Apo-Telyt 135mm f/3.4 ... chromatic aberration

[Knightspirit]

Advertisement (gone after registration)

29 minutes ago, adan said:

Hmmm - why? See post #14, page 1 of this thread.

Getting the colors focused and sharp (repeat, focused and sharp) in the same plane is the only thing that defines an apochromatic lens. What happens in the unfocused and blurry parts of the image is not a factor in determining "apo performance."

........................

Not aimed anyone in particular.

But if one asked one of the 99% of optical engineers NOT working in the photo industry (and maybe even some who do), about "color bokeh" or "color effects in the out-of-focus areas," one would want to leave some space between one and them.

So they would have room to fall off their lab stools laughing hysterically. 😉

Blur is blur - a defect in the plane of focus, and completely irrelevant anywhere else.

I was referring to CA - not out of focus color shift. So in that respect - it probably does have less. Spherochromatism is actually a thing - as this thread proves - I guess it isn't something that gets any correction from lens designers? You make it sound like everyone just ignores it, thinks it doesn't exist or it just doesn't matter - but it does show up and can be very difficult to eliminate (as was the case with this example). A lot of talk here about CA, color fringing etc but what we saw in that photo was specific to the out of focus areas (focus plane was color correct). So I guess what we can conclude is that an APO lens does NOT have any correction for spherochromatism and we should not expect any! 

[Advertisement]

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

[adan]

4 hours ago, Stuart Richardson said:

We do not perceive the world as shifting towards magenta or green as we observe it with our eyes, so seeing it in a photo is somewhat jarring.

Generally true - because "perception" takes place in the brain, not our eyes' physical lenses.

And our "perception" can filter out all kinds of things - one obvious example being that our brain will generally perceive a "white" piece of paper as being "white," regardless of the  wavelengths of light actually falling on it (6500°K blue-white outdoor sunlight, 2800°K yellow-orange indoor LED, etc.)

Another visual cortex filter makes it generally very difficult to physically "look at something out of focus" - the brain will try very very hard to "auto"focus the eyes on whatever they are pointed at. Try it - (but don't blame me for the literal headaches trying to do so may cause.) 

I will bet no-one here can actually describe or characterize the "bokeh" of their own eyes - despite having used them constantly for decades. The brain simply filters that out.

Christopher Isherwood notwithstanding, we are NOT cameras. And cameras are not us.

See also: https://en.wikipedia.org/wiki/Optical_illusion

[adan]

5 hours ago, Knightspirit said:

I was referring to CA - not out of focus color shift. So in that respect - it probably does have less. Spherochromatism is actually a thing - as this thread proves - I guess it isn't something that gets any correction from lens designers? You make it sound like everyone just ignores it, thinks it doesn't exist or it just doesn't matter - but it does show up and can be very difficult to eliminate (as was the case with this example). A lot of talk here about CA, color fringing etc but what we saw in that photo was specific to the out of focus areas (focus plane was color correct). So I guess what we can conclude is that an APO lens does NOT have any correction for spherochromatism and we should not expect any! 

Roughly speaking, that is a fair summary.

Spherochromatism is not "a thing" in its own right - it is simply a synonym for chromatic aberration.

"In optics, chromatic aberration (CA), also called chromatic distortion and spherochromatism,  is a failure of a lens to focus all colors to the same point." - Wikipedia

Someone on the web (not blaming you!) decided it would be cool and innovative and click-bait to break that obscure and "sciency-looking" word out on its own and limit it to backgrounds and foregrounds, but it is faux-science. Just substitute "chromatic aberration" for "spherochromatism" anywhere you see it, and all will be clear(er).

Do optical designers, even for photography, "just ignore it, think it doesn't exist or just doesn't matter?" No - although "doesn't matter" is closest to the truth. "Give it a lower priority" would be a better phrase.

The job of a lens designer is difficult. They have to corral and bend (refract) light that is photons travelling at 299792458 meters/186000 miles per second, and rigorously obeying Newton's First Law of Motion: If a body (read: photon) is at rest or moving at a constant speed in a straight line, it will remain at rest or keep moving in a straight line at constant speed unless it is acted upon by a force.

