Panasonic G 20mm underexposure with G1

Yesterday I finally managed to get hold of a Panasonic 1.7/20. I drove to Germany for it, in The Netherlands it doesn't seem to be available (and believe me, I checked!).

But let me get to the point. When testing it on my G1 the pictures look distinctly underexposed by about 2/3 stop at all apertures. The 14-45 zoom doesn't show this, direct comparison between the two in ACR shows a difference of almost 1 stop. I took pictures in artificial light with the white balance set to incandescent lighting.

When I used a flash with a fixed output and M(anual) mode on the G1 both lenses showed equal exposure at the same aperture within 0.1 stop or so, so aperture calibration is not the problem. Using the built-in flash yields a difference of 0.1 stop, so that's OK too. It looks like the exposure measurement is confused by the 20mm lens under certain circumstances.

I checked the Panasonic site. My G1 has the latest firmware V1.4, the 20mm has V1.0. I couldn't find any updates for the 20mm.

Has anyone else noticed this? Does anyone know what's wrong here? Thanks.

Checked in daylight today: no underexposure in daylight even when dim, but still under incandescent light. So it has something to do with the colour temperature. Very strange!

No, not experienced that I'm afraid!
--
Regards,
Rich Simpson

its also happening to me with my GH1 latest firmware, there was a thread on here earlier also describing the same thing. Im trying to figure it out also

Thanks, I've read through the other thread and now I have a theory: maybe the 20mm transmits a larger fraction of infrared light than e.g. the 14-45mm. Because incandescent light has a lot of IR this could cause underexposure.

ad2005 said:

Thanks, I've read through the other thread and now I have a theory: maybe the 20mm transmits a larger fraction of infrared light than e.g. the 14-45mm. Because incandescent light has a lot of IR this could cause underexposure.

Click to expand...

I doubt that is the case, the difference in transmission of near-IR in most glasses is vanishingly small. The UV cutoff does move a bit, but really very little happens on the IR end. Besides this, the G cameras have essentially zero IR response, they have a very, very strong IR blocking filter (hot mirror). Finally, and most importantly, the m43 cameras meter with the imaging sensor itself (unlike DSLRs) so if there was a difference in IR response sufficient to cause the metering to change it would be blatantly obvious in the final image as well!

But yeah, something weird is going on - especially if it is color temperature dependent. I just got a 20mm on Friday but have hardly had a chance to try it on my G1. I'll definitely do some exposure tests now. Thanks for the heads up!
--
Ken W

Rebel XT, XTi, Pany G1, LX3, FZ28, Fuji F30, and a lot of 35mm and 4x5 sitting in the closet...

I too have noticed this with my GH1 and the 20mm. And while it's better outside, many times I still have to bump EC +1/3 or +2/3. I know the G cameras seem to error on the side of underexposure anyway, but you're definitely right that it's more pronounced inside in low light.

What does the Live Histogram show when you are taking these underexposed photos? Can you tell before you take the shot they are going to be underexposed? I know the LCD screen is probably not a good indicator because the camera auto gains in order to make it brighter. But I would think the histogram should give you an indicator that the shot is going to be underexposed?

I haven't taken the time yet to do some testing, but I haven't really recognized a pattern yet. Even my 14-140 errors on the side of underexposure sometimes, but I haven't actually taken side by side shots with both lenses to see if it's more prevalent on the 20mm.

Travis
--
http://travisimo.smugmug.com/

Yes, I can see the underexposure in the live histogram: the highlight room is very obvious. When I check the histogram with the 14-45 it is often spot-on or shows blown-out highlights.

I will try to post a video comparing it to another lens when i have time

I just checked the metering of my G1 with the 20 and 14-45 in A mode set at 5.6 on a tripod with the zoom carefully set to match the framing of the 20. Scene included beige carpet and black and brown objects under dim tungsten light at ISO 100. In this case the 20 mm actually metered the scene at +1/3 eV compared to the 14-45 (that is the 20 mm metered for a slightly longer shutter speed). I have a custom WB set.

Anyway, just a single data point as I said. I did notice with this scene I could easily get the 14-45 to change metering + or - 1/3eV with just slight changes in the zoom setting (evaluative metering set).

So my camera doesn't seem to exhibit the problem in this particular case. Of course there are many, many variables!

Good luck, let us know if you find a definite pattern I might be able to try to reproduce.
--
Ken W

Rebel XT, XTi, Pany G1, LX3, FZ28, Fuji F30, and a lot of 35mm and 4x5 sitting in the closet...

It is very scene-dependent, especially when using fluorescent lighting of low colour temperature. One scene consistently shows 2/3 stop underexposure of the 20mm w.r.t. the 14-45mm and another shows slight overexposure. In both scenes (and a lot of others) the 14-45mm yielded the best exposure.

I couldn't find a clear-cut dependence of the colours in the scene. A blue carpet got slightly more exposure from the 20mm, a book closet with a lot of different colours and fine detail could yield up to 2/3 stop underexposure, a yellow curtain was very slightly overexposed.

It's a bit alarming that the review is darker for the underexposed pictures than the finder image (clearly seen in auto-review mode); it looks like the camera changes its mind during picture taking/processing. This isn't the case for the 14-45mm, ever!

I must say this drives me a little bit crazy. In principle the 20mm is eminently suitable for low-light pictures, but the erratic underexposure creates a lot of unnecessary noise, even at ISO 400. Let's hope that Panasonic can fix this with a firmware update of the lens and/or camera... At this point I'm a bit weary to use the 20mm in low-light.

Nothing is perfect. I have been dealing with slight variations and differences in lenses and cameras for years. That is what exposure compensation is for. I regularly compensate with all my cameras and different lenses.
--
http://digitalphotonut.zenfolio.com/

Now I'm almost sure what's going on with this underexposure. It appears that (on my G1 at least) the diaphragm closes down when there's more light during viewing. When it's dark, and that's when underexposure occurs, the lens is wide open during viewing. My guess is that the software of the camera and/or lens do not correctly account for this varying aperture during viewing.

I remember having seen a similar story for the Olympus E-P1 w.r.t. closing down the aperture in brighter light. There it was stated that the Pany camera (a GF1 I think) left the aperture always wide open. Not true for my camera!

Wondering if you ever verified that the 20 1.7 actually has what would appear to be a design defect that causes it to underexpose under some lighting conditions. I've just gotten one and am trying to determine if this is a flaw in design or I have a defective lens. I've got a new thread on this going at dpreview, but so far nobody seems to be experiencing what you and I have. Also, I have noted that the lens does not show this issue when I use spot metering, so I'm not convinced of the stop-down theory you expressed in your thread on the subject. Wouldn't this also affect spot metering?