90 Apo Purple Fringing

[Csacwp]

Advertisement (gone after registration)

I did some test shooting of the 90 Apo today.  I'm very impressed by the resolving power it has wide open, and the colors are pleasant.  One thing I noticed is that I am getting a ton of purple fringing shooting wide open.  It appears on the edges of in-focus and out-of-focus areas.  Is the 90 Apo known to have a lot of fringing wide open?  I should add that I shot it with a B+W MRC Nano UV filter attached.  Could the filter be causing the purple fringing?

[happymac]

Some pictures showing the phenomenon would be appreciated. ;-)

[Advertisement]

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

[Csacwp]

Shoot me a private message with your email and I'll send you a few DNG's via Wetransfer.

[jip]

Strange, APO lenses should not exhibit this. I have 2 APO lenses and no matter what I do I can't get them to show CA. 

 

P.S.

120mm APO-Makro-Planer for C645 on Leica S

180mm APO-Tele-Elmar-S for Leica S

[Csacwp]

I had heard somewhere that the early APO lenses are corrected for one type of CA but not longitudinal CA, which is the type that shows up as purple fringing.  Is that correct?

[ZHNL]

Yep, that is what 90APO does. Good news is your lens behave normal. Bad news is 90APO is not first grade APO lens. However at infinity, it is still way better than 90summarit on both PF and sharpness and it is still better than most sharp short tele lens out there in term of PF.

 

Jip, the two you mentioned are might few best APO out there. I have them too. I don't shoot 180S enough but C120 is one of best in color correction.

[Csacwp]

Advertisement (gone after registration)

Wide open it seems to be a huge improvement over the Pre-ASPH 90 Summicron in terms of sharpness, and the colors are better all around.  As for the poor performance I hear about at close distances, it seems to be a product of poor focusing.  Every once in a while I was able to get an incredibly sharp image at minimum focusing distance, but normally the images are lacking.  

[ZHNL]

Not true. About close focus performance. Unless we have different standard about close focus performance. Do you have experience of modern zeiss glass like 135APO, OTUS or latest S glass? Check those, the difference is not small. It is simply dull by comparison.

[adan]

The thing about "purple fringing" is that it has many causes, often overlapping and simultaneous in a given lens/situation. And it also occurs in different places, for different reasons.

 

Purple fringing (in its various forms) can be caused by: 1) longitudinal chromatic aberration; 2) lateral chromatic aberration; and 3) the fact that blue/purple/violet light is more energetic than green or red light, and thus can spread farther before dissipating to unnoticeable levels. In addition, in digital pictures, the very structure of a digital sensor (AA filter, IR filter, cover glass, Bayer filter, microlenses) can also produce or exaggerate color fringes, that may or may not have anything to do with the "camera" lens.

 

The chromatic aberrations are simply examples of Newton's discovery that light passing through glass at an angle (as it does in any lens element) gets broken up into a spectrum, with different colors travelling in different paths, unless corrected.

 

https://i.pinimg.com/originals/02/3c/d8/023cd830add548d3e11fe1deeacda85d.jpg

 

1) Longitudinal CA occurs when a lens focuses the different primary colors in different planes. This will result in concentric, non-directional color fringes around light or dark pinpoint details, resulting from, for example, a sharp image in green being overlaid with fuzzy blurs of red and blue (a purple fringe around the outside).

 

An APOchromatically-corrected lens will eliminate or greatly reduce longitudinal CA and its associated color fringes around objects - in the plane of focus. Sharp points will not have fringes. HOWEVER, the definition of APOchromatic correction does not apply to defocused areas of the picture. Take a look at this diagram of what an APO lens does - it focuses light broken up into red, green, and blue rays into one plane. But look either side of the plane of focus, left or right, and you will see that the color rays are not tracking together except at the plane of focus.

 

http://www.subclub.org/fujinon/close-ups.gif

 

An APO lens may easily produce color fringes in fuzzy backgrounds or foregrounds. The only place that counts, in optical engineering, is what it does with the sharply-focused parts of the picture. If it focuses three primary color at the same place, it is APO. Even if the fuzzy backgrounds and foregrounds show color fringes.

