Chromatic Abberation

[zapp]

Advertisement (gone after registration)

We have discussed many aspects of Leica lenses here. One discussion, I could not find so far, is a chromatic aberration (longitudinal, color finging) discussion on M and R lenses for digial use. Schneider and Zeiss advertise with low CA. Can one get numbers/opinions for the Leica lenses? MTF, distortion, vignetting is in the official pdf's, bokeh is discussed in great detail here, well what about CA. How do the latest ASPH and APO lenses perform? Especially looking for opinions on R lenses - that should no be a problem since many of you mount those on Canon EOS cameras for quite some time. How will the 90AA from 2002 do compared to the 80Lux from 1980 considering that longitudinal CA was not an issue in the 80s ...

[Advertisement]

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

[lars_bergquist]

Quantifying chromatic aberration (transversal and longitudinal) is theoretically possible, but I have never found any numbers anywhere. Opinions there are, of course ...

 

Several recent M lenses do carry the APO designation of course. This was originally minted for microscope optics for visual observation, so the classical definition says only that the system should be corrected for three colours (normally blue, green and red) close to the optical axis, which is enough for eyeball work. Leica's definition is that it should do this across the field, more or less. So Leica's apo lenses are really more 'apo' than those from many other manufacturers, who have been pretty loose with the term.

 

Generally speaking, the effects of chromatic aberration are subsumed with those of the other aberrations when MTF measurements are made. They are simply not sorted separately, but they are of course more important with long than with short focal lengths. I do not think that there is any reason to worry specifically about them.

 

The old man from the Age of Anastigmats

[zapp]

Quantifying chromatic aberration (transversal and longitudinal) is theoretically possible, but I have never found any numbers anywhere. Opinions there are, of course ...

 

Several recent M lenses do carry the APO designation of course. This was originally minted for microscope optics for visual observation, so the classical definition says only that the system should be corrected for three colours (normally blue, green and red) close to the optical axis, which is enough for eyeball work. Leica's definition is that it should do this across the field, more or less. So Leica's apo lenses are really more 'apo' than those from many other manufacturers, who have been pretty loose with the term.

 

Generally speaking, the effects of chromatic aberration are subsumed with those of the other aberrations when MTF measurements are made. They are simply not sorted separately, but they are of course more important with long than with short focal lengths. I do not think that there is any reason to worry specifically about them.

 

The old man from the Age of Anastigmats

 

I appreciate your reply.

After buying and working with a lot of lenses, I stopped taking anything for granted when working in a digital workflow. I want to test it myself or study opinions of real users. A great film lens may not at all be a great lens for digital use - a lot of M8 users are confronted with problems not heared of (by them) before when film was used.

Several foto papers/testers as well as manufacterers provide numbers for CA. Well, who has hands on experience when it comes to CA on the M8, DMR, or Leica lenses on EOS ...

[jaapv]

I appreciate your reply.

After buying and working with a lot of lenses, I stopped taking anything for granted when working in a digital workflow. I want to test it myself or study opinions of real users. A great film lens may not at all be a great lens for digital use - a lot of M8 users are confronted with problems not heared of (by them) before when film was used.

Several foto papers/testers as well as manufacterers provide numbers for CA. Well, who has hands on experience when it comes to CA on the M8, DMR, or Leica lenses on EOS ...

 

Judging by the lack of interest for this thread I conclude that my opinion: A theoretical consideration without practical application in actual picture taking with Leica lenses is shared by many if not all Leica photographers. As a matter of fact the only CA thread we had here in relation to the M8 was about the Zeiss ZM Biogon 21/2.8. One poster reported somewhat bemused that he had found CA on this lens, I tested mine and and had to reply "egad, you're right" End of thread.....It is interesting to note that when I was still using my Canon DSLR and thus active on FM forums every third Canon lens thread seemed to be about CA, 100% crops and all. I draw my conclusions, not based on lab measurements.

[zapp]

Judging by the lack of interest for this thread I conclude that my opinion: A theoretical consideration without practical application in actual picture taking with Leica lenses is shared by many if not all Leica photographers. As a matter of fact the only CA thread we had here in relation to the M8 was about the Zeiss ZM Biogon 21/2.8. One poster reported somewhat bemused that he had found CA on this lens, I tested mine and and had to reply "egad, you're right" End of thread.....It is interesting to note that when I was still using my Canon DSLR and thus active on FM forums every third Canon lens thread seemed to be about CA, 100% crops and all. I draw my conclusions, not based on lab measurements.

Right now I would consider the lack of the discussion just ignorance and not a lack of CA in Leica lenses. The image below shows the corner of an image taken with a Canon EOS 1Ds and the 19 mm (latest version). Well, an excellent result for a FF camera with a 19 mm lens, but still a couple pixels of shift (enlarge to see). For my application this shift matters, and we are not talking about strong contrast edge effects in difficult lighting here. My Schneider Apo Digitar 47 mm shows less CA, the 72 mm Apo Digitar is superb, the effect vanishes with longer focal length.

(added a larger image)

[jaapv]

Right now I would consider the lack of the discussion just ignorance and not a lack of CA in Leica lenses. The image below shows the corner of an image taken with a Canon EOS 1Ds and the 19 mm (latest version). Well, an excellent result for a FF camera with a 19 mm lens, but still a couple pixels of shift (enlarge to see). For my application this shift matters, and we are not talking about strong contrast edge effects in difficult lighting here. My Schneider Apo Digitar 47 mm shows less CA, the 72 mm Apo Digitar is superb, the effect vanishes with longer focal length.

