Color shift on the right with Elmar 24/3.8

[dasjak]

Advertisement (gone after registration)

Hi,

 

today I recognized ugly color shift happening on the right side (here it´s the top) with my Elmar 24/3.8 on my M 240. Aperture was set to f11. I don´t use that lens very often so I just saw it today shooting seascapes with cloudy sky.

 

Is that normal? Cannot be normal, right?

 

Thanks for your input!

 

Best,

Andreas

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited May 21, 2016 by dasjak

[lct]

May happen, easy to adjust in PP. See http://www.l-camera-forum.com/topic/138133-red-edge-with-m924mm-elmar-can-leica-fix-it/

Now before you junk your camera i would check if Lens Detection is set to Automatic:

- If yes, it may be that there is some dirt on the coding pits of the lens and/or the 6-bit sensor;

- if not, it may be that Manual Lens Detection is not set to the right lens. I would try "24 f/2.8..." instead.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jaapv]

Actually this means that the lens is minimally out of true to the optical axis which may be due either to the camera mount or to the lens. I would send the file to Leica and ask whether they can correct it, or decide to live with it and deal with it by flatfield correction if required.

[gpwhite]

The posted image also looks quite soft, by Elmar-M 24mm ASPH standards. Unless you experienced a fair amount of camera movement during exposure, the lack of characteristic super crispness seems more of a problem than just color fringing. Have you tried achieving perfect focus on your M240 via LV, just to be sure it is possible with the lens?

[Paulus]

I had a lot of colour shift experience with the Zeis 4,5/21mm. Especially on the M9 in this kind of weather/ clouds. 

 

Decided to change it for a Super-Elmar 21. Did not regret it a moment.

 

Still it's a very nice picture. Reminds me of Caspar David Friedrich. Just one question, did you hold the camera with the release-button down or up?

Edited May 22, 2016 by Paulus

[dasjak]

The posted image also looks quite soft, by Elmar-M 24mm ASPH standards. Unless you experienced a fair amount of camera movement during exposure, the lack of characteristic super crispness seems more of a problem than just color fringing. Have you tried achieving perfect focus on your M240 via LV, just to be sure it is possible with the lens?

the lens is amazingly sharp. i just scaled the image down, i just wanted to show the color shift, not the sharpness

[gpwhite]

Advertisement (gone after registration)

the lens is amazingly sharp. i just scaled the image down, i just wanted to show the color shift, not the sharpness

 

I see; well that is good news. Jaap's comment is likely correct. The SEM 18mm has been shown by a number of shooters to have one corner that is off, as the lens must be very difficult to assemble precisely.

 

My Elmar 24mm was "perfect," as was the other member of the pair of new lenses I chose from. I am afraid your lens needs a visit to the spa in Wetzlar. 

[pico]

I cannot see it. That's why I shoot nothing but monochrome.

[sdk]

On my M9-P my 24mm/3.8 had a red cast in the lower edge and a slight cyan cast on the top edge (the camera's lens profiles also didn't perfectly correct other wideangles like the 28mm/2 ASPH or 28mm/2 ASPH). The least problematic wideangle lens for my M9-P was the Zeiss 35mm/1.4 Distagon ZM, which had minimal color shifts in the corners, just noticeable to me. On my M-P 240 all these lenses all seem better behaved, even the Zeiss with no manual camera lens profile, and need no correction.

You can use the Adobe DNG Flat Field plugin to correct this sort of color cast with any DNG raw file. You need to take a DNG format reference image of a solid white object with the same lens having similar aperture settings, the same lens profile (or none if the original image had none), and similar light. I use a credit card sized piece of translucent white perspex (plexiglas, acrylic) for reference images. The best reference images are those shot at the same time, and place, and light, as the image to be corrected. But a later reference image shot under close to the same ambient light is usually good enough. The plugin can be set to correct color cast alone or color cast plus vignetting, and generates a new, corrected DNG file. It works very well if you are in the Lightroom workflow.

[Monochrome]

I have the same problem with my 24/3.8 on a M262. Using the "automatic" lens detection helps a lot of course, but it is still there.

Correcting the problem in LR isn't very easy, it takes a lot of time. 

The problem doesn't show up in every picture, but is very apparent in vertical shot's with a lot of sky or clouds.

 

(i posted this problem a few months back, may be worth to look at the comments there too)

Edited May 29, 2016 by Monochrome

[willeica]

I have the same problem with this lens on the M240 even if the 6 bit code identification is used. I did not have the problem with the lens on my M9. My wider WATE shows no signs of this issue on either the the M9 or the M240. The 24mm f3.8 lens itself is excellent but this issue is a deterrent in respect of using it for photos with a lot of sky. You can do some adjustments in Lightroom but it is a bit of a pain to have to do this. I think that Leica might be be able to adjust the correction applied to this lens based on the 6 bit code identification. Perhaps this could be done through some form of firmware upgrade, that is if there is such an upgrade before the next truly new M is released.

 

William

[lct]

Most wides produce color shifts below 35mm folks. Do not shoot a white or grey PC monitor if you want to believe that perfect lenses exist in this matter. Easy to solve the problem in PP but using an UV/IR filter may help to a lesser extent. I mean on the 24/3.8 at least as it may produce too much cyan shift with other lenses.

 

Without filter:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

With UV/IR filter:

 

Without filter (bottom left corner):

 

With UV/IR filter r (bottom left corner):

[willeica]

Thanks ict, but based on my experience with different lenses, I have come to the conclusion that Leica could do more to apply correction in respect of wide lenses. It seems to be very hit and miss with different generations of camera and lens models. If my WATE does not give 'Italian Flag' issues, then the 24mm f3.8 should have no issues either. Something for firmware upgrade and certainly for the next M generation.

 

I prefer not to use filters because of reflections etc. I can think of no other make of camera that produces such issues with 'native' lenses. This is a pity as I believe that the issue is very fixable. 

 

William

[lct]

Fact is the 21/3.4 asph has less color shift issues than the 24/3.8 but other camera makers don't design their bodies to fit M lenses do they. Suffice it to compare the size of M lenses to that of Sony ones to see why the sensor stacks or native A7 cameras is so thick compared to digital Ms. My A7s mod has a thin sensor stack made by Kolari Vision and behaves more or less like my M240 with green/cyan shifts instead of red/magenta ones. With a thicker sensor stack i would get smeared corners the same way as i got with my Fuji X-E2. I don't know if a firmware update can improve the 24/3.8's IQ and i do hope so but as far as i'm aware Leica and Kolari vision are the only companies working on actual solutions. Now building a LCC profile is no rocket science and Cornerfix can be downloaded for free...

Edited May 29, 2016 by lct