90mm Apo-Summicron-M not apochromatic?

[pedaes]

Advertisement (gone after registration)

41 minutes ago, LarsHP said:

This thread is about the lens, not how to fix its shortcomings. 

Haven't you done the lens to death and isn't it time to move on to solutions? (My last post in this thread).

[ianman]

2 hours ago, LarsHP said:

Most, if not all, lens reviews I have seen say it's trivial to remove LaCA and (more) difficult to remove LoCA.

I read this too. The solution given was to use focus stacking as mentioned many pages ago.

[Advertisement]

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

[lct]

There is an inexpensive 85/2 lens with good CA performance i've been told, the Jupiter-9. Sort of clone to the Sonnar 85/2. Rather soft at f/2 though and QC is not Russian lenses' forte but i have no experience with this lens. Now some CA (loca or laca?) can be reduced by stopping down. Would be interesting to see if the problem persists at f/2.8. Or to try an f/2.8 lens but my Leica M ones are hardly better than the 90/2 apo as far as CA is concerned. I have not my Tele-Elmarit 90/2.8 with me but i seem to recall that it does rather well in this respect. Otherwise good 90/4 lenses with low CA are not difficult to find. FWIW.

Edit: Now why not a Nikkor Z 85/1.8? I have no experience with it but it seems to have litle CA and very good optical performance overall. I would be tempted if i had a Nikon Z.

Edited May 23, 2021 by lct

[LarsHP]

58 minutes ago, ianman said:

I read this too. The solution given was to use focus stacking as mentioned many pages ago.

Thanks, but focus stacking will not help to remove bokeh fringing.

[LarsHP]

5 hours ago, farnz said:

Actually, no, it's the current definition of apochromatic.  If you're going to resort to arguing semantics, have you found out what Leica's definition of apochromatic is?  Surely that must be at the heart of your premise if (as per your thread title) you believe the 90/2 APO-Summicron asph is not an apochromatic lens?  Without Leica's own definition you have nothing to argue over.

This is a massive generalisation that requires supporting information before it could be taken seriously.

@jaapv will be able to confirm either way but I understood Jaap to be referring to the term "APO", rather than the term "apochromatic", which has an unequivocal definition.

 

Frankly I've lost sight of what you're trying to achieve through this thread so perhaps it's time I put the time I have spent in trying to help you, despite your out-of-hand rejection of everything I've written, to more beneficial use.

If you wish to believe that the 90/2 APO-Summicron asph doesn't deserve its status as apochromatic then it's no skin off my nose so, as our American friends say: "Knock yourself out".

Pete.

PS, no hostility intended.

PPSS, Since you've mentioned "expert opinion" several times along the way perhaps I should disclose that for the past 35 years I have been a professional Senior Design Engineer working as the optical subject matter expert for a company with 40,000+ employees in the optical fibre sector.  But I doubt that that'll make any difference to your trenchant position.

Thanks, Pete. I don't think I have seen anything resembling hostility from you.

Regarding the semantic of "apochromatic" I am relying on my time in the photography business and what I read from various sources that appear as having good or high credibility. This includes this forum as well as various other photography sites. I am not an optical engineer. As to your professional background, it certainly makes a difference to me. If you had made it clear that you have design of optics as your work and education earlier, I would have taken your input more serious than I have. In a forum like this it's not obvious what kind of knowledge people have just based on what they write.

There has been a lot of discussion about what the term apochromatic means in this thread. I didn't start the thread in order to argue about the semantics, but it appears to be unavoidable considering the thread title and subject. Instead, I'd like to ask you a question or two, given your background:

1) What is causing "bokeh fringing" (color aberrations is the out of focus areas)? Is that caused by longitudinal / axial CA? (This is the explanation I see in the sources I have been referring to.) If not, then what causes bokeh fringing?

2) If a flat subject with sharp black and white figures (like a test chart) is rendered without color fringing of any kind, can we then conclude that the lens performs as an apochromatic lens? (This is how I understand it, at least.)

