I need some education... 35's and distortion.

[Juan Valdenebro]

Hi,
I’d like to read a good article explaining optics from a design point of view, but I haven´t found one yet… Can someone recommend any? I have no idea at all on that subject, so everything would start educating me… Thanks.
In particular, I wonder why is barrel distortion present in several 35mm rangefinder lenses… I’ve read only summarons and V1 Summicron have close to zero distortion, (maybe summilux 1 too? I don’t remember), and of course Zeiss Biogons, but the majority of (more recent) 35’s, including those considered the best, the summicrons made for the last 50 years, prefer some distortion: I imagine such varying amount of distortion is there to find a formula that’s superior in other fields…
That’s exactly what I’d like to read or understand clearly: maybe it’s different depending on the type of optical formula, but is it possible to say which aspects of a 35mm lens would be (slightly) sacrificed if distortion is kept out the formula?
Vignetting? Closest focus? Center or edge sharpness? Would it affect the way the lens defocuses the background?
Thank you!

 

[Roger Hicks]

Dear Juan,

Try Cox's Optics. It's widely available second hand.

Meanwhile, here's a quote from Dr. Nasse at Zeiss (from memory, in conversation): "The only way to tell what a lens is like is to build it."

Cheers,

R.

 

[FrozenInTime]

At a high level, symmetrical lens designs have lowest distortion.
The near symmetrical Biogons are far better than tele-centric Distagons.
I think there was an issue of Zeiss camera lens news that went into the details.

The tiny fast MS Optical Apoqualia 35/1.4 is almost free of distortion.
The Zeiss C-Biogon 35/2,8 similarly so.

 

[ferider]

Hi Juan,

Distortion is one of the 5 basic (first-order) Seidel aberrations:

- Spherical Aberration
- Coma
- Astigmatism ("focus shift")
- Curvature of Field
- Distortion

Basically, a lens designer starts from a given lens design with multiple unknowns (curvature and distance of glass-to-glass and glass-to-air surfaces, and glass types), then builds an equation system via ray tracing, which is then optimized mathematically, including trade-offs of one vs other optimization targets. Depending on the design target, some aberrations will be better corrected than others. For example, my VM 35/1.7 Ultron has almost zero distortion, but some curvature of field. Or, the pre-asph 50mm Leitz Summilux has almost zero Coma, but some barrel distortion. The same design can be found optimized for different targets. For example, when I compare 3 of my classic Sonnars with identical diagram, I observe this ((+) for good, (-) for bad):

- 1951 Jupiter 3 (basically the "original" Sonnar): ........ (-) Spherical Aberration, (+-) Coma, (+) Curvature of Field, (+) Distortion
- Late S-mount 50/1.4 Nikkor: ...................................... (+) Spherical Aberration, (-) Coma, (+) Curvature of Field, (+) Distortion
- Late Contax Mount Sonnar: ....................................... (+) Spherical Aberration, (+-) Coma, (-) Curvature of Field, (+) Distortion

Today, with the availability of computers, also 2nd order aberrations and more complicated non-spherical glass surface shapes can be accounted for.

Here is a nice summary: http://www.quadibloc.com/science/opt0505.htm

Wikipedia has a longer write-up with references at the end: https://en.wikipedia.org/wiki/Optical_aberration

Roland.

PS: as a side note, computers have been used for lens optimization since over 60 years (Mandler with the rigid Summicron is the earliest that I know of).

 

[icebear]

Unless you shoot brick walls, the importance distortion (or lack thereof) is highly over rated😉

 

[sepiareverb]

Unless you shoot brick walls, the importance distortion (or lack thereof) is highly over rated😉

True, architecture (brick or otherwise) is the area where distortion is apparent.

Thanks Roland and Roger, I could use some reading material, those sound appealing!

 

[Juan Valdenebro]

Wow, thanks everyone for all the great information!
Roland, great to hear you again... Thanks...
That was a great and deep introduction! I really appreciate your precision: it's a pleasure as always. I'll read more now...
It's amazing lenses with the same focal length and the same type of design, can be made to produce so many differences in rendering... The quote shared by Roger makes a lot of sense... I imagine it must be a huge pleasure to work in lenses' design... It would be a dream to talk an evening with Berek and Mandler about how things changed from before WW2 to Mandler's first findings in the late 40's and 50's... Someone should write a comprehensive book on both of them and those two decades, or even make a good movie...
Thanks again!
Cheers,
Juan

 

[Roger Hicks]

Talking to Dr. Nasse is certainly very interesting, because it also shows how much influence personal preferences have on lens design. For example, one designer might prize compactness, and put up with a degree of vignetting in return. Another might regard focus shift as trivial and easily overcome by moving back and forth a little when shooting. Yet another might not care too much about speed, but be looking for ultimate technical quality even if it means making a very slow lens.

Of course they pay attention to what the market might want, but the simple truth is that there's generally an adequate sized market for most things, including cost-no-object lenses -- though there are always those who fail to appreciate that (for example) there's a good reason why a Zeiss Otus 55/1.4 costs as much as it does, but live in a fantasy land where you could have a 35/1.4 the size of the last pre-aspheric Summilux with razor sharpness from edge to edge at full aperture, no coma, no vignetting, and costing the same as a cheap zoom.

It's also worth remarking that from my understanding, which is far from expert level, telecentric designs are often optimized for digital while non-telecentric designs (including older designs) may offer significant quality advantages on film. This may be one reason why so many Hollywood directors prefer film.

Cheers,

R.