Anamorphic Lenses and physics at play

[Chris Adler]

Can anyone answer why when I shoot through my massive Kowa anamorphic projection lens with my 50mm I have to stop the lens down to about f8 or greater for the image to not be smeary and in focus? I can shoot wide open on my 28mm. This is on a BMCC MFT. What are the physics at play that require the stopping down?

 

[Ryan Patrick O'Hara]

Can anyone answer why when I shoot through my massive Kowa anamorphic projection lens with my 50mm I have to stop the lens down to about f8 or greater for the image to not be smeary and in focus? I can shoot wide open on my 28mm. This is on a BMCC MFT. What are the physics at play that require the stopping down?

Haha, great question. It's a tough one to explain. I'm not a lens designer, but I fancy myself to having a somewhat decent hold on how things more-or-less work. I'll do my best and hopefully some brilliant mind can step in and do a better job! First, there is a lot of stepping back to do.

I should first state that an anamorphic lens is best described as two lenses in one. There is typically a spherical lens at heart and then an anamorphic element (grouping? I'm not sure if it's a single element or multiple) most commonly in the front (although there are also rear anamorphic lenses).

Before we dive into anamorphic lenses, we should examine why spherical lenses typically also benefit from closing the aperture down. The answer has to do with how incoming light acts. Reflected light can be imagined as a cone shape of light emitting from a point at the subject. The cone of light enters/strikes the lens at various incoming angles and with the exception of the light entering through the center of the lens, those various angles must be bent and refocused to a point upon the sensor. Bending light is a tricky business.

First of all, light is made up of many different wavelengths. These all travel on different frequencies, yet must be bent to the same point. The faster the lens, the more the light the lens needs, thus the aperture must be more open. A larger aperture accepts a larger portion of the incoming 'cone' of reflected light, and thus must bend and refocus the additional light at greater angles. It is difficult to bend light at greater angles of acceptance, thus we get aberrations. Aberrations come in many different forms, but one easy example of a type of aberration is chromatic aberration. CA is an optical effect when all color wavelengths (at different frequencies) of light are not bent and focused equally. You may know this effect as color fringing, often a red, violet or green color, depending on which wavelength wasn't bent the same as the others and fell slightly offset from the rest. So you see, by restricting the aperture, the lens is accepting the less radical portions of the 'cone' of incoming light, in which the image is not comprised of the light that needed to be bend greatly. The result is a sharper image free of the aberrations and other optical effects softening the image.

In fact, the reason we even use optical lenses, is because we wish to have a large opening (aperture) to gather more light for photography. However, because the aperture is so large, the incoming light rays are of such variance, they will not focus on a plane with any sort of sharpness. Are you familiar with the term/concept of 'Camera Obscura' (latin for dark room/chamber)? (In fact, that is where the word 'Camera' came from, as techniques involving the concept of Camera Obscura gave birth to photography.)

Anyhow, the best example of Camera Obscura is a pinhole camera. A pin hole camera is a camera without a lens (as we know it.) In fact the aperture *is* the lens. The aperture is so small, (a pin hole), that the only light allowed to enter through the hole is the very direct and center portion of the incoming 'cone' of reflected light. Thus the vast amount of incoming light needs little or no bending. Of course small apertures have their own flaws, such as the law of diffraction. This will actually create small apertures to lose sharpness. So small apertures don't always make a sharper image. So if you understand that smaller apertures only accept incoming light that does not need to be 'bent' to the focus plane as much as wider apertures, you can understand that these images suffer less from the difficult optical issues of bending light.


Now it must be stated that the optimal aperture of a lens is actually different from lens to lens based on lens design. Because a lens has to bend lightwaves of different frequencies to the same point, the lens designers can design the lenses to perform optimally at different apertures, so there is not a hard rule such as "a lens is always best two stops from wide open." This is not true. It honestly depends on the lens design.

Now let's focus on anamorphic lenses.

