Exposure differences between Panasonic and Olympus Lenses????

Discussion in 'Native Lenses' started by Swandy, Jun 7, 2016.

  1. Swandy

    Swandy Mu-43 Veteran

    Dec 15, 2009
    Has anyone noticed a difference in exposure between Panasonic and Olympus lenses? I have the Pen-F which really - at least for me - screams for small primes and pancakes. I had the Olympus 14-42EZ pancake and got the Panasonic 12-32 Pancake to compare. And while doing some VERY unscientific comparisons (trying to determine which lens I preferred) I noticed that very often the Pen-F will (depending on how I look at the numbers) either overexpose the shot from the Panasonic or underexpose the one from the Olympus pancake. (Now, to be honest I was shooting Aperture Priority with the aperture set at 5.6, but the camera was on Auto ISO - so not a real scientific test at all.) But there were many shots - especially the ones indoors - where the camera will either push up the ISO on the Panasonic (in one shot - sorry already deleted the images - the Aperture, Focal Length and shutter speed were the identical - but the ISO on the Olympus was 500 and the Panasonic was 2000 - same basic framing and same amount of light in the room). The outdoor shots, the ISO was generally set - it was a pretty sunny day - at 200, but the shutter speed would shift a bit - again slower on the Pana than the Oly lens.

    I intend to do some better testing - using manual mode and not Auto ISO - but curious if anyone else has noticed this? Thanks.
  2. wjiang

    wjiang Mu-43 Hall of Famer

    It's not an Olympus/Panasonic thing, it's a lens thing.

    The camera will attempt to meter wide open, then close down for the actual exposure. It will subtract the f/stop (aperture) difference from wide open to the desired f/stop from the metered exposure

    Why does this lead to different results? T/stops are not f/stops. F/stops don't actually measure the amount light, T/stops do. If the lens vignettes wide open, or somehow changes its transmittance as you stop down, a 1 stop difference in f/stops (aperture) won't necessarily give a 1 stop difference in t/stops (transmission).

    Different lenses will have different characteristics in this respect. Because such aspects are not precisely characterised and corrected for in the m4/3 standard, YMMV.

    The only surefire way is to meter with DoF preview turned on so the camera gets the actual transmission characteristics at the desired aperture.
    • Winner Winner x 2
  3. David A

    David A Mu-43 All-Pro

    Sep 30, 2011
    Brisbane, Australia
    Haven't noticed but I do think you need to do more testing.

    1- As wjiang said above, T/stops are what counts and the reason for that is that f/stops are based on the ratio of the aperture opening to the focal length of the lens and don't take into account the loss of light as it passes through the elements of the lens. Different lenses have different element structures, often different thicknesses to the elements and different sorts of glass. The loss of light as it passes through the lens can vary from one lens to another by enough to make an exposure difference.

    2- you don't want the camera changing more than one thing in each test so if you're going to use aperture priority, turn auto ISO off. There should be no need to have auto ISO on for comparisons of this sort. All you want is the camera to change one parameter, the shutter speed. That makes your exposure comparisons a lot easier and more accurate.

    3- you don't want the light to change as you change lenses and that means more than just no changes to the light illuminating the scene. You need to compare the 2 lenses shooting from the same spot at the same thing so use a tripod. You need to use exactly the same focal length with each lens since you're using zooms, and you need to use exactly the same metering method and measure exactly the same area of the scene. The reasons for those other points is that the sensor measures the light falling on it. If you change focal length you gather light from different fields of view which may have different proportions of light and dark areas within the frame and that can change the result if you're using averaging or centre-weighted metering. If you're using spot metering, different fields of view will result in the spot, which is a fixed area of the sensor, measuring different areas of the scene with each lens and that can change the result also. The light meter must measure exactly the same area of the scene with exactly the same mix of light and dark areas for the comparison to be reliable.

    4- a good comparison to try is to hang a white sheet or something of a uniform colour on the wall, shoot from close enough so that the only thing in the frame is the sheet, and shoot with the same focal length and camera position with both lenses. This way you really minimise any possibility of a different amount of light being reflected to the camera with each shot. If there's a difference between the 2 frames, then there's a difference in transmission. I'd also shoot RAW and do the comparison in my processing software to avoid the possibility that the camera's JPEG conversion software has been tuned to make small adjustments as it does the conversion.

    As an alternative approach to redoing the test you did taking into account the things above, and I'd still do that anyway, what I'd also suggest you do is this. Set the camera to manual and measure the scene. Ideally it would be an interior scene with controlled lighting that you know will not change (it's hard to guarantee the sun because of cloud and atmospheric effects). If you have access to a hand held incident light meter I'd measure the scene using that and use the recommended exposure for my shots with both lenses with the camera in manual mode. If you don't have access to an incident light meter, do the usual reflected light measurement with the camera's meter using the first lens you shoot a frame with, but still use manual mode. Change lenses without changing the exposure at all and take a shot with the second lens using exactly the same focal length and camera position (tripod again if you have one. Then compare the 2 shots in your processing application and see how much you have to change the exposure of one of the shots in order to match the exposure on the other. Do no other processing other than moving the application's exposure slider for this comparison. If your software's exposure slider's scale matches the way exposure adjustments on a camera work, such as Lightroom/Adobe Camera Raw where a shift from the 0 neutral point to + or - 1 equals a difference of 1 stop, you'll end up with a much more accurate figure for the difference in transmission between the 2 lenses than you will by letting the camera adjust exposure since most cameras can only change exposure in third or half stop adjustments.
  4. Klorenzo

    Klorenzo Mu-43 All-Pro

    Mar 10, 2014
    Two stops of difference are a lot. In modern lenses I've never seen more then half stop of difference due to transmission. Metering on slightly different framing could explain 2 stops of difference, especially if the sky or bright elements are involved where just tilting up the camera a few degrees makes a lot of difference.

    Here there are a couple of other suggestions to test transmission:

    Is that Pentax Auto 110 lens REALLY f/2.8?
  5. ijm5012

    ijm5012 Mu-43 Hall of Famer

    Oct 2, 2013
    Pittsburgh, PA
    Real Name:
    As others have said, it's the T-stop value of a lens that actually dictates the exposure parameters of an image, not the f-stop.

    As for your statement about there being a 2 stop difference between the lenses, I would highly suggest doing a far more controlled test to determine any differences there may be between the lenses. Something with a fixed subject and constant light source. From there, set your ISO and Av, and let shutter speed float so that you can evaluate the differences in exposure using the same metering modes. I would suggest doing this with both "averaged" metering as well as "spot" metering, to see what the difference is.

    Based on your description of the "testing" done, no conclusive evidence can be drawn at this point.
  6. Conrad

    Conrad Mu-43 Veteran

    Were the indoor shots with fluorescent light? This matters because fluorescent lights typically flicker at 100 Hz (Europe) or 120 Hz (US). If your shutter speed is not significantly longer than 1/100 s, you will get random exposure (and likely also metering). A large number of energy saving "bulbs"are actually also fluorescent, btw.
    • Informative Informative x 1
  7. BigOwl

    BigOwl Mu-43 Regular

    May 3, 2013
    San Antonio, Texas USA
    Real Name:
    Just one more data point for what it's worth. I have only one Oly lens, the 75 mm f/1.8, so I compared it to the Lumix 45-200 zoom set at 75 mm. With both set to f/5.6 using aperture priority at ISO 400, I got an indicated 1/20 sec shutter speed for both lenses. No measurable difference. This was an indoor shot under incandescent lighting.