My ideal camera's

Weia | Opinions | Published Dec 5, 2012

Recently some articles have been published here about 'ideal' camera's. Here come my ideas. I start with the opinion that for general work a lot of camera's and lenses already are good enough, except for one thing: it would be nice to have a camera that accepts lenses of all brands and all mounts... In a world where capitalism rules this is a faraway ideal.

In shooting non general work camera's and lenses and tripods turn out to be clumsy things. With one-issue camera's this could be overcome and pictures of unheard quality would become possible. When I were rich..., I would have them developed. I sketch two extreme camera's.


For architectural pictures two aspects of lenses are very welcome: a hugue field of view and an unseeable distortion. For full frame camera's the shortest existing focal length is 14 mm, giving a horizontal view of 104.3 degrees, a vertical angle of 81.2 and a diagonal angle of 114.2 degrees. For four-thirds Olympus has a 7-14 mm zoom and Panasonic has the same specs for a micro four-thirds lens. This gives horizontally 102.0 degrees, vertically 85.8 and diagonally 114.2. The winner can be found among the APS-C lenses. Sigma has a 8-16 mm zoom which gives on its own APS-C (not all APS-C's are the same): field of view horizontally 111.5 degrees, vertically 88.9 degrees and a diagonal field of view of 121.2 degrees. Compare with large format lenses where the Schneider Super-Angulon XL is the thing to beat, it has an image circle of 120 degrees.

The lens is one thing, control over composition is very important too. Below a rough sketch of a camera with an extreme field of view and build into it all control over tilt and pan you can want.

A is a normal quick release plate for use on a tripod. S is a circular sledge rotating exactly through the front principal point of the lens (it can be found where the front principal plane H and the optical axis intersect). C is the screw for sledging. D is a horizontal panning device. B is a short circular sledge perpendicular to S. With screws B and C you can position the optical axis exactly horizontal, first B , than C – the camera has built in sensors (gps or so) which give exact information about the direction of the axis and this information comes into the exif. The camera itself is a simple box with a tiltable screen and a cable outlet to use an iPad or a computer as viewer, for having more control over the composition. I often use a very high tripod, so the camera is much above my head and it is difficult to work with and judge composition carefully enough.

The camera itself is a simple box with some knobs and buttons that are useful (missing a lot of fantasy possibilities of todays camera's – 'now with more artist effects than ever'). But wait: the sensor is very special. I personally have the above mentioned Olympus 7-14 mm. At 7 mm it has a distortion of 5%, focused at infinity. In Photoshop I can remove this distortion (except from a tiny bit higher order distortion). The Olympus lens has zero distortion at f=10, but the lens of the Arkhitek should have a much shorter focal length, giving a distortion free image circle of 130 degrees (or someting like that). It is likely that it is easier to construct an ultra wide angle with say 8% distortion than with zero %, so we make an 8% lens and construct a sensor with the same distortion! A non square sensor. Focusing definitely is not internal focusing because this had much influence on distortion, so there only is a shift of the camera body back and forth (not sketched). The sensor by the way has the same width and height and portrait or landscape is just a button. You also have the possibility of using the whole picture, with bad edges, choosing portrait/landscape later. (It would be nice to see pictures of that Sigma 8-16 on full frame, where does it become how bad?)

As the lens has only one purpose and no internal focusing it can be optimised for just one situation, say sharp until infinity at f/8. Optimisation is done by way of a lot of aspheric surfaces or whatever needed to gave a distortion free system. As the camera too has only one purpose, working on a tripod with objects in rest, it could be a good choice to have very many pixels and a bit more noise at ISO 200.

Just for fun I show an f=7mm picture of the Domkerk in my home town Utrecht http://en.wikipedia.org/wiki/St._Martin%27s_Cathedral,_Utrecht . The picture is taken just some feet away from the place where Pieter Saenredam made a sketch on september 15 of the year 1636. The exact place is inattainable as the west part of the church has been destroyed by a thunderstrom in 1674. Saenredam is using very wide angle views, sometimes over 120 degrees! See some of his work at http://commons.wikimedia.org/wiki/Pieter_Jansz._Saenredam I show the picture after perspective control in photoshop. It is scaled down to 1000 pixels.


Another field I am active in is quite extreme macro work. A wasp of 0.8 mm long, what to think about that? I use an apo macro 150 mm of Sigma and often even an Olympus 2x extender. Together details of about 1/400th of a millimeter can be made visible with my old E-3. In this realm Canon is the winner with its 65 mm that can shoot only at 1x up to 5x. How small would be the finest details possible? Where more resolution isn't possible anymore for physical reasons? One micron or half of it? I have no idea, but the Macrobeast exactly gives that. The lens is optimized for only its highest magnification, with aspherical surfaces etcetera and the very best diaphragm. It most likely has many elements of slight curvature, being diffraction limited as a result. It has no iris diaphragm but a turret with circular holes with such details that influence of diffraction is as low as technically possible. Without diaphragm the lens has a speed of f/1.0 or even better, this diaphragm being completely useless photographically, except for focusing on the optical axis.

Again the camera itself is a simple box with some elementary knobs, dials and buttons. Plus a tiltable viewer – through the lens with such high magnification one too easily looses sight of those tiny creatures that have no idea of photography. And a twin macro flash is built in.

Apart from details depth of field is the main problem in macro photography. A nice way to overcome this is to stack several or many pictures. On a tripod with a focusing rail this gives rather good results, but only with an object in rest. Here a moth of 5.6 mm long, stacked with Zerene stacker out of too few originals: http://waarneming.nl/waarneming/view/72127577 (my own version is nearly three times wider). The rail can best be built into the camera, here a sketch.

A again is a quick release plate. S is a horizontal sledge for moving the lens plus camera back and forth. B is another sledge for moving only the camera for using other magnification rates (drawn much too short; think more of the old fashioned bellows). This is enough to make perfect series of pictures on a tripod for stacking.

But often insects and so on are not at rest at all. In reality I take many pictures without use of a tripod and often stack two or three of them by hand. This one exists of three handheld pictures with only sunlight: http://commons.wikimedia.org/wiki/File:SympetrumVulgatumFemale.jpg A stacker programm can't do that. Nothing special up to now has been said about the camera part of the Macrobeast, but here it comes: the whole sensor can move a few millimeters back and forth. So I set lens and camera on a certain maginfication rate (where maximal is the most interesting), move the handheld camera until the object is in view, press the button and very rapidly a set of flashed pictures is made while the sensor moves say 2 millimeters. For tiny insects 2 mm depth of field is huge! 

These two simple boxes with their not so simple lenses and not so simple sensors and not so simple mechanical additions are very much out of the box compared to todays general purpose camera's. 

Weia Reinboud