And do it with, basically, melted sand in spherical shapes (common, available, cheap, and easy to form - anyone can crush snow into a reasonable sphere with their bare hands).

But that basic, cheap, and easy method is highly imperfect (more earthily, "crap"). So then they have to adjust it get rid of what imperfections (aberrations) they can: spherical, coma, astigmatism, distortion, curvature of field, vignetting, long CA, lateral CA, Gaussian errors, Seidel aberrations (5), Schwartzchild aberrations (9), and so on. Add some minerals to change the refractive index of the glass in element 2; figure out how to produce a wavy surface (an asphere) economically, and so on.

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

And that process is recursive - fix one aberration and it will likely make another aberration worse. (e.g. flatten the field, and the corners may acquire more coma - or (on topic) correct the chromatic aberration in the plane of focus, and it may make the chromatic aberration in the out-of-focus areas worse or at least more obvious).

Therefore the designers have to pick and choose, or prioritize, or optimize, what they can and will fix, for the best overall imaging. Leitz/Leica (Dr. Walter Mandler specifically) developed an optimization algorithm for their own lens-design software. It would compute various versions of a lens, with different aberrations prioritized, and score each with a "merit-number" - the highest merit-number was the best overall image quality, given the tradeoffs.

Which produced, for example, the "tradeoff" for the Noctilux f/1.0: accepting quite high corner vignetting for better corner resolution at the larger apertures.

The 28mm lens on the Q series is similar - it corrects a lot of aberrations very well, using the trade-off of quite high distortion (straightened out by the camera's firmware).

.......

We had a similar thread a year or two back, about this same problem of color fringing in the OOF backgrounds made with the 90mm APO-Summicron-ASPH.

I proposed the idea that perhaps an additional element or group at the back of APO lenses, to linearize the RGB light rays as they exited the lens, would reduce the color spread (pink here, green there) in the blurry areas. 

However, even if possible, that could well 1) turn the $4595 135 APO-Telyt into a $24595 135mm Super-APO-Telyt-ASPH, and 2) would likely introduce or increase other aberrations. Or just simply look weird - the backgrounds wouldn't be blurry ever, and at any aperture.

[lct]

2 hours ago, adan said:

And our "perception" can filter out all kinds of things - one obvious example being that our brain will generally perceive a "white" piece of paper as being "white," regardless of the  wavelengths of light actually falling on it (6500°K blue-white outdoor sunlight, 2800°K yellow-orange indoor LED, etc.

OK but if we perceive this piece of paper with a color fringe it is a fact for us and if we don't perceive this fringe anymore with another lens we can compare those lenses accordingly. Those loca, laca and ca things are just technical explanations of those perceptions.

Edited March 6, 2024 by lct

[lct]

25 minutes ago, adan said:

[...] However, even if possible, that could well 1) turn the $4595 135 APO-Telyt into a $24595 135mm Super-APO-Telyt-ASPH, and 2) would likely introduce or increase other aberrations

Or sufice it to choose a lens doing less color fringing whatever theory can explain it. Some of them are not that expensive, the superb Tele-Tessar 85/4 springs to mind. Sample at f/4 on M11 below.

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 March 6, 2024 by lct

[Knightspirit]

9 hours ago, adan said:

Roughly speaking, that is a fair summary.

Spherochromatism is not "a thing" in its own right - it is simply a synonym for chromatic aberration.

"In optics, chromatic aberration (CA), also called chromatic distortion and spherochromatism,  is a failure of a lens to focus all colors to the same point." - Wikipedia

Someone on the web (not blaming you!) decided it would be cool and innovative and click-bait to break that obscure and "sciency-looking" word out on its own and limit it to backgrounds and foregrounds, but it is faux-science. Just substitute "chromatic aberration" for "spherochromatism" anywhere you see it, and all will be clear(er).

Do optical designers, even for photography, "just ignore it, think it doesn't exist or just doesn't matter?" No - although "doesn't matter" is closest to the truth. "Give it a lower priority" would be a better phrase.

