90mm Apo-Summicron-M not apochromatic?

[LarsHP]

Advertisement (gone after registration)

I have owned my 90mm Apo-Summicron-M for about nine months now and have mixed feelings about it. On the positive side, it is really sharp at long and semi-long distances and shooting a landscape where everything in the image is near infinity, it performs impressively. For that reason I don't hesitate shooting it wide open when it's dark, and I don't have a tripod around. It also shows very little field curvature and zero LaCA's, so for those infinity shots, it's stellar. However, at shorter distances (about 2 meters and less) the sharpness decreases more than I like and LoCA will show up.

Yesterday I took the below image. Focus is on the tree several meters away and the background is several hundred meters and several kilometers away respectively. It was shot wide open to get that depth of field impact. When I opened the image on my computer, I was shocked to see the amount of longitudinal chromatic aberrations (LoCA) in the background. Lots and lots of green and purple too! To me, this is unacceptable - particularly for a lens with the "apo" designation. The lens is obviously far from apochromatic when talking LoCA correction.

Is this just how the lens is? Or is my sample much worse than a good copy of the lens should be?

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!

[Advertisement]

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

[leicameech]

Because a lens is an apochromatic design does not mean it will have no chromatic aberration under any circumstances. The image you've shared (which is cool, by the way) is about as challenging of a scene for any lens as it gets for mitigating the undesirable effect. Looking at it, as a matter of fact, I first thought "wow, that's pretty impressive." Have you shot the same scene with any other M lenses (if you brought any others with you that day?). I suspect my 50 Summilux ASPH would not have done so well wide-open, and perhaps even at f/2. At the 90mm focal length, the only lens I can imagine would do better than this would probably be one of the SL lenses, like the 24-90 @ 90 (which wouldn't be fair because that would be at f/4), th 90-280 (also an APO lens, but huge, and still at f/2.8 when at 90mm), and the 90 APO-SL, which would be the best of them all, I'd suspect. 

To answer your question, I think that this is just how the lens is, and not sample variation. I don't own the lens, though. Would be curious to hear from owners of the lens. As I said before, though, I think your image is a testament to how good the lens is and not how it is lacking. Could just be me..

[UliWer]

I am not really sure where to look at the problem you describe: I don't see much - wrong - green and purple on your example: The branches in the foreground as well as the little bushes in the range behind the focal plane don't look bad to me and so does the tree in focus. 

Perhaps you could show some crops of regions where you see "lots and lots of green and purple too" which you regard as totally unacceptable.           

[vanGeist]

vor 20 Minuten schrieb leicameech:

(…) As I said before, though, I think your image is a testament to how good the lens is and not how it is lacking. Could just be me..

Very well said. Most other lenses would have been worse, even most of the best Leica ones.

[LarsHP]

1 hour ago, UliWer said:

I am not really sure where to look at the problem you describe: I don't see much - wrong - green and purple on your example: The branches in the foreground as well as the little bushes in the range behind the focal plane don't look bad to me and so does the tree in focus. 

Perhaps you could show some crops of regions where you see "lots and lots of green and purple too" which you regard as totally unacceptable.           

Right. I will post a crop tomorrow. 

The LoCA is mosly seen in the plains, but also in the island in the background. There is just a touch in the branches, which is acceptably weak and easily fixed in editing, but the snow-earth transitions in the far background has loads of green and purple. I even used a Zeiss UV filter which cuts sharply at 410nm, so there's no UV light "pollution" causing purple fringing. 

[LarsHP]

55 minutes ago, vanGeist said:

Very well said. Most other lenses would have been worse, even most of the best Leica ones.

Without having tested it properly, I think my 90mm Elmarit-M seems to be on the same level as the Apo-Summicron-M regarding LoCA. My Nokton 50mm f/1.2 Asph VM also seems to be comparable at f/2. I actually took a shot of the scene with the Nokton 50, so I will share that for reference. In other words, I don't think most other lenses would do worse. 

[lct]

Advertisement (gone after registration)

1 hour ago, leicameech said:

Because a lens is an apochromatic design does not mean it will have no chromatic aberration under any circumstances.

+1. I could show much worse purple fringing out of the Summicron 50/2 apo but i don't want to demoralize the OP . Just kidding but true.

[ianman]

The only issue I have with this lens is the onion rings I noticed once under a particular set of circumstances which were unusual to my style. That being said I don’t pixel peep and hardly ever use it at f2

[lct]

Great lens indeed, even at close distance contrary to what i read here and there. Better use an EVF to focus it though. Here at about 0.5m on digital CL with Macro-Adapter-M v2 :

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!

