See also: Leica DG 8-18 mm F/2.8-4.0 lens report

See also: E-M1 Mk.II Lens Fest, Part 1

Angle of view

Let us start from a quick comparison of the angle of view for various focal lengths. Here are four pictures, shot from exactly the same position (but with some changes in the direction), and converted to RGB by the camera's firmware.

Update: I have added a fifth picture, shown at the right. It is from the same shot (read: captured data) as the first one (below left), but the RGB conversion was done by a third-party program, ACDSee, without correcting the image for geometric distortion. This could be referred to "as captured".

I am discussing distortion in the lens report; the image is shown here just to make a point that the image "as captured" may have a wider angle than one after a factory-default processing.

Back to Leica DG 8-18 mm F/2.8-4.0

Back to E-M1 Mk.II Lens Fest, Part 1

F = 8 mm

As expected, for this scene and lens angle I needed a bit more than the usual -1/3 EV of exposure compensation — the brightly-lit, red-roofed house in the frame center is not large enough to be noticed by the ESP matrix-metering system (even the bright patch of grass in front of it does not help).

Just in case, I shot the whole session (all focal lengths) in duplicate, using both -1/3 and -2/3 EV, and for each length I chose the correction only after a closer look at images.

Other parameters: aperture priority, exposures from 1/3200 to 1/50 s, ISO 200, WB at Sunny (5300K), Natural Picture Mode with sharpness and contrast at -1, noise filtering at Low. Saved as 1:2.7 JPEGs.

1:1 samples from marked areas follow as usual, but if you need to scrutinize full, original frames yourself, here they are:

F/2.8 | F/4.0 | F/5.6 | F/8.0 | F/11 | F/16 | F/22

The lens is impressively sharp wide open, both center and corners. There is (almost) no improvement going from F/2.8 to F/4. Image deterioration due to diffraction starts above F/5.6; it is just visible at F/8, quite obtrusive at F/11, and obscene at F/22.

Although it may seem counterintuitive, better lenses start suffering from diffraction effects earlier (i.e. at wider apertures). This is just a simple game of numbers: other optical flaws are corrected earlier so that the increase in diffraction effects (usually masked by a decrease in other, non-diffractive components of sharpness) becomes clearly visible.

In a lens of lower quality the non-diffraction component of unsharpness may keep dropping longer, more than offsetting the rise of the diffractive one. The optimum (lowest unsharpness) is reached later, already well in the diffraction zone of F-numbers, but its magnitude will be higher.

There are no signs of chromatic aberration (or purple fringing); this is now, I believe, handled at the stage of demosaicing of the sensor signal. Still, see the lens report for (a bit) more on this subject.

` Leica DG 8-18 mm F/2.8-4.0

Back to E-M1 Mk.II Lens Fest, Part 1

F = 10 mm

The matrix metering still does not respond to the small, bright house, so I decided to stick to the -2/3 EV exposure compensation used for 8 mm. The resulting shutter speeds were between 1/2500 and 1/50 s.

All camera settings were like in the previous series.

The full-size originals are here:

F/3.2 | F/4.0 | F/5.6 | F/8.0 | F/11 | F/16 | F/22

Very similar to the previous series, perhaps with the sharpening artifacts in the branches easier to notice (but this is not about the lens).

I'm getting an impression that the lens is slightly better at 10 mm than at 8 mm, but this may be an illusion due to higher magnification. To check that, I should compare images shot in the same scale, from a varying distance. Maybe one day.

Back to Leica DG 8-18 mm F/2.8-4.0

Back to E-M1 Mk.II Lens Fest, Part 1

F = 12 mm

At this image magnification the camera's ESP (Electro-Sensitive Pattern, which means exactly nothing) metering did properly react to the brighter area, so I was able to use the images shot with -1/3 EV compensation. Not that it matters, really.

All camera settings were exactly like described above. The shutter speeds varied from 1/2000 to 1/40 s.

The full-size originals, if you want to see any:

F/3.4 | F/4.0 | F/5.6 | F/8.0 | F/11 | F/16 | F/22

Nothing really new, compared to the previous series.

Back to Leica DG 8-18 mm F/2.8-4.0

Back to E-M1 Mk.II Lens Fest, Part 1

F = 18 mm

Again, the shots with -1/3 EV of exposure compensation were selected for this series. All settings as previously, shutter speeds ranged from 1/1600 to 1/60 s.

The full-size originals:

F/4.0 | F/5.6 | F/8.0 | F/11 | F/16 | F/22

Bad luck with the sample at F/11: I wasn't paying attention and did not wait for the small cloud to go away. Oh, well, nothing much lost: this is in diffraction zone anyway.

Back to Leica DG 8-18 mm F/2.8-4.0

Lake sample set summary

All signs show that we are dealing with a praiseworthy lens. I would love to compare it against the Four Thirds ZD 7-14/4.0, but any such comparison would require the same camera to be used with both, and I no longer have that lens. The samples I took with it used a 10 MP camera; compared with 20 MP here that's a factor of 1.4× in resolution; a major handicap for the older lens.

What next?

First of all, a High Res trial. Then, during next trip to Poland, some urban shooting. Unless something else comes up.

Extra: Geometric distortion

This section has been largely expanded and is now a part of the Leica DG 8-18/2.8-4.0 Lens Review.

Back to Leica DG 8-18 mm F/2.8-4.0

Back to E-M1 Mk.II Lens Fest, Part 1