Lens design and resulting colour saturation

[Ricoh]

Apparently the lens of the Lomo LC-A results in saturated colours, according to Lomography, and to be fair based on what I’ve seen I’d say it’s fairly true. By contrast, the Leica glass that I have produce natural results.

I’m interested to know how lens design can alter / effect the colour from neutral to saturated.

 

[Rob-F]

Whatever else may be involved, I believe that contrast is very important. I think higher contrast results in more saturated colors. I think so because it makes sense that a lens with low contrast will scatter more stray light into the image areas; and that stray light will "dilute" (if that's a good word for it) the purity or intensity of the color. Great question! I will be coming back to see what I can learn.

 

[Retro-Grouch]

Whatever else may be involved, I believe that contrast is very important. I think higher contrast results in more saturated colors. I think so because it makes sense that a lens with low contrast will scatter more stray light into the image areas; and that stray light will "dilute" (if that's a good word for it) the purity or intensity of the color. Great question! I will be coming back to see what I can learn.

I'm not sure that it's always a question of light being scattered and "diluting" the colors (although certainly that can and does happen). I suspect that there are materials in the composition of optical glasses that simply absorb certain wavelengths and lead to lower saturation. Of course, how this would work across the entire visible spectrum is beyond me, and my essentially non-existent knowledge of optical design. Then we have the question of coatings as well. Probably a whole other set of variables!
Like Ricoh, I've been puzzled by this. We all know how easy it is to diminish saturation and contrast, but how does a lens designer increase them, as in the Lomo? Certainly, someone in the RFF community has an explanation! Perspiring minds want to know!

 

[bluesun267]

Years of working in camera shops and using different cameras has led me to have certain overall opinions of lens manufacturers vis-a-vis color rendition, not necessarily in line with contrast which can be all over the place even within one brand or another. I've found Nikon and Canon colors to be less saturated--or is it more 'realistic'?

Leica (post 1970), Minolta and Zeiss give, in my opinion, the richest, most striking & saturated colors. Pentax, Voigtlander, Olympus, Konica etc, fall somewhere in between those extremes.

Pre 1970 glass is less uniform in color behavior and there are lenses well known for soft, muted colors (Summitar and the early Summicron to varying degrees); yet some lenses made in the same era give incredibly saturated, vibrant colors. Topcor 50mm f2 LTM is one I'm currently enjoying. I've also had big saturation from the Schneider lenses on a Kodak Retina. I'm talking about coated lenses only of course

 

[Retro-Grouch]

I agree with bluesun267 about the variation in color rendition between lens manufacturers. For my taste, Pentax lenses, followed closely by Olympus, have provided the most natural and pleasing color rendition and contrast. I wonder also whether lenses post-1970 were designed around a certain standard of film response, say, Kodachrome? That's always been regarded as the "cleanest" and "most natural" film, but of course we're veering off into the treacherously murky waters of subjective response. Let's not talk about Velvia...

 

[Ko.Fe.]

I have several examples for same manufacturer lenses been different. I have been turned off by cheap alternative to Canon 17-40 L from Vivitar. Colors where so dull. I had camera store person logic applied 🙂 on Vivitar until I got special version of Vivitar wide prime for FD mount. It was incredible lens, including colors. While testing F mount, I found Vivitar primes to be just awesome on colors.
Canon 18-55 EF-S lens was improving from version to version. Last STM versions of it is good, first versions were dull on colors.
Canon came from so-so USM non L lenses on colors to just as good STM version on colors.

 

[Sonnar Brian]

Lenses have to be corrected for color- and this can be dramatically different. Most lenses are achromats and are corrected for two colors to form images in exactly the same place. APO lenses, three colors, Ultra-Achromats- 4 or more colors. This does not address how quickly the colors deviate from forming at the same place. The colors can "muddy" and blur. A lens coating handles one color, get farther from it- absorbs on a curve. Multi-coated lenses, different layers are optimized for different colors. The net result- not all colors are passed equally. Other factors- air/glass interfaces, internal reflections, surfaces that might cause hot-spots and flare.

I tried looking for an online site for some of this- did not find anything as good as my 1950s Book by Kingslake, "Lenses in Photography". Ebay, a few dollars.

1934 5cm F1.5 Sonnar, wide-open uncoated optics, on the M9.

Warm November Day by fiftyonepointsix, on Flickr

6 air/glass interfaces. The blue of the sky comes out darker than the coated equivalent. Maybe it passes blue less efficiently? The eye is not sensitive to Blue, always looks darker to the eye than a modern lens captures it. My uncoated Sonnars seem to capture it closer to how I see it.

Pentax 85/4.5 Ultra-Achromat.



