You need to pick different ISO corrections for each film, just as though they were HP5 and Ektachrome 400, or Velvia and Pan F.

[photohc]

Andy, thanks for the reply. I think you have put it in prospective for me. The Sekonic has a calibration setting, allowing a +/- 1 EV on it. By applying a -1/3EV, I now have the Sekonic agreeing with my M9. My original concern was whether I should send in the M9 for recalibration. It now sounds like I just need to keep the calibration adjustment in place and use it that way. Harv

[photohc]

Yes. It's not big, but it is big enough to be noticeable. Sometimes it is the difference between highlights being clipped or not.

 

Agree. It is even more evident when viewing on a large screen. It might be that I'm more than 1/3EV off, maybe 1/2EV. Has anyone else tried to make this same comparison I have done?

[GlennB]

To add to what Andy has written; I remember in the film days (yes I am that old) I remember reading or hearing (I could hear better then ) that leica's meters (read here M6, M6ttl, I don't think the old MR meters were that accurate) were designed for slide film (Kodachrome ?) which would have better color saturation when underexposed slightly (about 1/4 to 1/3 stop). Other camera makes had their own unique "features" such as Nikon's viewfinder showing less than 100% (98% ??) so that when a slide was projected, your image sides (what you were expecting to see) wouldn't be cut off by the cardboard mount's cutting into the image.

 

Each manufacturer seemed to have their own unique philosophy when it came to these sort of things. Seems like they still do !

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

The relevant ISO norm is 12232. The only part in that text relevant to digital cameras is this:

 

The methods for assigning an ISO speed rating to electronic cameras should harmonize with current photographic standards and practice. In order to be easily understood by photographers, the ISO speed rating for an electronic camera should directly relate to the ISO speed rating for photographic film cameras. For example, if an electronic camera has an ISO speed rating of ISO 100, then the same exposure time and aperture should be appropriate for ISO 100 rated film / process system. (Final Draft International Standard ISO 12232, October 1997)

 

That " norm" is so soft that it leaves the interpretation open to the camera maker with plenty of leeway, as the word "appropriate" is not defined in any way. For instance, a sensor cuts of highlights abruptly, film has recoverable highlights. At which part of the exposure curve of film should we place the cutoff of a sensor? The same for the shadows. Or should we match at 18% grey? The norm does not say. So we should not be surprised at the discrepancies we see in practice. I am sorry I cannot quote the complete norm. It costs 92 CHF to buy and is then locked for reproduction by full copyright.