The job of a lens designer is difficult. They have to corral and bend (refract) light that is photons travelling at 299792458 meters/186000 miles per second, and rigorously obeying Newton's First Law of Motion: If a body (read: photon) is at rest or moving at a constant speed in a straight line, it will remain at rest or keep moving in a straight line at constant speed unless it is acted upon by a force.

And do it with, basically, melted sand in spherical shapes (common, available, cheap, and easy to form - anyone can crush snow into a reasonable sphere with their bare hands).

But that basic, cheap, and easy method is highly imperfect (more earthily, "crap"). So then they have to adjust it get rid of what imperfections (aberrations) they can: spherical, coma, astigmatism, distortion, curvature of field, vignetting, long CA, lateral CA, Gaussian errors, Seidel aberrations (5), Schwartzchild aberrations (9), and so on. Add some minerals to change the refractive index of the glass in element 2; figure out how to produce a wavy surface (an asphere) economically, and so on.

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

And that process is recursive - fix one aberration and it will likely make another aberration worse. (e.g. flatten the field, and the corners may acquire more coma - or (on topic) correct the chromatic aberration in the plane of focus, and it may make the chromatic aberration in the out-of-focus areas worse or at least more obvious).

Therefore the designers have to pick and choose, or prioritize, or optimize, what they can and will fix, for the best overall imaging. Leitz/Leica (Dr. Walter Mandler specifically) developed an optimization algorithm for their own lens-design software. It would compute various versions of a lens, with different aberrations prioritized, and score each with a "merit-number" - the highest merit-number was the best overall image quality, given the tradeoffs.

Which produced, for example, the "tradeoff" for the Noctilux f/1.0: accepting quite high corner vignetting for better corner resolution at the larger apertures.

The 28mm lens on the Q series is similar - it corrects a lot of aberrations very well, using the trade-off of quite high distortion (straightened out by the camera's firmware).

.......

We had a similar thread a year or two back, about this same problem of color fringing in the OOF backgrounds made with the 90mm APO-Summicron-ASPH.

I proposed the idea that perhaps an additional element or group at the back of APO lenses, to linearize the RGB light rays as they exited the lens, would reduce the color spread (pink here, green there) in the blurry areas. 

However, even if possible, that could well 1) turn the $4595 135 APO-Telyt into a $24595 135mm Super-APO-Telyt-ASPH, and 2) would likely introduce or increase other aberrations. Or just simply look weird - the backgrounds wouldn't be blurry ever, and at any aperture.

That was an excellent wrap up - thank you!

[UliWer]

Advertisement (gone after registration)

Am 5.3.2024 um 12:28 schrieb lct:

don't know if it is CA or not but my 135/3.4 apo  does more color fringing than my Tele-Elmar and even my old Elmar 135/4 so i use the apo when i need f/3.4 mostly. It is also sharper at f/4.

A direct comparison (more than 100% crop):

135mm Tele-Elmar at f/4:

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!

Tele-Elmar at f/4.5:

 

Apo-Telyt at f/3.4:

 

Apo-Telyt at f/4:

 

Apo-Telyt at f/4.5:

 

 

 

[lct]

Interesting thank you. I have not the same experience on branches but i never did side by side comparos so far.

[UliWer]

Now I have to admit that my "direct comparison" was manipulative. I tried as well with the 135mm Elmar as I was sure that it showed rather obvious chromatic aberrations - though it really shows none in my example...

Elmar at f/4:

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!

Comparing this result with the Tele-Elmar it was clear that I didn't focus the Tele-Elmar properly (even LiveView isn't reliable with these lenses...) - which is obviously the cause for the strong blooming.

 

[adan]

Some key points in making tests like these:

A fully-blown highlight will produce a purplish color fringe regardless of the lens.

THAT is usually what causes problems in "leaves against skies" tests for fringing (or the bright whites against black tiles in Uliwer's examples) - unless you expose so that the sky is its natural blue or gray tones.

Note in Uliwer's samples the black tiles against the correctly-exposed blue sky produce no fringing, with any of the lenses.