The scan does not do this justice. Uncoated Optics, the most color-corrected lens you might ever find. There are 40 of these known to exist.

 

[Sonnar Brian]

The uncoated 5cm F1.5 Sonnar, wide-open on the M9.

 

[TenEleven]

The uncoated 5cm F1.5 Sonnar, wide-open on the M9.

Nice sample pictures.
The other thing I noticed with uncoated lenses is that they seem to "clip" highlights earlier than their coated brethren. You do get more "open" shadows with them too, so the image may still have lower overall contrast. But I always wondered about the highlights - what the explanation for that would be.

Wondering if you've observed this, too and what your thoughts are.

 

[Retro-Grouch]

To attempt to condense what Sonnar Brian is saying, it seems there is a mind-numbing quantity of variables that could affect saturation and contrast. It's astonishing to consider what designers accomplished before CAD programs, when all the design and calculation was done by hand. Some of those old lenses are still without equal, in my opinion.

 

[boojum]

I agree with bluesun267 about the variation in color rendition between lens manufacturers. For my taste, Pentax lenses, followed closely by Olympus, have provided the most natural and pleasing color rendition and contrast. I wonder also whether lenses post-1970 were designed around a certain standard of film response, say, Kodachrome? That's always been regarded as the "cleanest" and "most natural" film, but of course we're veering off into the treacherously murky waters of subjective response. Let's not talk about Velvia...

I have had 35mm cameras since 1953. That does not mean I know how to use cameras well, it is merely a note of ownership. While stationed in France as a part of my military obligation the young French woman in the Fontainebleau PX told me about Agfachrome. Her take was that if photographing a chateau in Kodachrome it will look freshly painted. With Agfachrome it will have a more natural look. She was right and I switched to Agfachrome there and back here in the US. Costco also sold Agfachrome under their label which made it not only better but way cheaper. For me, Kodachrome was kind of like turning it up to 11. Agfa does have a downside of slightly orange-ish skin tones, ever so slight. Or maybe all my friends had jaundice, take your pick.

With this in mind would the German lens designers design to Agfa or Kodak?

 

[Retro-Grouch]

I remember Agfachrome being advertised as having "Old World" colors. It did have a warm, mellow palette. Unfortunately, it was nowhere near as stable as Kodachrome. I have slides from the seventies on both films; the Kodachromes look like they did on the day they came from the processer, but the Agfachromes are blue and faded nearly beyond recognition.

 

[Sonnar Brian]

My Sonnar was designed before color was in wide use. Color correction was just as important for Black and White film- badly corrected color meant blurry images.

The APO Summicron was introduced with the M Monochrom in 2012. I keep my APO Lanthar 50/2 on the M Monochrom.

 

[Sonnar Brian]

Nice sample pictures.
The other thing I noticed with uncoated lenses is that they seem to "clip" highlights earlier than their coated brethren. You do get more "open" shadows with them too, so the image may still have lower overall contrast. But I always wondered about the highlights - what the explanation for that would be.

Wondering if you've observed this, too and what your thoughts are.

The Sonnar, and most Zeiss lenses, are aimed at providing higher contrast than the Leica lenses of the day. My Nikkor 5cm F1.4 will blow highlights even on film, as opposed to the low-contrast Summarit. There is a trade-off between resolution and contrast. Leica went for resolution, Zeiss for contrast,

 

[TenEleven]

The Sonnar, and most Zeiss lenses, are aimed at providing higher contrast than the Leica lenses of the day. My Nikkor 5cm F1.4 will blow highlights even on film, as opposed to the low-contrast Summarit. There is a trade-off between resolution and contrast. Leica went for resolution, Zeiss for contrast,

Yeah all of the above is true and meshes with my own experiences. However, I guess I worded my question poorly. To put it more succinctly:

I have two nearly identical 1936 Sonnar 50/1.5 lenses - apart by less than 100 units - one is coated the other isn't. Both are in excellent condition. Yet, strangely enough the highlights are less likely to blow on the coated lens than on the uncoated one.

My working theory is that it's actually not so much the highlights blowing (further) -- instead it could a sort of blooming out of the highlights which gets layered on top of them boosting the overall exposure. The same effect that lifts the shadows in uncoated or weakly coated lenses.

 

[Sonnar Brian]

I also have two 5cm F1,5 Sonnars, one coated and the other not- less than 100 apart in the 190xxxx batch. I've not done a test for this- next time I have both out, will try. I think the way to do this is have both measure a light source and then display the histogram from the Raw files. I used to get paid t do things like this- in the 1980s. I rewrote some of my software from the 1980s to process DNG files from the M9 and M Monochrom. All in Fortran and Assembly, of course.

