8
\$\begingroup\$

Please refer to this photo:

http://img600.imageshack.us/img600/94/1600full.jpg

The image on the left was taken with a Panasonic GH2, a micro-4/3rds camera with a crop factor of 2, and the right one was taken with a Nikon D700, a full-frame camera. Both of the cameras used the same aperture, shutter speed, and ISO, and yet the D700 photo came out about one stop brighter.

My question is, why did it come out so drastically different, given the same settings? I've asked the same questions several months back theoretically:

Do the same camera settings lead to the same exposure across different sensor sizes?

But recently I was able to test this with the two cameras. My guess is that both cameras measure ISO differently from each other, but isn't ISO a standard that's shared across camera systems, much like how shutter speed and aperture are? Any insight into the matter would be greatly appreciated.

Improve this question
\$\endgroup\$
2
  • 2
    \$\begingroup\$ "The great thing about standards is that there's so many of 'em!" \$\endgroup\$
    – mattdm
    May 26, 2011 at 13:37
  • 3
    \$\begingroup\$ ImageShack seems to have deleted your image, and replaced it with an ad banner. If you can, please reupload the image (or something equivalent) using the image upload button in the editor toolbar (which will upload it to Stack Exchange's imgur account). \$\endgroup\$
    – Ilmari Karonen
    Aug 17, 2015 at 17:23

4 Answers 4

Reset to default
6
\$\begingroup\$

You're right to be skeptical of ISO sensitivity. Yes it is a standardized, but the standard effectively allows for manufacturers to specify their own sensitivity. For a long time ISO320 on a Nikon was equivalent to about ISO400 on a Canon.

But that difference, whilst noticeable, was only a third of a stop. I wouldn't put a whole stop down to ISO differences. It would be silly for a manufacturer to label their camera as ISO3200 when it's the same sensitivity as another manufactuer's ISO1600, that way it will just make their camera look more noisy, as people compare cameras at the same ISO number, not the same actual sensitivity!

I think it's more likely to be a difference in tonecurves applied during raw processing that makes one image seem darker.

Image processing is the weak link, where it can all break down. Shutter speed is measured in seconds, a one second exposure on one camera is the same duration as a one second exposure on another (ignoring relativistic effects!). Ok there are tolerances so the duration wont be exactly the same, but the point is there is a unit, which is well defined. Same with aperture. But there is no unit of brightness in an image.

ISO [badly] defines how much actual light (a proper measurable quantity) it takes to saturate a sensor. But it's the image processing software that translates that into the numbers that get displayed on your screen. Even if the software maps a fully saturated sensor to the value 255 (which would seem to be the only sensible thing for it to do), a half saturated value might also get mapped to 255 or any other value for that matter. And that mapping can dramatically affect the perceived brightness for an image.

Improve this answer
\$\endgroup\$
5
  • \$\begingroup\$ Does that mean that if the experiment were repeated but with raw files and identical raw development settings were used you would expect to see a closer match? \$\endgroup\$
    – fmark
    May 26, 2011 at 13:46
  • \$\begingroup\$ Yes, shooting raw with linear tonecurves ought to get a lot closer, the only disparity would then be differences in ISO sensitivity (which I would expect to be within a third of a stop) and any nonlinearity in the sensor response. \$\endgroup\$
    – Matt Grum
    May 26, 2011 at 14:01
  • 4
    \$\begingroup\$ By the way, a half-saturated sensor is usually also mapped to 255 in raw processing (with the default settings, assuming a typical DSLR). That's exactly the reason why you can recover the highlights in an overexposed photo by adjusting the settings in raw processing: the sensor was not really saturated, it only looked like overexposed because the mapping between sensor readings and the final image clipped the highlights. \$\endgroup\$
    – Jukka Suomela
    May 26, 2011 at 14:19
  • \$\begingroup\$ @Jukka that's right, the point is the camera can map intermediate values to just about anything. I've edited the answer to be more realistic, however. \$\endgroup\$
    – Matt Grum
    May 26, 2011 at 15:01
  • 1
    \$\begingroup\$ Both were shot in RAW and converted to DNG in Lightroom. \$\endgroup\$
    – Daniel T.
    May 26, 2011 at 19:19
6
\$\begingroup\$

In addition to differences in tone curves and sensor ISO sensitivity levels, there are more factors that could be affecting exposure with same settings.

