CV 21mm F4 Color Skopar on Leica M9

[artspraken]

Advertisement (gone after registration)

There are some reports on internet (eg. ken rockwell) which claim there is a vignetting / falloff / magenta corner shift problem when the Voightlander 21mm F4 is used with M9. Apparently, the back of the lens protrudes and is too close to the M9 sensor.

 

However, I checked that some of these reports pre-date M9 firmware 1.116. As you know, one of the key improvements in FW 1.116 is that the magenta correction for wide angle lenses.

 

Please can somebody comment whether the FW1.116 has fixed or improved the performance of CV 21mm F4 on Leica M9? Or did FW1.116 only improve Leica lenses, not the CV lenses?

 

Thank you in advance for your help. I am looking to buy this lens as an affordable wide angle 21mm for my M9, because I cannot afford the Zeiss or Leica 21mms at the present.

[Archiver]

Firmware upgrades are unlikely to directly improve Voigtlander lens performance on the M9. Leica will only address their own lenses. Having said that, any improvements in Leica lens performance might be pass over to the CV's and Zeiss' as well.

 

I have the CV 25/4P. It is a stunning little lens on a film body, but has a distinct red edge on the M9. The red edge can be fixed with some PP, but if you want problem free master images, you won't get them, regardless of the lens you manually select.

[lars_bergquist]

Did anybody mention coding? You should.

 

The old man from the Age of the Enigma Codes

[k_g_wolf ✝]

On the M8 the V/C 21 mm worked like a charm. When I tried it on the M9:

too much colourshift/ red corner-syndrome (on the left of the frame),

even with the latest firmware.

 

Sold it.

 

In case you are good at postprocessing, you might give it a try.

The newer version shows a recess at the flange and can easily be coded.

 

 

Best

GEORG

Edited November 16, 2010 by k_g_wolf

[bocaburger]

I have the 21/4 Skopar (LTM version). It worked well on the M8 coded as a 21 Elmarit-ASPH, but on the M9 it gives lots of red-edge. The only code that eliminates the red edges is the one for the pre-ASPH 21 Elmarit, and only at ISO 160. From there on red edge creeps in progressively worse.

 

There are 2 solutions, both involving Cornerfix. One, code it as the 21 pre-ASPH and make separate Cornerfix profiles for each higher ISO. Two, leave it uncoded and then you only need one Cornerfix profile no matter what ISO. I code mine as a 90/2.5 Summarit. It receives no cyan vignetting correction from the M9, so one Cornerfix profile suffices, and it provides a means of ID for finding them to process in Cornerfix (I don't own a 90/2.5). Similarly, I code my 15 Heliar as a 135/2.8, another lens I don't own which the M9 doesn't attempt to correct cyan vignetting.

 

Given the sharpness and small size of the Skopar compared to my Elmarit, I'm keeping it as my travel-21.

[Shootist]

I must say this red edge problem is specific to the M9. In other words it is a camera problem and is/was noticeable with Leica lenses also.

[Dan States]

Advertisement (gone after registration)

Couldn't you always just use Cornerfix for this issue?

[adan]

Couldn't you always just use Cornerfix for this issue?

 

I guess that depends on how much one enjoys screwing around with images in multiple programs. Personally, if a lens requires CornerFix, that lens goes on the auction block (or gets crossed off my "buy" list) - I have better things to do with my life.

 

(No knock on Sandy's work in creating CornerFix. I'd "use" it if Leica paid him a nice licencing fee and built it into the firmware )

 

My pre-ASPH 21 Elmarit fortunately only shows visible red-edge in about 1 photo situation in 10 (depends on the color/tone distribution in the subject matter). If it actually required CF for every shot, I'd have bitten the bullet and moved on to a 21 ASPH long ago.

 

Realistically, red-edge is not an "M9" problem - it is a "full-frame sensor with old-style wide-angle rangefinder lenses designed for film" problem. The M8 shows little or no red edge because the red edges are cropped off - crop an M9 image to the M8 sensor area, and the red edges disappear with the M9 in exactly the same way.