As stated, an anamorphic lens is easiest to be thought of as two lenses in one. In fact, it's easier to explain it as one spherical taking lens with an anamorphic optical element in front (in rear in some designs). The anamorphic element(s) self adjust in accordance with the shifting of the 'spherical' lens optics when focusing. This is something many amateur tinkering minds don't realize. An anamorphic element that simply screws or clamps onto the front of a lens is not how it should work. A proper anamorphic lens has the anamorphic element adjust in concert with the spherical lens optics to maintain optimal performance.

Anyhow, an anamorphic element, typically a x2 squeeze, is a difficult element to make/incorporate when trying to avoid compromising the optical quality of a lens. This is because of the curved nature of the anamorphic element. Anamorphic elements are commonly 2x squeeze lenses, meaning they allow the lens to see double the horizontal field of view. Take a close look at the anamorphic element; at the very dead center, the element is more or less parallel to the rest of the optical elements in the lens. However, as you move away from the center, the curvature increases drastically and quickly departs from being parallel. This curvature creates added optical issues for reasons mentioned before; the difficulty of bending light. From my knowledge I would venture to say this curvature complicates the image in multiple ways. Among many things, I would expect to see an increase in aberrations an thus a drop in sharpness. The lens would likely suffer greater distortion and increased uneven field illumination from the law of reflection (although lens coatings and/or lens shading helps tremendously.) There would potentially be a reduction in contrast, which could also end up reducing apparent sharpness.

Anamorphic lenses also have a bent focus plane. This is a major reason you may be experiencing seemingly soft edges (if you were looking at a focus chart or something across the image and in one focus plane. This bent or curved focus plane will appear to 'straighten' out as you stop down the lens, as the increased DoF will better hide the bent focus by encompassing a greater area. However, when wide open this bent focus plane will become obvious and sometimes troublesome.

Does that help?

The short of it is that a wider aperture gathers a larger portion (at a greater variance of angle) of the incoming reflected light and it must be bent at more extreme angles to focus to a point on the sensor. Bending light is not easy and aberrations are often the result which soften the image in many ways.

 

[Grug]

What a wonderful post! Thanks for that Ryan.

 

[Chris Adler]

Thanks, Ryan. That was a great explanation. Much clearer than everything else I've read up to this point. Appreciate it.

 

[j1clark@ucsd.edu]

Anamorphic lenses also have a bent focus plane. This is a major reason you may be experiencing seemingly soft edges (if you were looking at a focus chart or something across the image and in one focus plane. This bent or curved focus plane will appear to 'straighten' out as you stop down the lens, as the increased DoF will better hide the bent focus by encompassing a greater area. However, when wide open this bent focus plane will become obvious and sometimes troublesome.

Most things are improved by using the center part of the lens, which is what the aperture 'does'. However, there is no free lunch, and so as one gets to smaller apertures, one has diffraction to deal with.

As for the 'physics'... if one traces the 'rays' through the lens, one can 'see' that a point on the object which is in 'sharp focus' on the image plane, has a number of 'rays' which the lens 'captures'. The ray that passes through the 'center' of the lens, is called the principle ray. The other rays from the point are called marginal rays.

Those 'rays' that pass through the lens at some distance from the center, end up being distorted due to the imperfections of the lens towards the edge.

The aperture 'blocks' those rays, leaving more of the 'principle' ray to form the image... and unfortunately with 'less' intensity...

Well maybe a picture is in order... (Here the term 'chief' ray is used, rather than 'principle')...

 

[visugeek]

Put another way, your 28 has a closer minimum focus distance than your 50. Wider lenses tend to. When you close the 50 to f8 or greater you extend the near end of your acceptable focus to reach the Anamorphic element. My 50 has a close focus of .45 meters, my 24 has a close focus of .18 meters. These are measured from the focal plane inside the camera, the sensor. You may also notice that when you focus your main lens further away it may get soft for the same reason.