3) In terms of lens designation, what distinguishes lenses like Voigtländer 65mm f/2 Apo-Lanthar and Zeiss 135mm Apo Sonnar which shows practically no color fringing in the out of focus areas from a lens like the 90mm Apo-Summicron-M? In other words, what would a meaningful distinction be, when calling a lens "apo" etc.?

Edited May 23, 2021 by LarsHP

[adan]

1 minute ago, lct said:

Now some CA (loca or laca?) can be reduced by stopping down.

LoCA can be, since it is basically a focus issue. Stopping down will shrink the "blur circles" of whatever wavelengths are "wandering" away from the prime focus plane. Remove/reduce the blur in red (or green or blue, or combinations (red + green = yellow)), and the blur-fringe also shrinks to invisibility.

LaCA won't change with stopping down, because it is a scaling or magnification issue - any more than stopping down a 28mm lens will give it the field of view of a 24mm or a 35mm lens.

However, LaCA is easier to fix in post - if one's software allows working on the individual RGB color channels (Photoshop does - not sure about LR or others). Just requires figuring out which color channel is "mis-sized" compared to the other two, and using Transform > Scale to enlarge or shrink it to align with the others.

I was "fixing" LaCA that way virtually as soon as I began scanning film in the 1990s - I had a Contax/Zeiss Sonnar 300mm MM with so much LaCA that, when switching between the scan color channels, objects with LaCA fringes would visibly "jump" left/right/up/down/diagonally several mm on my screen when switching between the R, G, and B color channels. Same when I first went digital, and got a Panasonic FX-10 for "tele" pictures alongside my Digilux 2. it produced substantial LaCA fringes at its longest focal length - "realigned" them in PS.

Here's that LaCA example I showed previously, with its fringes removed in PS by scaling down just the G/B color channels 3% or so, from the top-right corner. (note NEW fringes on the frame edge and black type - which are, of course, not a part of the lens image itself).

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!

There is not as easy a way to fix LoCA - unsharp masking is not powerful enough to adequately sharpen a blurred single color channel. In 2009, before Adobe added a "defringe" control, I messed around with "color noise reduction" and some local desaturating, to fix LoCA from a 75 Summilux wide-open. But it was a lot of work per picture.

Not that Adobe's defringe controls are that much better. It uses AI to find hard edges that have a fringe on one side, and desaturate it based on color and size setting. But it can easily get confused by, say, a red sign with a sharp edge, and start eating into the red of the sign, thinking it is a "red fringe."

The AI is not very "I".

It does work with "uncommon colors" like violet and magenta - right up until one takes a picture with a magenta or violet sign (or swimsuit, or building) in it.

[luigi bertolotti]

Advertisement (gone after registration)

This deep discussion has made me to consider this problem of color fringing on my only APO M lens... the Apo Telyt 135: so that I took it today (instead of a stronger tele... didn't want to carry also the 250 or 400... not a top day for quality of air/light)  for a trip next to my town, in a wet area where you have some chance of birdwatching.... here is one example (jpg OOC, AT 135 at 5,6 , M240)

Full frame

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!

100% crop - focused subject

200% crop, OOF area.

What about LoCA LaCA in your opinion ? I frankly can't decide about...

 

Edited May 23, 2021 by luigi bertolotti

[luigi bertolotti]

.. and another at a much closer distance (3 m around) , always the AT 135, between 4 and 5,6

FF

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!

2 crops (100 and 200%)  in OOF zones

Opinions welcome... (I love the lens, anyway...)

 

Edited May 23, 2021 by luigi bertolotti

[lct]

3 minutes ago, luigi bertolotti said:

What about LoCA LaCA in your opinion ? I frankly can't decide about...

Simple. If vegetation is green behind the duck it is laca err... loca . Just kidding. And saluting adan's efforts to make me understand what my brain cannot. 

[adan]

32 minutes ago, LarsHP said:

1) What is causing "bokeh fringing" (color aberrations is the out of focus areas)? Is that caused by longitudinal / axial CA? (This is the explanation I see in the sources I have been referring to.) If not, then what causes bokeh fringing?

2) If a flat subject with sharp black and white figures (like a test chart) is rendered without color fringing of any kind, can we then conclude that the lens performs as an apochromatic lens? (This is how I understand it, at least.)

