Automatic setting of hyperfocal distance?

Unless I'm missing something, this seems too simple to have not been

implemented, but I've not noticed it on any equipment....

 

The things that effect hyperfocal distance are aperture and focal

length (and sensor size for DSLRs). Today's cameras know what focal

length a lens is set at, and know what aperture is set (in aperture

priority or manual mode etc), so why shouldn't a camera's mechanics be

able to pre-focus the lens to a specific distance? Is there a

question over accuracy if the camera were to attempt focusing at a set

distance?

 

I would love to have such a mode available to avoid the maths involved

in calculating things myself.

 

So, am I missing something here? Go easy on me, please :-)

 

Pete

It WAS already made. The Pentax PZ-1p and those who love this film camera swear by this feature. Personally the camera was too bulky and complex for my needs, but just to let you know that it is tried and true technology.

Part of the problem may lie in the fact that autofocus occurs at maximum aperture, where the hyperfocal range is very short, anyway. The camera would have to be able to focus at an aperture where the range is significant, say f22, for instance, and I don't know if the AF sensors are sensitive enough for that.

 

Easiest thing is to use a chart of hyperfocal distances and focus manually......

Didn't know that, Douglas....thanks.

Hyperfocal distance focussing for landscapes gives you the minimally acceptable sharpness.

 

If that is what you want, are used to, need, are pining for, by all means use it. It is miniamlly acceptable, but never anywhere near the optimal resolution of objects on the sensor/film.

 

It is much more convincing - resultwise - not to use this minimally acceptable crutch and focus on the farthest object you desire to be sharp in the pic. Overall sharpness and detail will improve dramatically.

 

So, maybe the camera manufacterers know why they do not include this dumb trick.

Using the hyperfocal distance leaves the background out of focus. You are better off to focus at infinity (or the farthest important object) and use the aperture setting to get closer objects into acceptible focus. If that's not sufficient, there is the Scheimflug effect and a tilt-shift lens or camera.

<p>I don't know the innards of how all AF systems work, but AF doesn't require that either the body or the lens know at what distance the lens is focused. You simply move the lens' focused distance in or out, as required, until the subject is in focus, and then you're done.</p>

 

<p>So while I'd expect that it's possible to design an AF system in which the body can communicate with the lens and instruct the lens to focus at a distance of x, it's also certainly possible to design an AF system which has no knowledge of the actual distance. Such an AF system would not be capable of setting focus to the hyperfocal distance for a given focal length and aperture.</p>

 

<p>The AF system I know is the Canon EOS system, and I can't claim to know its innards. Many of the lenses for this system can tell the camera the distance at which they're focused (though rumour has it that in at least some of these lenses, the granularity of this is not great; to pull numbers out of a hat, a lens might be able to tell you it's focused at around 2m or at around 2.5m but might not have anything between those), but many cannot. That suggests to me that distance information is not an integral part of the EOS AF system, but rather is an optional extra. It's not even clear that absolute distance* information travels in both directions; the only references I've seen to lens/body communication involving the absolute distance is that some lenses can tell the body this information. I've never seen a reference that says the body can tell the lens what distance to use.</p>

 

<p>*: By "absolute distance" I mean the actual distance from the camera to the subject in some unit of measurement. The AF system certainly can tell the lens to focus nearer or farther, and give it some indication of how far to adjust focus. But AFAIK this is not done by saying "focus at 3m" or even necessarily "focus 3m farther away than you're currently focused." It could be "focus 15 motor rotation units farther away" where the camera and body have some shared understanding of what one unit is, or maybe "focus 1/3 of the way from the current distance to infinity" or something like that.</p>

I have several (inexpensive ) autofocus cameras in the family and have begun to wonder how they work. One is a Konica C35AF, which, if I read the skimpy manual correctly, starts from a position of 1.1 meters and moves until it finds an object to focus on and stops there. I assume that it will alays be impossible to have a distant object in focus if there are one or more closer objects in the field of view. Can any one confirm this or shed more light? The next model, C35MF, has a focus lock switch which may or may not work. On both of these there is a rangefinderlike moving mirror device which does not appear in the viewfinder.

Responding to that last: I have a C35MF which now does not work, but when it did, the focus lock did work. That is to say, it functioned. However, to prevent focusing too close, it required that one focus on something other than the actual subject and then recompose. What it was good for was placing a relatively near subject off center without losing the focus. I found that the result in general with this camera was good for shots of people and other fairly large and singular subjects, but not very impressive for scenic shots, perhaps due to its reluctance to go to infinity. The problem solved itself when the camera failed.

Some Canon bodies have either DEP or A-DEP modes. With DEP, you can select and focus on a near and far detail in the image (recomposing as necessary), and the camera will set focus at 7/17ths of the range from the near point and stop down to give a nominal "acceptable" degree of sharpness. The 7/17ths factor is really designed for group portrait work: the ideal factor ranges between 1/2 for macro work and a theoretical 0 for true hyperfocal conditions (since the rear DOF is then infinite!) A couple of the bodies have shiftable DEP, allowing you to adjust the aperture/shutter speed combination in the same way as a shiftable P mode, so you could stop down additionally to get added sharpness. The A-DEP mode works similarly, except that it judges the near and far point simultaneously from the points that lie under the AF sensors when you focus - which can be a bit limiting in practice. However, Canon have dropped DEP modes from their more recent high end DLSRs (their excuse being that it wasn't popular enough and they wanted the firmware space to implement other features) - to the consternation of some photographers who had valued the feature.

 

With today's AF lenses, this sort of mode is definitely a bonus, since the focus throw is so short, and distance scales on the lens are extremely imprecise and you may be lucky to even have one aperture marked to show DOF. Moreover, effective focal length and aperture often changes significantly with focus distance with internal focus and internal zoom designs, making calculation based DOF tricky to achieve. Another difficulty is that with the small viewfinders in many DSLRs it can be very hard to appreciate the real DOF even if you camera offers DOF preview.

 

Some P&S cameras offer a "snap" mode (Ricoh models come to mind) which does set hyerfocal focus as an aid to minimise shutter/focus lag.