Canon FL 17mm f3.5

The FL 19mm mounted on a Bell & Howell Canon with the special 19mm view finder necessary since the mirror is in the “lock up” position.

FL 19mm f/3.5

An Early Super Wide

I was in Ottawa this past April (2019) and I came across the Camera Trading Company in the basement of a building on Bank Street. They deal in used cameras and they have lots of them. I asked what they had in the way of interesting Canon gear. It seems that they were just going through some equipment from a collector who was disposing of his collection and they showed me some of it. Well, they had more than I could afford but I picked up two interesting lenses: an 

FLP 38mm lens for the Pellix and an FL 19mm f/3.5 wide angle lens which we are looking at here.

Researching this lens on the internet yields surprisingly little in the way of description or comments and what comments there are tend to confuse the various wide angle lens available in the early 1960’s and 1970’s. Kitchingman in his excellent book on Canon’s M-39 mount lenses says that this lens was available from 1964 through 1974 in a screw mount version and his photographs appear to show the identical lens without the FL designation on the lens name ring.

Apparently this lens was simply transferred to the FL Mount for the SLR cameras. Introduced in August of 1964 but Canon followed it up with the FL 19mm f/3.5R in November of 1965. The “R” stood for Retrofocus. This latter lens was an inverted telephoto design that did not interfere with the mirror operation on SLR’s thus allowing the normal viewfinders to function. However, the result has been much confusion in the internet comments about the three 19mm lenses of this period.

This is an interesting lens in many ways. Firstly, it covers an exceptionally wide angle, 96 degrees, which, in its day, was unusual. To accomplish this the rear elements of the lens extend back into the mirror box so far that the mirror has to be locked in the up position as the lens leaves it no room to function. And of course, it could not be used on the Pellix with its fixed mirror.

Canon FL 17mm f3.5

The lens and  shoe mounted viewfinder came in a Canon fitted leatherette box making a very neat package.

Canon FL 17mm f3.5

Inside was the FL lens itself and the viewfinder in a separete compartment under a velvet flap.

Follow the Cameras on Facebook

If you're interested in old cameras, "like" us on Facebook . Get news of what we're doing sent to you when we post something new.! Facebook Page
Canon FL 17mm f3.5

The lens from the front looks like a normal wide angle lens of typical construction.

Canon FL 17mm f3.5

The lens and viewfinder make a solid little package that is easy to hold and use.

Canon FL 17mm f3.5

From the side, however, the unique construction of this lens becomes obvious..

Canon FL 17mm f3.5

The camera operating controls are not obstructed and around your neck it seems a small package.

Canon FL 17mm f3.5

From the inside with the shutter open it is obvious how far the lens extends into the mirror box.

 The Canon Museum says that this lens was composed of 9 elements in 7 groups. The iris has 6 blades. Kitchingman says that the filter ring on the M-39 version is 55mm but on my FL version that is actually 58mm.

Mounted on the camera anyone who was not aware of what extended back into the mirror box would think they were looking at a normal pancake lens so thin is the external outline of the lens.

Because the lens is locked up the camera viewfinder will not work. So you are left with the auxiliary  viewfinder mounted in the equipment shoe and the lens, aperture and focusing, is totally manual. Because the depth of field on this kind of lens is so extreme focus is seldom a problem.

Having said that, this is an awkward lens to use. Because the profile is so narrow the focus and aperture rings are squeezed together and they interfere with each other. I have small hands and it is difficult for me to manipulate the settings easily.

Other than that, the camera and lens work well together. They feel balanced and tight and easy to manipulate (except for the setting of the two rings). The viewfinder is exceptionally bright and distortion free and pleasant to use. 

And how good is this lens?

My first reaction was to simply throw this lens on my Canon R, via an adapter, and snap a few pictures. A 30 mp image should show what this lens is about. Right? Well, that was wrong, wrong, wrong!

The lens fit the camera and adapter well enough. There was no mirror to concern ourselves with and the viewfinder had an image. But there was something wrong. The images are terrible. The two pictures below show heavy violet vignetting. The sharpness in the corners was abominable. This lens appeared to be useless. What was going on? 

My go to solution to these kinds of problems is to go to bed an worry about it tomorrow.

An image of our front yard taken on the Canon R using the 17mm f/3.5 lens. This is the full frame without cropping.

Same setup but an image of the backyard. This confirms that there is a serious problem with this lens on this camera.

I could not stop thinking that Canon would not put out a lens like this. As I thought about my setup I began to think about the sensor surface as opposed to a film surface. The sensor in a digital camera is composed of tiny cells which can set up interference patterns with light causing all kinds of colorful reflections. Not only that but the surface is covered in micro lenses over each light cell. Film is not like that.

This lens extends way into the camera so that the light coming from the rear element must approach the sensor surface at a very shallow angle to get to the edges of the frame. Could it be that the color vignette is really caused by an interference pattern combined with the light not being able to get down into the light receptor cells?

This lens was designed to be used with film so the obvious next step was to test it with film.

With film the problem is gone. There is no discernable vignetting and, although the image is Black and White, the color cast in the corners is gone as well.

Sharpness in the corners is good although somewhat distorted as ultra wide lenses will do. All these test images were at f/11 to give the lens a chance to show its best.

I had Ilford HP5 in my bulk film loader. It is not a really sharp film being designed for ISO 400 but it would be sufficient for our test purposes. So I prepared a 10 exposure roll and put it into my Bell & Howell / Canon FP using exactly the same setup pictured at the top of this page. I then took the same scenes over again and processed them normally.

Ecco lo! A totally different result. My musings about the effect of the low angle of incidence and the surface of the digital sensor must have had some truth to them. These images are now uniform corner to corner with no vignetting.

Examining the fine detail in these images you quickly see that this is an excellent lens. It may not be up to the standards of modern wide angle lenses but this is 55 year old technology and for its day was really impressive.

Zooming in the grain in the image introduces unsharpness before the lens does. This is an excellent lens and still a good shooter after all these years.

So I was right: Canon would not offer an obviously crippled lens. But to get the best performance from it you must use it as originally contemplated by its designers. Why is that not a surprise? And so we learn an interesting lesson. Not all old lenses can be mounted on modern digital cameras and perform well. In particular, it seems that wide angle lenses suffer problems. I do not know where the problems begin but I will next try 35mm and 24mm lenses and see if they show problems. I’ll let you know how that turns out.