1) It is not caused by LoCA, because LoCA is only "defined" in the plane of focus. It is caused by the same core attribute of light passing through air/glass surfaces, that can also cause LoCA in the plane of focus. Which is dispersion. LoCA is an effect of dispersion in the plane of focus.

I.E. LoCA (or Apochromatic performance) is a subset of dispersion effects, limited to the case where the image is otherwise sharp and focused.

If Newtonian dispersion of colors occurs anywhere other than a plane of focus, it takes a different name. Such as "spectrum," "rainbow," or if one likes, "bokeh color fringes."

It is rather like testing a lens's MTF (resolution x contrast). Which is calculated or measured only where the image is correctly focused. The contrast and resolution of an out-of-focus picture cannot be called its MTF.

2) If the chart is actually flat (which means parallel to the sensor/film plane, and not tilted in 3D space, back to front or side-to-side, and thus equally focused from corner to corner) - then in that case, YES, we can conclude that no apparent color fringes on the black/white edges very likely means the lens is apochromatic (and also has no LaCA), for all practical purposes.

(Sorry about all those underlined qualifiers - but that is how scientific (and in my view, any rational) analysis has to be done.)

To be absolutely certain, you'd have to get down to the scale of individual photons of different frequencies/wavelengths, and study where they are going. But at that scale, diffraction and quantum mechanics and "spooky action at a distance" and Heisenberg's Uncertainty Principle all come into play.

https://en.wikipedia.org/wiki/Action_at_a_distance

[adan]

Luigi - there appears to be a small amount of "bokeh color fringing" in the monochromatic bare branches in the background ofthe duck picture. The green leaves, of course, are poor test subjects, becasue they inherently produce green blurs.

There appear to be some color speckles in the duck, but those are probably aliasing or de-mosaicing artifacts from the Bayer color sensor.

Here is a Euro's-worth of free advice for everyone - write it down in your notebooks 100 times.

A consumer digital camera in the hands of an amateur scientist is a lousy, crummy, unreliable laboratory tool.
A consumer digital camera in the hands of an amateur scientist is a lousy, crummy, unreliable laboratory tool.
A consumer digital camera in the hands of an amateur scientist is a lousy, crummy, unreliable laboratory tool.
A consumer digital camera in the hands of an amateur scientist is a lousy, crummy, unreliable laboratory tool.
.
.
.
.
.

 

[Robert Blanko]

Luckily, I rarely take pictures of ducks with the 135 APO. And if so, Heisenberg‘s cat would for sure prevent any LoCa and LaCa with my Leica.

To be on the safe side, I prefer the 50 summilux which is reportedly an APO lens - however without the APO in its name, so that no wrong expectations can be violated. 😉

[luigi bertolotti]

1 hour ago, adan said:

Luigi - there appears to be a small amount of "bokeh color fringing" in the monochromatic bare branches in the background ofthe duck picture. The green leaves, of course, are poor test subjects, becasue they inherently produce green blurs.

There appear to be some color speckles in the duck, but those are probably aliasing or de-mosaicing artifacts from the Bayer color sensor.

 

Me too have the feel of a small amount of vague purplish fringing around the OOF branches...  and the OOF leaves (2nd pic 2nd post) probably haven't a bit of magenta component which can produce fringing... but probably have a slight blue one... and (at least on my monitor) seems to me to perceive a certain fringe towards the blue...

I did notice the speckles on bird's plumage... frankly I thought they were small sunrays reflections on waterdrops - the bird had just came out from underwater and sun was in front of me 

 thank you for your observations.

(globally, I think we are in the hair/splitting... 🙄)

Edited May 23, 2021 by luigi bertolotti

[pgk]

We all too easily get side-tracked and diverted from the 'true path' which is of course; 'does a lens take photographs that we are satisfied with?'. All mine do!

[ianman]

13 hours ago, adan said:

there appears to be a small amount of "bokeh color fringing" in the monochromatic bare branches in the background ofthe duck picture.

They seem to be more like quantumly unaligned colour kerfuffles, otherwise known as QUACK.

