Dario wrote:
> Of course, you cannot find a mechanism called circle of confusion buried > into a lens. However, an acceptable circle of confusion is a design task, > dependent on the print size and the distance you are supposed to look at it, > hence dependent from film/sensor size (a smaller sensor aimed to get > comparable results needs more enlargement). > So the lens must be designed for giving a circle of confusion matching well > the largest print size you are expected to get from the picture (depending > on the film/sensor size). Otherwise, even the focused details will look > blurred (out of focus). But what you are talking about here is lens quality; in particular lenses that have so low resolution that whats further away frome the plane of focus will be unacceptably fuzzy earlier because whats in the plane of focus isn't that great to begin with. However, this factor is completely unrealistic in real life and perhaps more than compensated for by the fact that DOF is relative. A very sharp lens will have apparently less DOF than a not so sharp lens because unsharpness in the DOF zone is by the brain measured by how it compares to whats in the plane of focus. The sharper whats in the plane of focus look, the more unsharp the stuff that isn't will appear. What you was saying was that an MF lens used on a 35mm body would yield smaller DOF that a 35mm system lens of similar focal lenght used on a 35mm camera. This isn't true because the circle of confusion is NOT a function of the lens but the format the lens is used on. For 35mm the circle of confusion will be a constant (whose value will be debatable!). As DOF is for all practical purposes the function of COC divided by the square of the focal lenght, it follows that lenses of a certain focal lenght, say a Pentax 645 lens and a K mount, will display the same DOF on a 35mm camera. Using a a lens of a fixed focal lenght on a fixed format (in this case 35mm) means that all variables relating to DOF will constants in this example. In other words, eg. the Pentax FA 645 200/4 will display the same DOF as a K-mount 200mm lens on a K mount body. It is true though that 35mm cannot be magnified as much as medium format! Hence 35mm show less sharpness and DOF than larger formats all things equal because it has to be magnified more to reach a certain end result. But this has nothing to do with how the lenses are designed. It isn't really particularly relevant either because no one use 35mm in order to blow it up to similar sizes as medium format can manage, and if you do the least of your problems will be DOF. Hence, you could say that, say, a 645 200mm lens will yield less DOF on a 35mm camera than on a 645 if the end print size is the same. They will give the same DOF if end magnification is the same. A 200 for a 645 will give the same DOF on a 35mm camera as a K 200mm on the same camera regardless. > A lens capable to give a smaller circle of confusion (hence a lens designed > for a smalller format) will render dots as dots even when farther from the > focus plane than a lens designed for a larger format can do, hence will show > a wider DOF. Why lenses for a larger format cannot be designed with the same > circle of counfusion as the lenses designed for the smaller format? It's > because designers have to correct a larger film/sensor size. If you have to > keep a larger format acceptable up to its corners, the average quality > lowers. What you are saying is that a rotten lens will give a smaller COC than a good one. This is independent from format. Theres no theorethical limits for making a lens with large coverage (larger format) of similar or even better resolution than 35mm system lenses. There are cost and practical issues for not doing so, but it is not material for generalizations as there are MF lenses than can give as small COC as virtually anything made for 35mm. Pål