If you blow out the highlights/sky, you get exactly what you deserve. 🤪

Digital sensors are more sensitive to the near-visible wavelengths (UV and IR - remember the M8?) than the eye, so the overexposure happens in those wavelengths too, into the red and violet ends of the visible spectrum. Thus the predominant "magenta" color of the flaring (red + violet = magenta).

AND the near IR/UV wavelengths do not focus in the same plane as visible light, even with "apochromatic lenses" (which are usually tuned for correct color focus only in the visible spectrum), unless intended for scientific/laboratory purposes). E.G the UV-Nikkor 105mm f/4.5 (BTW, if you think Leica lenses from the 1970s-1980s are pricey now......!)

https://www.nikon.com/business/industrial-lenses/lineup/uv/

https://www.ebay.com/itm/225635858387?chn=ps&_trkparms=ispr%3D1&amdata=enc%3A1MSAmhd5PS6CH1H8wf3rgCw9&norover=1&mkevt=1&mkrid=711-117182-37290-0&mkcid=2&mkscid=101&itemid=225635858387&targetid=1587262742097&device=c&mktype=pla&googleloc=1014485&poi=&campaignid=19894961968&mkgroupid=148855406073&rlsatarget=pla-1587262742097&abcId=9307911&merchantid=116328027&gad_source=1&gbraid=0AAAAAD_QDh-5fYDk4Po7jGMxRXrCQn5On&gclid=EAIaIQobChMI3rPH2pbjhAMVuDyHAx28ywVJEAQYASABEgJx6fD_BwE

Those of us who've been around a while remember lenses with a little red dot or R on the focusing ring - to indicate how much to "defocus" from visible light when making IR-only landscapes and such.

See section 5c Focusing issues here: https://robertreiser.photography/infrared-photography-tutorial/

Takeaway:

Misfocused and blurry IR/UV light - overexposed and blown out - yeah, that will produce a purple fringe with any lens. One with CA will just do it slightly sooner or more obviously.

[adan]

On 3/6/2024 at 1:02 AM, lct said:

Or sufice it to choose a lens doing less color fringing whatever theory can explain it. Some of them are not that expensive, the superb Tele-Tessar 85/4 springs to mind. Sample at f/4 on M11 below.

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!

If you had exposed this picture the same way you exposed the picture in post #43 (sky blown to white) - it would show substantial fringing also.

Go ahead and try it.

[lct]

I'll see this after my surgery (carotid artery) if you and the gods don't mind 😉 but in my most modest experience, there are no lenses with or without color fringing but with more or less of it and the Tele-Tessar is very hard to beat there.

[Robert Blanko]

vor 26 Minuten schrieb adan:

...

Misfocused and blurry IR/UV light - overexposed and blown out - yeah, that will produce a purple fringe with any lens. One with CA will just do it slightly sooner or more obviously.

I thought that most sensors including Leica M sensors have a filter stack in front of the sensor, including inter alia an IR filter and a UV filter? 🤔

[adan]

You assume those man-made filters are perfect? 

They all have some kind of roll-off (fade-out) that reduces UV or IR increasingly, but does not start or stop "perfectly" at a given wavelength.

How much will depend on how much of the visible wavelengths the camera/sensor manufacturer wants to interfere with (which will affect color accuracy of the extreme visible colors - intense visible violets or reds would get "filtered out" also).

So there will always be some leakage of UV/IR - and enough (too much) exposure will capture it.

Consider the graphs here (bottom of the page for Kolari). None of the transmission curves form perfect right-angles (pass wavelength 704 nanometers, block wavelength 705 nanometers or higher) - most are not even close.

https://infraredatelier.files.wordpress.com/2010/11/common_filters.jpg

https://kolarivision.com/product/kolari-vision-color-correcting-hot-mirror-filter-uvir-cut-filter/

We have a couple of "black-light" display areas at the gallery. Here is how one appears, to the eye, and to the "filtered" but otherwise uncorrected M10-P sensor. Lots of "invisible" UV being captured.

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!