[pedaes]

9 minutes ago, ianman said:

They seem to be more like quantumly unaligned colour kerfuffles, otherwise known as QUACK.

That's a quacker!

[LarsHP]

15 hours ago, adan said:

1) It is not caused by LoCA, because LoCA is only "defined" in the plane of focus. It is caused by the same core attribute of light passing through air/glass surfaces, that can also cause LoCA in the plane of focus. Which is dispersion. LoCA is an effect of dispersion in the plane of focus.

I.E. LoCA (or Apochromatic performance) is a subset of dispersion effects, limited to the case where the image is otherwise sharp and focused.

If Newtonian dispersion of colors occurs anywhere other than a plane of focus, it takes a different name. Such as "spectrum," "rainbow," or if one likes, "bokeh color fringes."

It is rather like testing a lens's MTF (resolution x contrast). Which is calculated or measured only where the image is correctly focused. The contrast and resolution of an out-of-focus picture cannot be called its MTF.

2) If the chart is actually flat (which means parallel to the sensor/film plane, and not tilted in 3D space, back to front or side-to-side, and thus equally focused from corner to corner) - then in that case, YES, we can conclude that no apparent color fringes on the black/white edges very likely means the lens is apochromatic (and also has no LaCA), for all practical purposes.

(Sorry about all those underlined qualifiers - but that is how scientific (and in my view, any rational) analysis has to be done.)

To be absolutely certain, you'd have to get down to the scale of individual photons of different frequencies/wavelengths, and study where they are going. But at that scale, diffraction and quantum mechanics and "spooky action at a distance" and Heisenberg's Uncertainty Principle all come into play.

https://en.wikipedia.org/wiki/Action_at_a_distance

Thanks for yet another contribution to this thread. 

Regarding your answer to question #2, I am happy to know that we agree (including the qualifiers). To me, this also suggests that among modern quality lenses most (if not all) can be classified as apochromatic even though the manufacturer doesn't use that term. Sigma comes to mind: Their EX series had some with the apo label, but their current top of the line lenses that outperform the older designs (as well as most of the current competition) are "just" designated with the Art label. In other words, they dropped the apo designation for some reason which certainly isn't because the Art series lenses are inferior to the older EX Apo lenses. 

Regarding your answer to question #1, I hope I understand your reply correctly when putting it this way:

A) LoCA is a term only related to the plane of focus. (This has been repeated countless times here.)

B ) Bokeh fringing (color aberrations in the out of focus areas) is caused by the same optical phenomenon as LoCA, but it's not called LoCA since it's not happening in the plane of focus. 

Am I right in the latter point? If so, then I will be happy for that too! 

Finally, I edited my previous post and added one more question that you probably didn't see, so if possible, I'd like you to relate to that.

3) In terms of lens designation, what distinguishes lenses like Voigtländer 65mm f/2 Apo-Lanthar and Zeiss 135mm Apo Sonnar which shows practically no color fringing in the out of focus areas from a lens like the 90mm Apo-Summicron-M? In other words, what would a meaningful distinction be, when calling a lens "apo" etc.?

What I am after here is understanding if the bokeh fringing is telling us something about LoCA in the in-focus area. My question would then be:

Is the absence of bokeh fringing a sign of extremely precise focus of all three color channels in the plane of focus? (This seems to be the underlying understanding I get in the sources that I have referred to.) 

Edited May 24, 2021 by LarsHP

[LarsHP]

2 hours ago, pgk said:

We all too easily get side-tracked and diverted from the 'true path' which is of course; 'does a lens take photographs that we are satisfied with?'. All mine do!

Good for you! 😊 

Hopefully I have made it clear that I am not complaining about the M 90 Apo from a purely technical or theoretical viewpoint. It's directly derived from images I have taken and posted samples of here. Infinity performance is stellar, as I have noted more than once here.

[pgk]

3 hours ago, LarsHP said:

I edited my previous post and added one more question .....