 

[farnz]

3 hours ago, LarsHP said:

The LoCA is mosly seen in the plains, but also in the island in the background. There is just a touch in the branches, which is acceptably weak and easily fixed in editing, but the snow-earth transitions in the far background has loads of green and purple. I even used a Zeiss UV filter which cuts sharply at 410nm, so there's no UV light "pollution" causing purple fringing.

Apochromatic lenses are designed to precisely overlay three wavelengths (usually in the red, green, and blue wavebands) in the plane of focus.  But the plains and island that you mention are in out of focus areas where it's not reasonable to expect the three wavelengths to coincide perfectly owing to chromatic dispersion, ie the difference in the velocity of propagation of different wavelengths through glass.  The three colours will travel through the lens differently so when the rays pass the plane of focus they will spread differently and not coincide precisely in the out of focus areas.

Pete.

[LarsHP]

9 hours ago, lct said:

Great lens indeed, even at close distance contrary to what i read here and there. Better use an EVF to focus it though. Here at about 0.5m on digital CL with Macro-Adapter-M v2 :

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!

 

It is not about missing focus properly. And I am of course talking about sharpness wide open. 

After receiving the lens I was doing some tests in order to learn how the lens behaves. My tests showed that it stays reasonably sharp in the center near the minimum focus distance (but still not as well as at longer distances) while getting pretty bad in the sides and corners. Using a crop format camera with this lens for close focus shots will mitigate this issue of course.

I took some shots at 1 meter and about 2 meters. Even at 2 meters the lens became sharper in the sides when I added an achromatic diopter (Tokina AT-X +0.4). This means that when the lens is focused much further away, it performs better (including by adding a heresy Tokina "close focus filter" on the godly Leica Apo lens 😆 ).

[LarsHP]

7 hours ago, farnz said:

Apochromatic lenses are designed to precisely overlay three wavelengths (usually in the red, green, and blue wavebands) in the plane of focus.  But the plains and island that you mention are in out of focus areas where it's not reasonable to expect the three wavelengths to coincide perfectly owing to chromatic dispersion, ie the difference in the velocity of propagation of different wavelengths through glass.  The three colours will travel through the lens differently so when the rays pass the plane of focus they will spread differently and not coincide precisely in the out of focus areas.

Pete.

You are of course totally right that the apo designation is about the point of focus. However, the apochromatic correction usually translate into out of focus areas that are free of color fringing (LoCA) or at least has greatly reduced it. To me, that's the sign of a proper apochromatic lens.

My Voigtländer 180mm Apo-Lanthar SL and Coastal Optical UV-VIS-IR 60mm Apo Macro performs like that. They are both f/4 lenses though. Still, the Voigtländer 65mm f/2 Apo-Lanthar and Nikon AF-S Nikkor 200mm f/2 VR are just as fast as the 90mm Apo-Summicron-M and they don't have any color fringing in the out of focus areas = very well controlled LoCA.

So, to me it was a considerable disappointment to see the amount of green and purple in the out of focus area in the image I just posted. I was aware that the 90mm Apo-Summicron-M wasn't perfect regarding LoCA, but this was just so much that I couldn't believe that Leica had the guts to call the lens apochromatic. If Leica just called it "aspherical", then I wouldn't be that disappointed. 

[pedaes]

35 minutes ago, LarsHP said:

 to me it was a considerable disappointment

Sell it and move on. It will never meet your expectations and standard. Maybe stick with maximum aperture of F4 in those lenses deliver for you.

[LarsHP]

7 minutes ago, pedaes said:

Sell it and move on. It will never meet your expectations and standard. Maybe stick with maximum aperture of F4 in those lenses deliver for you.

I actually do consider selling the 90mm Apo-Summicron-M, but I see no reason to stick with f/4 lenses. As mentioned, the 65mm f/2 Macro Apo-Lanthar and Nikkor 200mm f/2 perform beautifully, including regarding LoCA. 

[lct]

34 minutes ago, LarsHP said:

they don't have any color fringing in the out of focus areas

Suffice it to overexpose and you will see fringing with any lens in my modest experience with M 50/2 apo, 75/2 apo,  90/2 apo as well as R 180/3.4 apo and 280/4 apo. A good apo lens is a lens showing less fringing than others but zero fringing i have never seen that so far. FWIW.

[MrFriendly]

2 minutes ago, LarsHP said:

I actually do consider selling the 90mm Apo-Summicron-M.