 

BTW - since long CA is literally a problem of focus differences in red, green, and blue light, it is not easy to correct digitally. Very hard to shrink a purple blur circle of 5 pixels width back into a single pixel.

 

2) Lateral chromatic aberration occurs when a lens focuses all colors in the same plane, but the images for red, green and blue are not the same size.  (i.e. a lens is 21mm in green, and 21.05mm for red light, and perhaps 20.92mm for blue/violet light. It can be identified because it results in TWO fringes of complementary colors (one of which may be purple, or not), on opposite sides of the subject, and with increasing severity getting closer to the picture corners.

 

This is lateral CA, with complementary colored fringes of cyan-green and red-purple: https://cdn.photographylife.com/wp-content/uploads/2011/11/Uncorrected-and-Corrected-CA.jpg

 

Lateral CA is most common in wide-angle lenses, due to the amount of bending of light rays needed (Newton's prism again). It is rather easy to correct digitally, since unlike longitudinal CA, there is no focus problem, just a scaling problem. "CA removal" algorithms simply scale the mis-sized color channels to all overlap exactly - and no more fringes.

 

3) The higher energy of blue/purple light means it will "spread more" in the case of over-exposure. Especially if the light is already bluish and bright (e.g. sky). If there is also some CA, it will exaggerate that fringing as well - but it can cause fringing even with an APO lens.

 

https://upload.wikimedia.org/wikipedia/commons/c/c1/Purple_fringing.jpg

 

Finally, digital photography is more prone to revealing or producing color fringes than film. It's easier to "blow" one color channel first, with overexposure, which will produce a fringe. And the higher the lens's overall contrast (read - 90mm APO), the easier it is to blow highlights - the highlight brightness is not being dispersed over the whole picture as veiling flare, but hammering only in one spot.

 

https://mjambon.github.io/mjambon2016/purple-fringe/tree.jpg

 

One really has to look carefully at any purple fringing in detail, to determine what the primary cause is in any given case.

[jaapv]

On Apo lenses: As Andy explains, they focus three (or four) wavelengths across the spectrum in one plane. But they do not have to focus the wavelengths inbetween in those three or four in the same plane; In fact, they usually don't. The amount of aberration between the key colours is reduced compared to a non-Apo lens, that is true, but it is certainly not zero. The best we can say of an Apo lens is that it has reduced chromatic aberration, but there will still be a residue present.

[ZHNL]

The thing about "purple fringing" is that it has many causes, often overlapping and simultaneous in a given lens/situation. And it also occurs in different places, for different reasons.

 

Purple fringing (in its various forms) can be caused by: 1) longitudinal chromatic aberration; 2) lateral chromatic aberration; and 3) the fact that blue/purple/violet light is more energetic than green or red light, and thus can spread farther before dissipating to unnoticeable levels. In addition, in digital pictures, the very structure of a digital sensor (AA filter, IR filter, cover glass, Bayer filter, microlenses) can also produce or exaggerate color fringes, that may or may not have anything to do with the "camera" lens.

 

The chromatic aberrations are simply examples of Newton's discovery that light passing through glass at an angle (as it does in any lens element) gets broken up into a spectrum, with different colors travelling in different paths, unless corrected.

 

https://i.pinimg.com/originals/02/3c/d8/023cd830add548d3e11fe1deeacda85d.jpg

 

1) Longitudinal CA occurs when a lens focuses the different primary colors in different planes. This will result in concentric, non-directional color fringes around light or dark pinpoint details, resulting from, for example, a sharp image in green being overlaid with fuzzy blurs of red and blue (a purple fringe around the outside).

 

An APOchromatically-corrected lens will eliminate or greatly reduce longitudinal CA and its associated color fringes around objects - in the plane of focus. Sharp points will not have fringes. HOWEVER, the definition of APOchromatic correction does not apply to defocused areas of the picture. Take a look at this diagram of what an APO lens does - it focuses light broken up into red, green, and blue rays into one plane. But look either side of the plane of focus, left or right, and you will see that the color rays are not tracking together except at the plane of focus.

 

http://www.subclub.org/fujinon/close-ups.gif

 

An APO lens may easily produce color fringes in fuzzy backgrounds or foregrounds. The only place that counts, in optical engineering, is what it does with the sharply-focused parts of the picture. If it focuses three primary color at the same place, it is APO. Even if the fuzzy backgrounds and foregrounds show color fringes.