 

Leica has told us for seven years that there were serious difficulties in getting digital images from short-focus non-SLR wide-angle film lenses on digital sensors. (And they aren't alone - Hasselblad says the following about using the short-focus, Leica-RF-like 38 Biogon on their digital backs: "Not recommended for critical work together with SWC models and ArcBody due to optical incompatibility." As in - color shifts at the edges, even with a crop.)

 

They weren't kidding.

Edited November 16, 2010 by adan

[250swb]

Its a great little lens if you get a good copy, but there is red edge no matter what you code it as. Post processing and/or Cornerfix sort it out, but as my prefered medium is B&W I don't normally find it a problem anyway. So you should consider the number of times you'll be needing to fiddle about with the red edge in pp to judge whether you'll get along with the lens or not.

 

Steve

[bocaburger]

I guess that depends on how much one enjoys screwing around with images in multiple programs. Personally, if a lens requires CornerFix, that lens goes on the auction block (or gets crossed off my "buy" list) - I have better things to do with my life.

 

The way I look at it is, if I had to buy all late-model Leica glass to avoid using Cornerfix, instead of spending a few minutes more post-processing I'd have to spend months working longer hours to earn the tens of thousands of dollars needed to upgrade. And I have better things to do with my life than that!

[adan]

Understood - we all have different priorities

 

My post-processing hours ARE my "working hours" and since I get paid by the assignment, picture or page, not by the hour (and have to meet deadlines) a lens that reduces post-processing time saves money, and my reputation.

 

I'd love it if the 21 Skopar (or the old Super-Angulons) and the M9 were more compatable, for those times I don't need f/2.8.

[Shootist]

I

 

Realistically, red-edge is not an "M9" problem - it is a "full-frame sensor with old-style wide-angle rangefinder lenses designed for film" problem. The M8 shows little or no red edge because the red edges are cropped off - crop an M9 image to the M8 sensor area, and the red edges disappear with the M9 in exactly the same way.

 

 

 

They weren't kidding.

 

Realistically it is a M9, FF, problem. That is why there is more of a red edge on the left side of the image then on the right. This happens with every wide angle lens by any maker, including Leica and has been documented here on this forum.

 

It is either a camera, M9, problem or every lens ever made by any manufacture for M mount cameras is skewed to the right the exact same amount.

[adan]

"Realistically it is a M9, FF, problem."

 

Not until you can show that a "non-M9", FF (rangefinder) camera is able to avoid the problem.

 

If I put a cow in a field and it gives strange-tasting milk, it could be the cow, or it could be the field. Until I have another cow and/or another field to use for comparison, I have no way of knowing. Basic scientific technique, called "controlling for all the variables."

[Shootist]

"Realistically it is a M9, FF, problem."

 

Not until you can show that a "non-M9", FF (rangefinder) camera is able to avoid the problem.

 

If I put a cow in a field and it gives strange-tasting milk, it could be the cow, or it could be the field. Until I have another cow and/or another field to use for comparison, I have no way of knowing. Basic scientific technique, called "controlling for all the variables."

 

That is just a load of bull dong. There is no other digital FF RF, as you well know. But please explain why there is more red edge on the left then the right of images. Do these same lenses show anything like this on film, IE different fall off from left to right of images? Or are they equal?

[adan]

(sigh!) No - Leica-mount lenses do not show digitally-induced artifacts when they aren't shot on digital. I gotta admit, ya got me there.

[sdh]

Whatever it is, the red edge(s) are definitely there on the M9. Offsetting that, the lens is incredibly small and light, and is capable of outstanding results. If you can tolerate an extra step in your workflow, CornerFix does a superb job of correcting red edge and vignetting. When I had the lens, I opened all my 21mm shots in CF, then pulled the corrected versions into Lightroom. I thought the results terrific, but I decided that life was too short for this bit of extra complexity and bought a 21mm f2.8 Zeiss instead. Couldn't be happier, even though my camera is heavier and my wallet lighter.

[Shootist]

(sigh!) No - Leica-mount lenses do not show digitally-induced artifacts when they aren't shot on digital. I gotta admit, ya got me there.

 

Andy you are really splitting hairs here and picking on everything I type that you can.

 

But please explain why all wide angle lenses on the M9 show more red edge on the left of images then on the right. I have asked you this many times in this thread and still no response to that question.