If I might say so, I would say that the fundamental problem in trying to assess what is caused by what, is the desire for simple answers to what are probably complex problems. I had a Canon 24/1.4L (so a 'good' lens) which exhibited similar and significant fringing to that which you describe. I looked into the possible causes and finally gave up. There are some which seem to be an obvious cause but then again there are others which may well be contributing too - especially on digital cameras. So I finally decided that it was, as usual, probably a result of several different causes interacting and reinforcing. The shape and colour of noof areas are caused by many things; the lens design itself, the diaphragm shape, the intensity, shape and location of the illuminated areas and overlays of these, and to all these and more add to the sensor's design and so on. Tempting though it is to pin a problem down to one root cause, I don't think that this is easy to do. Diffuse oof effects due to  and in any case aberrations show in the image plane because the oof areas are not actually imaged as such. So aything which could be caused by 'aberrations' is really difficult to tie down, because so much will be going on. What may seen logical and straightforward is in reality almost certainly the result of complex interactions.

[Advertisement]

Hallo,

Look here: Leica M

[adan]

4 hours ago, LarsHP said:

if possible, I'd like you to relate to that.

3) In terms of lens designation, what distinguishes lenses like Voigtländer 65mm f/2 Apo-Lanthar and Zeiss 135mm Apo Sonnar which shows practically no color fringing in the out of focus areas from a lens like the 90mm Apo-Summicron-M? In other words, what would a meaningful distinction be, when calling a lens "apo" etc.?

What I am after here is understanding if the bokeh fringing is telling us something about LoCA in the in-focus area. My question would then be:

Is the absence of bokeh fringing a sign of extremely precise focus of all three color channels in the plane of focus? (This seems to be the underlying understanding I get in the sources that I have referred to.)

Well, I like ianman's designation - QUACK.

More seriously, a hypothesis (untested theory) concerning the 65mm APO-Lanthar or 135 APO-Sonnar or other lenses that may not show color fringing in the out of focus areas.

For their specs, they are rather large, long lenses with a lot of elements. They may have adopted some advances in optical intent from non-photo lenses. Such as telecentricity, which projects an orthographic (without distance/size perspective) image. And is more "friendly" to the 3-dimensional pits and valleys of digital sensors overall (without regard to color or fringes as such).

https://www.edmundoptics.eu/knowledge-center/application-notes/imaging/advantages-of-telecentricity/

https://en.wikipedia.org/wiki/Orthographic_projection

https://www.closeuphotography.com/voigtlander-apo-lanthar-65-mm-f2-macro-lens-test

https://www.cameralabs.com/zeiss-apo-sonnar-135mm-f2-review/2/

But - we can also note how many low-dispersion elements each contains - 4 to 6 (the 90mm ASMA has only 5 elements altogether!). That itself may make some difference.

(future research question: do the recent - and very large - Leica 75mm f/1.25 ASMA (9 elements) and 90mm f/1.5 ASMA (8 elements) eliminate or subdue "color bokeh fringes?")

A classic apochromat focuses three colors in one plane. Outside that plane the colors are convergent or divergent (not parallel) and thus can create color fringes.

It may be possible to use additional glass elements to achieve something similar to telecentricity or orthagonal projection, as regards divergent color rays (as opposed to diverging lines or sizes/shapes) - restore the original parallel color light rays to parallel before they finally leave the lens.

It would not have to be complete or perfect, since the blurring of the image itself will hide residual misalignments (fringes) if they are simply reduced enough.

This (rough) schematic suggests a hypothetical orthapochromatic lens (light rays corrected ("orth" or "ortho") to parallel across all planes, by additional bending). I don't know if it is actually possible, or whether it might introduce other artifacts to the background blurs (!!!!). And it would be larger than a "simple" apochromat.

Whether or not this is actually how Zeiss and C/V do it, "orthapochromat" ("corrected" apochromat) might still be a reasonable name for such a lens, and "orthapochromatic" for such correction. Leica versions: OAPO-Summicron or OAPO-Summilux. And of course one might then create an "orthosuperachromat" as well. But those are, for the time being, "unofficial" designations.

[edit]: It is - interesting - that C/V's logo for their APO Lanthar lenses is red/green/blue lines in parallel.   /  /  / 

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!

 

Edited May 24, 2021 by adan