(Tom) wrote in
news: m:
> All my AF lenses transmit their actual focal length to the camera.
> The exposure data recording does show this.
But is that really /actual/ focal length, meaning not the marketing focal
lengths printed on the lens? I don't think so. For example, the 100-
400/4,5-6,7 APO is really a 101-382mm lens. The 135/2.8 is a 133mm. This
info is needed for the best AS operation.
> In the Magic Lantern book
> "Classic Minolta Cameras", the sections on the i series cameras does
> state that Minolta zooms have data on up to 12 zoom positions that are
> transmitted to the camera for use in "P" setting of desired shutter
> speed.
[snip]
> There was also some info on one of the
> KM sites that stated that FL and magnification were taken into account
> for AS.
The question is how fine-grained the info used by the AS system needs to
be. If it really is true that the lens transmits info in twelve zones, it
can't be that critical; but I don't think it does. Rather I'd guess that
previous cameras don't take exact zoom positions (transmitted by the lens)
in account but rather uses twelve zones in their internal programming.
If the FL info was that course (not taking actual FL into consideration,
dividing the zooms in "zones") there would be absolutely no reason to do an
infinity search at startup (as mag change when focusing is orders of
magnitude less than when zooming a 100-400 inside a "zone"!). So, using a
little backward logic, I indeed think that
1) The lens transmits a unique lens ID
2) It also transmits "marketing" focal lengths in pretty precise steps
3) This is matched to an in-camera database storing all pre-D-lenses to
determine real FL; the D lenses probably have means to transmit "real" info
4) Focus distance is also used either from D or the in-body decoder
> I think I disagree with Magus on where the database of lens parameters
> is stored. He says the camera, but I think it is in each lens.
> Reason? - Because my 24-105 lens was designed after my 700si or my
> XTsi were made. But that lens works correctly with both of those
> cameras to do the "P" mode scene recognition and subject program
> exposure settings.
There is no incompatibility, so why wouldn't it work with older cameras?
They can't use D info, but they will still do their calibration search at
the beginning to ensure that the focus is at a predetermined position.
> However, my Tokina and Sigma lenses don't with any
> of my 14-seg metering cameras.
It is well known that many third party lenses can't do proper "expert
system exposure", but the incompetence of these makers can't be taken as
proof for anything -- we don't know exactly where they screwed up
> (This has to do with setting the
> desired aperture and DOF for the type of shot, not for getting a
> corect exposure). So the data needed for the camera to know the
> focusing position of my 24-105 (or any other newly designed lens)
> probably has to be in the lens chip. I think that there is camera
> specific calibration data that customizes each lens for each model
> camera (for example, changing camera motor/AF drive ratios).
Of course lots of info is in the camera ROM IC -- but I am more thinking of
/actual/ focal lengths, zoom range (probably matched to a unique "lens ID")
and the like. I am sceptical to the 20 year old AF lenses being so
sophisticated and forward-looking, but of course everything is possible