How does the Summilux-M 90mm f/1.5 compare in this regard (LoCA)?  Is it APO?

[LarsHP]

Just now, MrFriendly said:

How does the Summilux-M 90mm f/1.5 compare in this regard (LoCA)?  Is it APO?

I have no idea. Haven't looked at it. 

[farnz]

1 hour ago, LarsHP said:

I was aware that the 90mm Apo-Summicron-M wasn't perfect regarding LoCA, but this was just so much that I couldn't believe that Leica had the guts to call the lens apochromatic.

I've explained that apochromaticity only relates to the plane of focus, which you accepted, and you didn't complain about LoCA around the elements in your image that are in focus so I'm left to assume there was none.  Therefore Leica is thoroughly entitled to describe it as an apochromatic lens firstly because it shows no LoCA in the plane of focus and secondly because the lens includes low-dispersion lens elements for the purpose of correcting for chromatic aberrations.

LoCA in out of focus areas is explained in my previous post.  The fact that your Voigtlander and Nikon lenses don't show colour fringing in out of focus areas is irrelevant because Voigtlander's and Nikon's design propositions, processes, and objectives are proprietary to them and inevitably different from Leica's so it makes little sense to compare them.  Please note that the 90/2 APO-Summicron asph was first released in 1998 so its design will be 24 or 25 years old whereas (I assume) that the design of your Voigtlander and Nikon lenses are much later and therefore in a position to benefit from knowledge gained and modern material science that wasn't available when the Summicron was designed so it seems pointless to me to draw comparison. 

To illustrate why  'all bets are off' with respect to LoCA in out of focus areas please note the diagram below showing corrected LoCA (credit: vision-doctor.com) where the rays clearly diverge after the plane of focus so any 'coincidence  errors' in wavelengths will be exaggerated the further the rays travel.

Pete.

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!

[Advertisement]

Hallo,

Look here: Leica M

[LarsHP]

22 minutes ago, farnz said:

I've explained that apochromaticity only relates to the plane of focus, which you accepted, and you didn't complain about LoCA around the elements in your image that are in focus so I'm left to assume there was none.  Therefore Leica is thoroughly entitled to describe it as an apochromatic lens firstly because it shows no LoCA in the plane of focus and secondly because the lens includes low-dispersion lens elements for the purpose of correcting for chromatic aberrations.

LoCA in out of focus areas is explained in my previous post.  The fact that your Voigtlander and Nikon lenses don't show colour fringing in out of focus areas is irrelevant because Voigtlander's and Nikon's design propositions, processes, and objectives are proprietary to them and inevitably different from Leica's so it makes little sense to compare them.  Please note that the 90/2 APO-Summicron asph was first released in 1998 so its design will be 24 or 25 years old whereas (I assume) that the design of your Voigtlander and Nikon lenses are much later and therefore in a position to benefit from knowledge gained and modern material science that wasn't available when the Summicron was designed so it seems pointless to me to draw comparison. 

To illustrate why  'all bets are off' with respect to LoCA in out of focus areas please note the diagram below showing corrected LoCA (credit: vision-doctor.com) where the rays clearly diverge after the plane of focus so any 'coincidence  errors' in wavelengths will be exaggerated the further the rays travel.

Pete.

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!

LoCA in areas that are in focus is low in the 90 Apo, but there is some. My point is that the lens performs below the expected standard regarding LoCA when comparing it to other apo lenses. 

The Voigtländer 180mm f/4 and Nikon 200mm f/2 lenses I refer to are not much never than the 90mm Apo. They were introduced in 2003 and 2004 respectively. The 65mm Apo-Lanthar is quite recent though. 

Edited May 16, 2021 by LarsHP

[farnz]

11 minutes ago, LarsHP said:

My point is that the lens performs below the expected standard regarding LoCA

I think this is where your problem lies: expectation.  There is no formal standard for LoCA corrected performance in lenses so it's a subjective 'performance standard' that you've created based on subjective performance of other unconnected lenses.

Clearly from your opening post the lens exhibits some performance features that appeal to you so the lens has value to you.

If the LoCA performance is unacceptable to you then you should send the lens to Leica with details of what you're not happy with backed up by pictures.  You might not want to be without the lens for a period of time and you might not wish to pay the price of shipping, inspection, and any adjustment that might be required - that is your decision.  Alternatively, you might decide that it is better to save your money and fix any unacceptable LoCA in post-processing.

Whichever you choose I wish you luck.

Pete.