 

BTW - since long CA is literally a problem of focus differences in red, green, and blue light, it is not easy to correct digitally. Very hard to shrink a purple blur circle of 5 pixels width back into a single pixel.

 

2) Lateral chromatic aberration occurs when a lens focuses all colors in the same plane, but the images for red, green and blue are not the same size. (i.e. a lens is 21mm in green, and 21.05mm for red light, and perhaps 20.92mm for blue/violet light. It can be identified because it results in TWO fringes of complementary colors (one of which may be purple, or not), on opposite sides of the subject, and with increasing severity getting closer to the picture corners.

 

This is lateral CA, with complementary colored fringes of cyan-green and red-purple: https://cdn.photographylife.com/wp-content/uploads/2011/11/Uncorrected-and-Corrected-CA.jpg

 

Lateral CA is most common in wide-angle lenses, due to the amount of bending of light rays needed (Newton's prism again). It is rather easy to correct digitally, since unlike longitudinal CA, there is no focus problem, just a scaling problem. "CA removal" algorithms simply scale the mis-sized color channels to all overlap exactly - and no more fringes.

 

3) The higher energy of blue/purple light means it will "spread more" in the case of over-exposure. Especially if the light is already bluish and bright (e.g. sky). If there is also some CA, it will exaggerate that fringing as well - but it can cause fringing even with an APO lens.

 

https://upload.wikimedia.org/wikipedia/commons/c/c1/Purple_fringing.jpg

 

Finally, digital photography is more prone to revealing or producing color fringes than film. It's easier to "blow" one color channel first, with overexposure, which will produce a fringe. And the higher the lens's overall contrast (read - 90mm APO), the easier it is to blow highlights - the highlight brightness is not being dispersed over the whole picture as veiling flare, but hammering only in one spot.

 

https://mjambon.github.io/mjambon2016/purple-fringe/tree.jpg

 

One really has to look carefully at any purple fringing in detail, to determine what the primary cause is in any given case.

Adan, nice writing. Always helpful. 90APO's problem is not fully corrected axial aberration.

[NB23]

The 90 PURPO is deceiving and doesn't deserve the APO designation.

[lct]

Purple fringing can come from sensor blooming as well and can be observed out of apo lenses like 90/2 apo or 50/2 apo in high overexposure situations.

[jlindstrom]

^ Exactly. I had similar experiences with nocti 0.95 trying to shoot it ”naked” wide open during day time. Even at iso100 I was constantly fighting over exposure at 1/4000. Purple fringing was hideous and constant pain in the rear side...

 

Then I bought a suitable ND filter & got the exposures under control at or close to 1/500 and no more purple fringing.

 

So high contrast situations (shoot against the sun) & over exposure will make the beast come out from hiding.

 

 

Sent from my iPad using Tapatalk Pro

[NB23]

Who is overexposing and why? We can't have a logical conversation if we aren't assuming the very basics: a normal exposure.

 

Let's make things straight: the 90 PURPO is a purple machine, wide open. That's what it does.

[lct]

Who is overexposing and why? We can't have a logical conversation if we aren't assuming the very basics: a normal exposure.

Let's make things straight: the 90 PURPO is a purple machine, wide open. That's what it does.

 

Where did you see purple fringing in case of normal exposure? Never seen this so far. Only in case of overexposure like other lenses including apo's. 50/2 apo here:

 

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!

 

[NB23]

The 90 wide open is a purple machine.

[lct]

The 90 wide open is a purple machine.

 

 

Show us your pics or stop spreading rumors please.

[onasj]

I’ve shot quite a few portraits with my 90 APO at f/2 and never noticed purple CA artifacts. Indeed, I consider the 90 APO to have fewer CA artifacts in my use, and much greater sharpness, than the wide angle luxes. But I have to admit that I use the 90 APO to shoot people, not overexposed black lines against a blown out sky, because the latter sounds pretty ugly to me.

Edited October 13, 2017 by onasj

[Advertisement]

Hallo,

Look here: Leica M

[NB23]

Show us your pics or stop spreading rumors please.

Isn't this thread about this exact issue? Now it's suddenly a rumor?