The M8's Wacky Cyan Wide Edges

[ken_tanaka]

Advertisement (gone after registration)

In the same vein as my previous "Wacky White Balance" thread I also want to point out the maddening cyan edge casts that the M8 continues to paint on most images taken with my 21mm f/2.8 Elmarit. I absolutely adore the sharpness and lack of distortion of this lens...which is why I spent $3,500 to buy it several years ago.

 

So imagine how thrilled I've been that I've not been able to use the lens with the M8 due to the as yet unsolved cyan cast that the camera records most often at the right side of frames captured with this lens. (Two samples below.) I've been just delighted.

 

 

 

Anticipating the predictable questions from the Leica Defense League:

 

a. Yes, my 21mm lens is coded. It was recently tagged with its little Dalmation spots by the Leica folks.

 

b. Yes, the lens was wearing a B&W 486 filter for both images.

 

c Yes, the camera was set to recognize the lens.

 

d. Both images were at ISO 160 and the lens was, I believe, at f/8 for both.

 

Again, my purpose in raising this old topic is to keep it vivid particularly to Leica and to prospective M8 buyers. Like the camera's primitive color balance programming this is another preeminent problem with the M8. In my case it's preventing me from using a wonderful $3,500 Leica lens.

 

The M8's been on the market for approximately 6 months at this writing. Come on Leica. Let's get this crap cleaned up to get the camera's performance and predictability to where it should, and must, be. Either you're going to be in the photography business or you're not.

[robertwright]

firmware?

my 21 did this under 1.091

under 1.102 there is no problem

 

I don't have the asph 21, I have an earlier canadian 21 2.8 elmarit

I am using a b+W thin 486, which is not even "hoyle"

 

I have it coded as the pre-asph 21.

 

otherwise, for me, the images are clean across the board.

[Advertisement]

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

[ken_tanaka]

My firmware level is 1.102.

[Jamie Roberts]

Ken--

 

Do your other coded lenses do this? Or is it just this one?

 

Are you saying this is an intermittent problem (hence wacky?). Or predictable? I also couldn't tell if you were saying "this only happens on the right side of the shot"--from a JPEG this size it's very hard to tell.

 

My coded 28 shows no sign of cyan whatsoever. My uncoded and BW'd 24 Elmarit does; the CV 21 less so (has a Leica filter) but it's there.

[ken_tanaka]

Jamie;

My 21mm f/2.8 Elmarit is the only lens that exhibits this problem. My (uncoded) Zeiss 28mm is fine, as are my other coded Leica lenses (50mm, 35mm, 90mm).

 

The problem was a bit more severe under firmware 1.09. Since 1.102 the cyan tint seems limited to the right side of the frame (in landscape orientation). These days it seems always to be there regardless of ISO or other exposure settings. But in late February or early March I believe I did use the lens indoors once without seeing the problem.

 

So "wacky" applies here, too. I'll be double-checking the lens' coding against the reference list of codes soon, but I really doubt that Leica mis-coded the lens.

[Guest guy_mancuso]

This is were i think the Leica filter is better than the B+W which is also the case on the WATE. I have no cast on the WATE at all with the leica filter and with the B+W i did at 16, 18 and 21. As Sean and myself have pretty much said 21mm and below the Leica filer just handles the cyan cast better. But we need a 55mm leica filter to hit the streets too, so we wait

[Eoin]

Advertisement (gone after registration)

I have a coded 21 Elmarit Asph and I don't have this problem, then again I also don't use a filter either. Leica 55mm filters are due week 21 which is next week so I'll see if the filter introduces the problem.

 

Strangely enough I've never felt the need to use a filter with the 21 and the M8 to the degree that I have with the 35mm.

[Venkman]

Have you checked the EXIF if the lense is correctly recognized?

[wlaidlaw]

My cyan is defintely less on the Leica filters than it is on the 486. I am using a Zeiss 21 coded as a WATE and set to 21mm. If you turn UV/IR on with the menu, using the Leica filter, cyan corners disappear almost completely although the vignetting seems worse with the filter fitted than without - not wholly unexpected on a 21mm lens I suppose. Turning lens recognition on or off does not seem to improve the vignetting a whole lot - just about noticeable. Using 1.102 and my lens recognition does appear on EXIFs

 

Wilson

[punktum]

 

Again, my purpose in raising this old topic is to keep it vivid particularly to Leica and to prospective M8 buyers.

 

Last week I stopped at Leica in Solms. Stefan Daniel told me that one has to use Leica IR Filter from 28mm on down to clarify the Cyan. The Leica Filters are different from the B&W which appears with the wide angles.

[scott kirkpatrick]

My cyan is defintely less on the Leica filters than it is on the 486. I am using a Zeiss 21 coded as a WATE and set to 21mm. If you turn UV/IR on with the menu, using the Leica filter, cyan corners disappear almost completely although the vignetting seems worse with the filter fitted than without

 

Assuming you are comparing vignetting with lens detection ON+UV/IR against vignetting with Lens detection merely ON, the reason is that ON has strong overall vignetting correction -- sometimes too strong, see Sean Reid's results with 50mm lenses -- and in 1.102 the new routines, ON+UV/IR do very little for overall vignetting, concentrating on the red vignetting corner effects.

