M9 and CV15 at the Getty Center (Images)

[Alnitak]

Advertisement (gone after registration)

I got my M9 back from getting the rangefinder calibrated, and things are spot on now. I am taking a few days off this week to be with family; my in-laws are in town to see us. Today we took them out to the Getty Center. I took the M9 and a small group of lenses. It was validation of why I switched to the rangefinder this past July--the whole bag with five lenses was nice and light. I took the 75 and 90 APO's, the 28 'cron, 50 'lux ASPH and the CV15.

 

First, the "red edge" problem is quite prominent with the CV15:

 

 

However, CornerFix 1.3.0.3 does a great job of correcting the problem:

 

 

Entrance Stairs; note the excellent flare control:

 

 

Entrance Hall:

 

 

I will post more images in a few replies to this thread.

 

Jeff

[Alnitak]

Garden Terrace and Central Garden:

 

 

Fountain, Main Plaza:

 

 

Cactus Garden and Wilshire Boulevard skyline:

 

 

Central Garden:

 

[Alnitak]

Final few images:

 

Central Garden #2:

 

 

Waterfall and Azalea Pool, Central Garden:

 

 

Great trip, and I love the full frame of the M9. Lots more creative opportunities with wider lenses.

 

Jeff

[doug_m]

Did you use Cornerfix for all images posted or just the first one? These are very nice images. I wonder how the 21 or 24 lux would do with the same subject matter?

[Alnitak]

Did you use Cornerfix for all images posted or just the first one? These are very nice images. I wonder how the 21 or 24 lux would do with the same subject matter?

 

I had to use CornerFix on all of them. The red is just way too noticeable on the images, no matter what the lens is set to. As for the 21 or 24 lux, I don't think it would be the best comparison. These were all shot stopped down to f/8 or f/16 for DOF, so the speed of the two luxes would be useless, and neither is very wide.

 

The better comparison would be the WATE. I actually have a WATE on the way and it should be here next week. Sean Reid's comparison showed the 15/4.5 to be a bit sharper than the WATE, but then there's the convenience of the flexible focal lengths.

[lars_bergquist]

We still do not have a satisfactory explanation for the Red Left Edge phenomenon. Sensor? Firmware? Lens? (this one is v-e-r-y improbable). Is Leica working on it? Of course. Why are they not telling us about it? They do not have a solution yet. The bizarre thing about it is that it is just the left edge. The M8 IR/UV cyan shift was at least symmetrical.

 

Until there is a known cause, we are just treating a symtom. Long term viability of super wide angle lenses on the M9 depends on finding the root cause, and rectifying it.

 

The old man from the Age of Black and White

[Alnitak]

Advertisement (gone after registration)

Lars, I guess I am pretty confident it will be fixed, whatever the cause. It's pretty easy to fix with CornerFix. It also doesn't show up until I use a 21mm lens or wider.

[doug_m]

Has anyone seen the red left edge phenom. with any of the 24s- particularly the 24 lux? What does a UV/IR filter do to the red edge when used on the M9

[Alnitak]

The one focal length I don't have in my collection is 24mm, so I can't check it. As for the filter, I have tried it and it doesn't really do anything.

[lars_bergquist]

Doug -- of course an UV/IR will counteract the Red Edge to the left, but create Cyan Corners on the right. As they said when I was a kid, however you turn, your arse stays behind.

 

One of the strangest things about the Red Edge is of course that it is not curved but straight -- it is not a corner thing, but an edge thing. (This points very strongly to firmware, or -- horror -- a sensor manufacturing error.) Cyan shift is a corner thing: it increases radially from the optical center. So no purely radial correction, be it internal or external, wil cure the disease. CornerFix is not a cure, it is an admission of defeat. The state of things is not acceptable.

 

And we would be very relieved if Leica could state at least that they are on the way to a cure.

 

The old man from the Age of Black and White

[doug_m]

Lars- I was told by an in-house Leica expert in the US that he thought the design of the front element on the 24 lux would tend to avoid the cyan drift otherwise thought to exist on wide lenses when using IR/UV filters with this lens and the M9.

 

It seems to me (an admitted tech challenged M user) like he is correct- based on my casual unscientific observations with this combo. Do you or anyone else have an explanation or knowledge or evidence to the contrary? I'm very curious since I'd like to keep the UV/IR filter on my 24 lux which I also use on the M8. My "evidence" request is sincere and not meant, in any way, to be argumentative.

 

Thanks,

 

Doug

[lars_bergquist]

This seems most improbable. The drift occurs not in the lens, but in the filter. In other words, the light is already 'drifted' when it hits the lens.

 

Admittedly, my only experience with the SW Summiluxes is with the 21mm.

 

The old man from the Age of Black and White

[UliWer]

We still do not have a satisfactory explanation for the Red Left Edge phenomenon. Sensor? Firmware? Lens? (this one is v-e-r-y improbable). Is Leica working on it? Of course. Why are they not telling us about it? They do not have a solution yet. The bizarre thing about it is that it is just the left edge. The M8 IR/UV cyan shift was at least symmetrical.

 

Until there is a known cause, we are just treating a symtom. Long term viability of super wide angle lenses on the M9 depends on finding the root cause, and rectifying it.

 

The old man from the Age of Black and White

 

Though it seems to be the most bizarre thing that even people who are reasonable and sticking to the facts do strongly believe that the issue is only one of the left side.

 

If you have the 18-mm Zeiss, a M9 and a M8, Lars, you could try to do a test, which might give us some enlightenment:

 

Please take a picture of a grey card under controlled conditions which make sure that there is no uneven lighting. Try with lens detection on and off, with and without UV/IR-Filter.

 

From your results of this test you will be able to make up your mind for the following questions:

 

Is it only a M9 issue?