Oh I got you there. No one can explain it, only Leica and they are talking.

 

Have it your way. You are right, right, right.

[bocaburger]

Whatever it is, the red edge(s) are definitely there on the M9. Offsetting that, the lens is incredibly small and light, and is capable of outstanding results. If you can tolerate an extra step in your workflow, CornerFix does a superb job of correcting red edge and vignetting. When I had the lens, I opened all my 21mm shots in CF, then pulled the corrected versions into Lightroom. I thought the results terrific, but I decided that life was too short for this bit of extra complexity and bought a 21mm f2.8 Zeiss instead. Couldn't be happier, even though my camera is heavier and my wallet lighter.

 

The last trip I took the 21 Elmarit and ended up taking 6 shots with it out of hundreds, all at f/5.6-8. In the future I'm going to save the weight and bring the Skopar, save the Elmarit for times when I'm sure I'll be shooting 21mm extensively, and/or need the extra stop. It's such a tiny little lens, and really a great optical performer, I really would hate to sell it.

[adan]

"It's such a tiny little lens,...." Therein lies the rub. And believe me, I use Leica Ms for compactness, too, so I'm not happy about it.

 

But if one looks at the sub-28mm lenses for Leica RF mounts:

 

Those that are big(gish) in length show little or no red-edge problem: WATE, 21/24 Luxes, Zeiss 21 Biogon, 24 Elmar (at least compared to the C/V 21/25 Skopars)

 

Those that are medium-sized show red-edge at a correctible level: 21/24 Elmarits, 18 Elmar (maybe - I'm not clear on its currrent red-edge status)

 

Those that are compact show the most red-edge: 15 SW Heliar, 21/25 Skopars, 21 Super-Angulons, Zeiss 21 f/4.5. The waters are muddied a bit because most of them are not Leica lenses, and therefore have no dedicated code to correct them precisely.

 

Digital sensors just behave better with light delivered through a long tunnel, at least for the time being.

 

Shootist, we seem to have two points in dispute:

 

a) Is the red-edge problem limited to the M9, or is it a problem that would show up with any FF sensor with short-focus (non-SLR) lenses. We'll have to wait for Zeiss or somebody to make a comparable non-SLR, 24 x36-sensor camera, and then put a 21 Skopar on it, to be certain.

 

There is one other way to put a short-focus wideangle in front of a digital sensor - use a Hassy CFV (or other digital) back behind a 38 Biogon. And, lo and behold, Hassy warns in their product literature that this combo is "Not recommended for critical work.....due to optical incompatibility."

 

Which brings me back to "it is a 'full-frame sensor with old-style wide-angle rangefinder lenses designed for film' problem" - not specific to the M9.

 

Why the assymetry? Could be lots of reasons. Last week I had a chance to shoot with a demo 21 Summilux. Part way through the shoot, I switched to a 135, and changed the camera to manual/135 lens corrections. And forgot to switch back when I remounted the 21. So the last part of the shoot was with the 21 "coded" as a 135. Neither before nor after the mistake did the Summilux show any signs of red edge (asymmetric or otherwise). On the same body that shows assymetric red edge with my 21 Elmarit pre-ASPH.

 

I don't know of ANY theory about red-edge that explains those results (except maybe that my 27-year-old 21 is itself decentered). Which is why I avoid trying to explain it.

[Advertisement]

Hallo,

Look here: Leica M

[ho_co]

Ed--

No one has been able to explain why the red edge is stronger toward one side of the image.

 

It isn't decentration, though that was a nice quickie instant explanation a year ago.

 

It isn't "just the M9," because the early Nikon- and Canon-based Kodaks also showed what was then known as the "Italian Flag Syndrome" in which the left edge was red and the right edge cyan.

 

To you it makes sense to say it's an M9 problem because that's the only camera where it's obvious today, but in fact that explanation is missing the forest for the tree in front of you.

 

 

I already expressed my opinion on a proper approach to the red-edge syndrome at http://www.l-camera-forum.com/leica-forum/customer-forum/116238-la-izquierda-roja—-homage.html.

Edited November 17, 2010 by ho_co