 

[TenEleven]

I also have two 5cm F1,5 Sonnars, one coated and the other not- less than 100 apart in the 190xxxx batch. I've not done a test for this- next time I have both out, will try. I think the way to do this is have both measure a light source and then display the histogram from the Raw files. I used to get paid t do things like this- in the 1980s. I rewrote some of my software from the 1980s to process DNG files from the M9 and M Monochrom. All in Fortran and Assembly, of course.

That would be super helpful, even if it just ends up confirming that I'm seeing ghosts...

 

[jcb4718]

This is an interesting thread. As others have said colour saturation depends of many things scene-to-print. I have five 50mm LTM lenses

1. f/1.5 Zeiss Opton sonnar in an Amedeo adapter
2. f/1.5 Canon, a copy of the Opton
3. f/2 Jupiter-8 with a distinctive blue lens coating
4. f/2 Carl Zeiss Jena sonnar, uncoated, made in 1937
5. f/1.8 Canon, double gauss

I once shot the same test scene with each lens in rapid succession. This eliminates things like the film, lighting etc. In terms of sharpness, the Opton, Canon f/1.8, Canon f/1.5. Carl Zeiss were all pretty similar. The Jupiter-8 was less sharp, mostly due to focus shift.

The Jupiter-8 produced the nicest colour saturation. The Opton, Canon f/1.5, Canon f1/8 images were all 'cooler', with the Canon f/1.8 the most 'cool'. The really like to look of my Jupiter-8 photographs even though under close examination they are less sharp. I took the Jupiter-8 colours to be my 'benchmark' and found in Photoshop a relatively simple adjustment of the 'blue levels', in the range 10-20%, made all the images about the same.

Not so with the uncoated f/2 Carl Zeiss lens! Firstly you have to adjust the contrast and even then the colours are different, denser somehow. I have not been able to adjust the images in Photoshop to make them about the same as those of the other lenses. This lens stands apart. It's still a great lens, just different. As you might expect it seems particularly suited to B&W film.

 

[raydm6]

I did some research on lens design and ray-tracing, and could not find a design reference to the OP's inquiry:

I’m interested to know how lens design can alter / effect the colour from neutral to saturated.

However, on reading about lens design approaches (especially pre-computer), it seems to have been part-science/part-art. There are now several special lens design applications and it appears designers use a suite of software packages in their approach.

For some reason, I felt compelled into looking how Walter Mandler designed lenses. His output was prolific to say the least. He produced over 45 high-performance Leica Lenses - (I own four Mandler lenses). Another list: https://www.l-camera-forum.com/leica...Walter_Mandler

It appears at Leitz, it was a group effort:

Reginald P. Jonas and Michael D. Thorpe, in the article quoted, also present a valuation of Mandler's significance. Mandler was Head of Optical and Mechanical design at Leitz Canada (ELCAN) for 20 years "during which time he was involved in the design of over 400 lenses, including many photographic objectives, as well as lenses for movie taking, movie projection, laser scanning, and other speciality optics. In those days optical design was very much a group effort. Walter Mandler's contribution as physicist and designer was to set out the general direction in which design solutions would proceed and to bring his experience and knowledge of optical design theory to select the shortest path to a solution. The solution chosen was not always the most excellent in the imaging sense but it would be the best solution, balancing performance, cost, and manufacturability. As a result many of these designs remained in production for decades".

from the article: Walter Mandler - A Valuation of Mandler's Significance.

This paper gets into the very technical approach to his double Gauss lens design: (PDF link)



Nice write-up of some of his lenses and their characteristics: In praise of the Mandler lenses

And some 'Gorgeous' examples here on RFF: Mandler Heaven: pictures of, and by, lenses designed by Walter Mandler

 

[Ricoh]

I did some research on lens design and ray-tracing, and could not find a design reference to the OP's inquiry:

However, on reading about lens design approaches (especially pre-computer), it seems to have been part-science/part-art. There are now several special lens design applications and it appears designers use a suite of software packages in their approach.

For some reason, I felt compelled into looking how Walter Mandler designed lenses. His output was prolific to say the least. He produced over 45 high-performance Leica Lenses - (I own four Mandler lenses). Another list: https://www.l-camera-forum.com/leica...Walter_Mandler

It appears at Leitz, it was a group effort:

from the article: Walter Mandler - A Valuation of Mandler's Significance.

This paper gets into the very technical approach to his double Gauss lens design: (PDF link)

Nice write-up of some of his lenses and their characteristics: In praise of the Mandler lenses

And some 'Gorgeous' examples here on RFF: Mandler Heaven: pictures of, and by, lenses designed by Walter Mandler

Excellent set of references, thank you!