Is it bound to the left side?

Do you see any red-shift without lens-detection?

Is there a substantial difference between the space the red shift takes with "lens detection on" and the cyan shift takes on the picture if you use UV/IR-Filters and "lens detection off" ?

 

From all I know (trying the Leica 3.8/18 with the M8) your answer will be "no" for all questions. Though I may be wrong and falsification of wrong theories may lead to progress (especially when "red" issues and "leftist" theories are concerned ... )

 

From what was reported about a Leica meeting in Wetzlar some weeks ago (I havn't been there but just read what other people said) Leica is working on the issue testing the results and tweeking their solution for different lenses and different conditions. The next firmware update which was announced to come out soon - for M9 as well as M8 - will show if and how far they have been successful.

[sandymc]

To pick up a few of the questions asked (or implied) here:

 

1. Is this a lens issue? - Depends on what you mean by "lens issue". However, among the image samples I have, red edge phenomenon (a) only occurs on lenses that are de-centered and ( is worse the more decentred the lens is. That doesn't however mean that the lens is "causing" the problem.

 

2. Is this only a M9 problem? - No, it existed on the M8 as well, but was almost never large enough to be visible. You can clearly see it on vignetting plots however. The only difference with the M9 is that the effect is large enough to be easily visible.

 

3. Is it bound to the left side? So far, of all the M9 images with red shift people have sent me, the red is always on the left, and the de-cenetring is always to the left. If anyone has a red edge to the right image, please send it to me!

 

3. Is there red shift without lens detection? - yes, in the sense of a asymmetry being there, although its effect is generally masked by the uncorrected "normal" vignetting. However, be aware that just because lens detection is off doesn't mean some firmware corrections aren't being made, so that doesn't rule out firmware problems/fixes.

 

Sandy

[SJP]

FWIW - I am convinced it is the sensor (microlens drift), can't see how it could be the lens as a lens is radially symmetric in its abberations.

[UliWer]

3. Is it bound to the left side? So far, of all the M9 images with red shift people have sent me, the red is always on the left, and the de-cenetring is always to the left. If anyone has a red edge to the right image, please send it to me!

 

 

I cannot give you an example of pictures where the red shift is only visible on the right edge, but here are some with regular left/right shifts:

 

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/103003-cornerfix-1-3-0-0-available-2.html#post1092262

 

I have posted more, but meanwhile it's hard to find them. Perhaps I find some more this evening.

[doug_m]

Sandy,

 

Since you are an expert in this area, can you comment on my last post and the questions I asked?

 

Thanks,

 

Doug

[Bo_Lorentzen]

Alnitak,

 

Great shots,

 

Maybe I missed it, I was wondering if you coded the 15mm..? I shot quite a lot with mine coded as a WATE and leaving it at the default 18mm which seems to look pretty good, a little vignetting but I don't feel its quite as much as the first image you show. again I could easily be wrong as I have not actually shot a solid wall yet. :-)

 

Great shots, Im in LA, Beverly and Western, shoot me a email if you ever look for a excuse to have coffee and talk cameras.

 

.

[sandymc]

Sandy,

 

Since you are an expert in this area, can you comment on my last post and the questions I asked?

 

Thanks,

 

Doug

 

Not an expert on this, I'm afraid - as Lars correctly points out, we don't know the root cause of the problem. Until we do, nobody qualifies as expert!

 

However, while I'd certainly agree that the design of the lens overall is a significant factor - e.g., that controls the exit angle, I'm not sure what was meant by the Leica expert's reference to the front element specifically.

 

Regards,

 

Sandy

[Advertisement]

Hallo,

Look here: Leica M

[adan]

FWIW - I am convinced it is the sensor (microlens drift), can't see how it could be the lens as a lens is radially symmetric in its abberations.

 

Well, an ideal, perfectly manufactured lens, yes. In the real world, lenses can have centering issues: either the whole lens is not perfectly centerd in the mount, or inidividual elements are either not perfectly centered in grinding, or in mounting.

 

E. Puts specifically refers to decentering issues visible in the first round of C/V lenses (15 and 25 scale-focus versions) in his review of them.

 

I woudl imagine some Leica lenses have left the factory not perfectly centered - within the tolerances of the time, but not perfect.

 

It wouldn't surprise me to find that my 21 pre-ASPH, made in Canada 25+ years ago, for film use, was not perfectly centered.

 

That's just for the record - microlens drift, as you say (I assume you mean the pattern of microlenses is not perfectly centered on the face of the sensor) is still a suspect, as is imperfect centering of the sensor behind the lens when installed, or even other factors.

 

It would take someone with multiple M9s and multiple versions of the same lenses to really test for red drift and red-left drift and see if it always-but-only happens with a certain lens, which would implicate the lens centering - or if it always-but-only happens with a certain body, which would implicate camera or sensor construction faults.

 

------------------------------

EDIT: Actually, I almost meet these qualifications. I used my current 21 pre-ASPH on an M9 demo camera (fw 1.002) on 9/9/09. I just went back and looked at those shots again, and I don't see any sign of the red left drift that I get from my production camera (or any red drift anywhere). Some from my 15c/v, but not the 21.

 

Which is too bad, because it may mean a manufacturing slip-up. Some cameras getting out of the factory with sensors or sensor boards out of place. Which in turn means the fix would be "Send it in for adjustment."

------------------------------

 

BTW, by my math, if the red/left drift I'm getting is caused by a miscentered sensor (or lens), it is out of place by about .5 mm. If it is caused by microlens drift, then the error would only have to be about 1 micron (1/7th of a pixel) or less. But chip makers routinely work to tolerances of less than 80 nanometers, or .08 microns. It would take a 1200% error in microlens placement to get the same amount of red-left drift.