 

The puzzle seems to be that only three of the four possibilities:

 

vignetting correction ON/OFF

and

filter correction for cyan drift ON/OFF

 

are implemented.

 

scott

[jaapv]

I can only confirm Wilson's asessment: With the Leica filter and coding/filter on, there is no cyan vignetting and slight exposure vignetting on the Zeiss Biogon 21. The exposure vignetting does not bother me, in fact I rather like it. It is however very easily corrected in post. The cyan edges do not appear, apart from the CA that this lens exhibits in general.

[wlaidlaw]

Assuming you are comparing vignetting with lens detection ON+UV/IR against vignetting with Lens detection merely ON, the reason is that ON has strong overall vignetting correction -- sometimes too strong, see Sean Reid's results with 50mm lenses -- and in 1.102 the new routines, ON+UV/IR do very little for overall vignetting, concentrating on the red vignetting corner effects.

 

The puzzle seems to be that only three of the four possibilities:

 

vignetting correction ON/OFF

and

filter correction for cyan drift ON/OFF

 

are implemented.

 

scott

 

Scott,

 

There does not seem to be any additional all light vignetting correction when you change from straight lens detection to lens detection plus UV/IR but obviously there is the cyan vignetting correction, which works well. The point I was making was that I think fitting the filter causes additional all light vignetting, which is not corrected in camera on the 21mm Biogon. Like Jaap it does not particularly worry me and on a real photograph is barely noticeable. If I could get my Biogon to code as a 21 Elmarit (it is stubbornly refusing to do this - it obviously has ideas above its station and insists it is a WATE), it would be interesting to see if there was more non-cyan vignetting correction, as the Elmarit being a f2.8 against the WATE max aperture of f4, theoretically should need more overall vignetting correction.

 

Wilson

[tom0511]

Leica-filters for me is theory. I am still waiting for the 2 free filters.

 

As much as I like the M8- the color and filter and coding confusion just not great at all.

 

What if you convert some older images (made with older firmware)....mmmh, switch back to older profiles....did I use a filter or not? Was it 486 B+W or did I allready have the original Leica filter? Now for landscape I use the M8-c1+UVIR profile to get saturated colors...but the lips of faces look pinkish agaain with that profile-so forthe portraits I could use a different profile....the Leica filter is recommended for wide angels...but the filter holder for the wate is huge and leads to reflections...so I could use a third party filter holder-shade...but now the free-wate-filter I have orderd doesnt make any sense and becomes useless...yes, the Leic filter should avoid the cyan corners...but not at any light temperature....

[scott kirkpatrick]

Scott,

 

There does not seem to be any additional all light vignetting correction when you change from straight lens detection to lens detection plus UV/IR but obviously there is the cyan vignetting correction, which works well. The point I was making was that I think fitting the filter causes additional all light vignetting,

 

Probably I wasn't being clear. If at the same time as you add the filter you change the camera from ON to ON+UV/IR, you will see more overall vignetting because the correction that was done for overall vignetting in ON is missing in ON+UV/IR.

 

scott

[sean_reid]

This is were i think the Leica filter is better than the B+W which is also the case on the WATE. I have no cast on the WATE at all with the leica filter and with the B+W i did at 16, 18 and 21. As Sean and myself have pretty much said 21mm and below the Leica filer just handles the cyan cast better. But we need a 55mm leica filter to hit the streets too, so we wait

 

Right, the firmware is tuned to the Leica filters and they work better than the 486 filter with ultra-wides (at 21 mm and wider). Still need a 55 to test a 24 Elmarit but the 486 filters do fine with 28 mm and longer lenses.

 

Cheers,

 

Sean

[wlaidlaw]

Probably I wasn't being clear. If at the same time as you add the filter you change the camera from ON to ON+UV/IR, you will see more overall vignetting because the correction that was done for overall vignetting in ON is missing in ON+UV/IR.

 

scott

 

Scott,

 

I had not realised that. I had assumed they would increase the amplification of the red channel for the corners with UV/IR filter set to ON, to compensate for the cyan because it is the red that is being filtered out by the UV/IR filter. I had assumed wrongly that the firmware would still be correcting to the same extent for overall vignetting.

 

Wilson

[jaapv]

Maybe the next firmware update, Wilson. I presume that that would have been one variable too many to be able to get the cyan correction into our cameras within a reasonable span of time.

[Shootist]

Right, the firmware is tuned to the Leica filters and they work better than the 486 filter with ultra-wides (at 21 mm and wider). Still need a 55 to test a 24 Elmarit but the 486 filters do fine with 28 mm and longer lenses.

 

Cheers,

 

Sean

 

I have the 24mm ASPH and am using a B+W 486 55mm filter on it and haven't seen any cyan corners on it.

[Advertisement]

Hallo,

Look here: Leica M

[Guest guy_mancuso]

So Ed it cleans up pretty good at 24mm that is good to hear. I was hoping the B+W would work on this focal length at least until we can get a leica filter or not get one , nice to know this works