  • The cameras have different optics attached. That means different number of elements, materials, coatings, all of which have effect on transparency of the lens.

  • Aperture diaphragms are not precision devices, even the same lens may close down differently each time.

  • Artificial light varies during AC power cycle (more noticeable with higher shutter speeds, should even out by 1/20s as in question).

  • Filters in front of lens (with polarizer being the easiest to forget).

  • Some cameras (such as Sony α55 and α33) have translucent fixed mirror that constantly tunnels part of the light into viewfinder and AF sensors (not the case for any of the cameras mentioned in question though).

When shooting JPEG, the following might be classified under different tone curve, or not:

  • Contrast setting.

  • Dynamic range expansion.

Improve this answer
\$\endgroup\$
4
  • 1
    \$\begingroup\$ +1. Also, shutter speeds can vary from the nominal. \$\endgroup\$
    – mattdm
    May 26, 2011 at 14:48
  • 1
    \$\begingroup\$ Another issue to keep in mind is that you will lose some light when you are focusing very close, and exactly how much you lose depends on many factors (see, e.g., this thread). \$\endgroup\$
    – Jukka Suomela
    May 26, 2011 at 23:04
  • \$\begingroup\$ shutter speed, aperture and transmission are going to vary by a few percent, if the aperture closing varies by a stop you have huge problems. Variations in lighting due to the AC cycle is far more likely. \$\endgroup\$
    – Matt Grum
    May 26, 2011 at 23:07
  • \$\begingroup\$ I don't think the aperture closing down differently each time varies enough to give it a 1-stop difference. I was not using filters or a camera with a pellicle mirror, and I was shooting in RAW, so the only difference one left is probably the different optics or different ISO measurements. \$\endgroup\$
    – Daniel T.
    May 27, 2011 at 2:33
0
\$\begingroup\$

If you took your 4/3rds lens and mounted it on a FF sensor it would give you a darker image when set at the same F/number. The lens is smaller.

The optical equation: "f-number = Focal length / Diameter" describes the controlling physics. In a 4/3rds lens, the focal length is 2X that of a Full Frame. In the equation, if we double the focal length, the f-number is also doubled if the diameter is kept constant.

So, we must double both the focal length and the f-number in a 4/3rds camera lens to compare it to a full frame lens.

Retake your images but close down the full frame f-number to be 2X the 4/3rds f-number and your exposures will be similar.

When manufacturers report the focal length equivalent (crop factor) in a 4/3rds lens they, strategically, neglect the f-number adjustment.

Improve this answer
\$\endgroup\$
2
  • 3
    \$\begingroup\$ The focal length is always the same for a given lens at a given setting. Does my crop sensor camera actually turn my lenses into a longer focal length? \$\endgroup\$
    – user
    Jan 14, 2017 at 20:41
  • \$\begingroup\$ Retake your images but close down the full frame f-number to be 2X the 4/3rds f-number and your exposures will be similar => nope. As a FF sensor is larger, it requires more light to get the same amount of light per unit area (= same exposure). \$\endgroup\$
    – Olivier
    Jan 15, 2017 at 12:51
-3
\$\begingroup\$

While there are a lot of good answers here, I noticed one major difference, which is that the Micro 4/3rds shot is taken at 20mm, while the FF is at 50mm. Since aperture is a ratio of the aperture diameter to the focal length, this means that the opening for ƒ2 on the 4/3rds setup was 10mm wide, while the opening for ƒ2 at 50mm full-frame was 25mm. In absolute terms, a larger amount of light was admitted inside the body of your D700.

I don't know for certain that this is what causes the difference, but it was the first thing that jumped out at me. I'd retake the shots so they are framed similarly at identical focal lengths and see if that made a difference.

Improve this answer
\$\endgroup\$
2
  • 2
    \$\begingroup\$ The D700 with a larger opening does let in more light, but it also has a larger sensor so light per unit area is the same so you would expect the same exposure. \$\endgroup\$
    – Matt Grum
    May 26, 2011 at 23:09
  • \$\begingroup\$ The micro 4/3rds has a crop factor of 2. 20mm x 2 = 40mm, which is pretty close to 50mm. And Matt Grum's response is correct as well. \$\endgroup\$
    – Daniel T.
    May 27, 2011 at 2:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.