Need help with diagnosing 21mm 'Lux

[arthury]

Advertisement (gone after registration)

Do you guys see any unusual drop in resolution (as in fuzziness) at the edges when subject is at infinity or closed to infinity and aperture is in this range f/4-f/8?

Is this lens only good for close range?

 

Look at the roof patterns and the main frames in the zoomed-in image. They seem very un-Leica-like.

 

 

EXIF:

ISO 200; f/4; 1/4000sec

 

Full-frame

L1001181-1 by phlog, on Flickr

 

Zoomed in

L1001181-1 TEST SAMPLE by phlog, on Flickr

[arthury]

Is this a de-centering issue?

[Guest Marc G.]

I think this is either a DOF problem or field curvature. Fast ultrawides always suffer from a certain kind of field curvature.

[pgk]

Firstly, your crop shows slight traces of chroma which should be corrected in post - this is in my experience, an issue with ultra fast wides and my Canon 24/1.4L is quite similar. So I'd first correct this and then try many more shots which include fine detail in the corners and try shooting under very different conditions. That said, I would not surprised if performance is lower than that of smaller aperture lenses and that, by looking at images from the lens at high magnifications, the consequences of a design which is pushing the boundaries of current lens design are visible.

 

Whilst it is I suppose possible that you may have an issue with a specific sample of this lens, its really a specialist lens whose performance is almost certain to be optimised, as far as possible, wide open. So when used in more 'normal' circumstances at mid or small apertures its performance may simply not be as good as smaller aperture lenses of the same focal length. If you look at Leica's MTF data, they only show data down to f/5.6 and at this aperture the graph does show lower performance than the SE. Obviously at f/2.8 and 1.4 it shows better performance;) but still not stellar in comparison to the fabulous performance of the SE throughout its aperture range.

 

Which says to me that it really is a case of horses for courses and that specialist fast aperture lenses need to be utilised with caution and in circumstances where they work best. FWIW my Canon 24/1.4L is no wonder lens stopped down either. Adequate but certainly nothing spectacular.

[algrove]

I think you should move to a different neighborhood, where the roofs are of a different construction and see if there are differences.

[arthury]

I think this is either a DOF problem or field curvature. Fast ultrawides always suffer from a certain kind of field curvature.

 

Based on the DoF table published by Leica, the areas of concern are well within the DoF.

 

So, may be it's the field curvature.

[arthury]

Advertisement (gone after registration)

Firstly, your crop shows slight traces of chroma which should be corrected in post - this is in my experience, an issue with ultra fast wides and my Canon 24/1.4L is quite similar. So I'd first correct this and then try many more shots which include fine detail in the corners and try shooting under very different conditions. That said, I would not surprised if performance is lower than that of smaller aperture lenses and that, by looking at images from the lens at high magnifications, the consequences of a design which is pushing the boundaries of current lens design are visible.

 

Whilst it is I suppose possible that you may have an issue with a specific sample of this lens, its really a specialist lens whose performance is almost certain to be optimised, as far as possible, wide open. So when used in more 'normal' circumstances at mid or small apertures its performance may simply not be as good as smaller aperture lenses of the same focal length. If you look at Leica's MTF data, they only show data down to f/5.6 and at this aperture the graph does show lower performance than the SE. Obviously at f/2.8 and 1.4 it shows better performance;) but still not stellar in comparison to the fabulous performance of the SE throughout its aperture range.

 

Which says to me that it really is a case of horses for courses and that specialist fast aperture lenses need to be utilised with caution and in circumstances where they work best. FWIW my Canon 24/1.4L is no wonder lens stopped down either. Adequate but certainly nothing spectacular.

 

Thank you for the explanation. It does help.

[jaapv]

Based on the DoF table published by Leica, the areas of concern are well within the DoF.

 

So, may be it's the field curvature.

Common misunderstanding. Areas within DOF are UNsharp except the plane of focus. It is just that your eyes are unable to discern the unsharpness on a print at normal viewing distance. The DOF tables are based on 1930-ies film rendering a 6x9 cm print at 30 cm.

 

They are totally useless for judging sharpness of a lens. You are looking at the equivalent of a print over a meter wide at short distance here when you blow your image up to 100%. The only area that is sharp will be the plane of focus that will be a few cm (because of pixel size) thick depending on subject distance.

[arthury]

Of all the Leica lenses I have bought in the past, this 21mm Summilux is the most troublesome and left a poor taste in the mouth.

 

A day ago, the front segment of the barrel came loose and the 1.4 f-stop calibration mark no longer match the white dot. In fact, it is flp-flopping between both sides of the white dot, depending on which direction you turn the focusing ring.This is most unexpected and it raises a question about the quality check for this very expensive lens before it leaves the factory.

 

This lens will be returned to NJ. I am waiting for Roxana or Carmen to reply but so far no one has since the employees have returned from the Easter.

 

I'll give it a couple of days more. This is a unforeseen and highly unexpected encounter.

[arthury]

Lens arrived at Leica, NJ. Estimated return is 4-5 weeks to tighten the front segment and CLA.

 

Is this a normal turnaround time?