 

I did not say lack of CA in Leica lenses, any optical system will have that to some degree, however small or large. I said it is clearly of no interest to Leica lens users, for it is corrected well enough for the use Leica lenses are put to for 99% of applications. I think, given the level of this forum, that the implication of ignorance does not apply. Full frame sensors are of course notorious for all kinds of problems in the corners, including enhanced CA ,one of the reasons that Nikon, Leica etc. do not use them.

[zapp]

Advertisement (gone after registration)

I did not say lack of CA in Leica lenses, any optical system will have that to some degree, however small or large. I said it is clearly of no interest to Leica lens users, for it is corrected well enough for the use Leica lenses are put to for 99% of applications. I think, given the level of this forum, that the implication of ignorance does not apply. Full frame sensors are of course notorious for all kinds of problems in the corners, including enhanced CA ,one of the reasons that Nikon, Leica etc. do not use them.

Let me continue to be a little bit provocative. Zeiss did not introduce a digital RF camera. One reason they claimed years (1-2) ago was focus shift when stopping down the lenses. Leica users had to buy the M8 first to figure out focus shift as well as foreseeable IR trouble. The M8 is not a bad product, but was the system and were the users digitally ready?

While I think you are correct in principle that we can assume very high quality glass from Leica, I am looking for confirmation. Leica told us that film was the best medium as long as no digital solution was available from Leica. They want to sell products, and I want to make sure they are as good as claimed.

[jaapv]

Let me continue to be a little bit provocative. Zeiss did not introduce a digital RF camera. One reason they claimed years (1-2) ago was focus shift when stopping down the lenses. Leica users had to buy the M8 first to figure out focus shift as well as foreseeable IR trouble. The M8 is not a bad product, but was the system and were the users digitally ready?

While I think you are correct in principle that we can assume very high quality glass from Leica, I am looking for confirmation. Leica told us that film was the best medium as long as no digital solution was available from Leica. They want to sell products, and I want to make sure they are as good as claimed.

 

Zeiss is indeed a bit funny in their rationalizing. The main reason give is: "our customers demand a 24x36 sensor and such a sensor does not exist and will not exist in the forseeable future so we cannot build a digital RF" Ok. On this assumption of the mindset of their customers they appear to be left on the platform whilst Leica is sitting in the first class of the departing train.

Your quote on Leica is is not complete It dates back to 2004, and had the addendum at the present time. With the implication that they would implement digital technology as soon as they deemed it met their standards. And they did. The users were more than ready. As far as they were concerned sooner was better than later...

[manolo]

CA is a relevent topic, I would certainly like to know more about how lenses work and understand why images like this come to be. It was shot with the DMR and the 280/4 APO and a close up filter. (image and 2 crops)

[jaapv]

No way to tell about the lens from this. Closeup diopters produce strong CA in their own right. Have you got a shot -without close-up attachment - without filter?.(but with the proprietary filter in the back - it loses quality if that is removed) Only then can you judge the lens itself. This lens is probably the best corrected lens there is. It got 98 out of 100 points for correction in an old test in Fotomagazin, as I recall. So it can provide a benchmark.

[manolo]

I agree, no way to find CA in this lens alone but I think with the close up filter you get CA that is only restricted to one step (the filter) and maybe predictable and usefull.

 

The close up filters are advertised as "acromat" by Leica; I get that this means that they are corrected but not as much to get the APO name.

[jaapv]

I agree, no way to find CA in this lens alone but I think with the close up filter you get CA that is only restricted to one step (the filter) and maybe predictable and usefull.

 

The close up filters are advertised as "acromat" by Leica; I get that this means that they are corrected but not as much to get the APO name.

An achromat is corrected in two colours, an apochromat in three. Basically an achromat is two, maybe three lenses kitted to one element. It will introduce curved field, which makes them suitable, even desirable, for three dimensional objects, but less suited for reproductive macro, which entails photographing flat objects.

[manolo]

curved field

 

can you explain this? Is this for the colors or for the entire immage? os is this that the lens focuses at different distances in the center and in the corners progresively?

[jaapv]

No, this is not a colour aberration, it is a geometrical one. The plane of focus is bent, making the focus on the edges of the image closer than the centre. When photographing a flat object it is impossible to achieve correct focus over the whole image, when photographing a three-dimensional object it gives the appearance of a deeper DOF. To make things complicated, the CFA may be different for different spectral lengths, thus creating a kind of chromatic aberration, as different colours will focus in different planes towards the edges, turning it into a colour aberration again...

[manolo]

I find this when reversing a lens directly to the camera but I guess this is not always a bad thing. Also it is more when the lens is wide open. Does the close up filter also affect the contrast inherent in the lens?

[jaapv]

Not necessarily. Macro contrast is of course less important these digital days. Micro contrast and edge contrast are virtually certain to be degraded, but it is, with good quality diopters, unlikely to make any visible impact on the final photograph.And I did not even introduce astigmatism yet, nor the higher orders of aberrations:eek: . Bottom line - take the photograph and see if it meets your standards. The rest is interesting and even useful to help understand, but don't even begin to consider all the aberrations your eye has. Your lenses are miracles of precision in comparison.

[manolo]

so am I just throwind away money if I use all this incredible Leica glass (like the 280/4) with diopters or reversing them?

[jaapv]

No, because the quality of all the links in the chain counts.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[manolo]

Erfahrener, thank you and I now understand these optical equations a bit more