home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: rec.photo
- Path: sparky!uunet!spool.mu.edu!uwm.edu!linac!att!att!dptg!ulysses!allegra!princeton!ernie.Princeton.EDU!ajackson
- From: ajackson@ernie.Princeton.EDU (Andrew William Jackson)
- Subject: Re: How Autofocus works?
- Message-ID: <1993Jan20.013011.27498@Princeton.EDU>
- Originator: news@nimaster
- Sender: news@Princeton.EDU (USENET News System)
- Nntp-Posting-Host: ernie.princeton.edu
- Organization: Princeton University
- References: <1993Jan19.172819.9064@sol.UVic.CA>
- Date: Wed, 20 Jan 1993 01:30:11 GMT
- Lines: 23
-
- In article <1993Jan19.172819.9064@sol.UVic.CA> mstoer@sol.UVic.CA (Marcell Stoer) writes:
- >I would like to exactly how the autofocus works in the new SLRs. I have
- >noticed that the EOS cameras have two photodiodes below the mirror and
- >I would like to know how they are used to focus the camera. I want
- >more than a handwaving explanation though. I thought you could offset
- >the two photodiodes and measure the ratio of their detected light intensities
- >and compare this to some calibration table, but that would be very awkward.
- >It probably wouldn't give much of a dynamic focussing range since you need
- >to offset the diodes about 2-5cm for things to work well. Clearly that's
- >not possible in a 4-5cm camera body.
- >
- >Does anyone know exactly how this works? If there are some publications
- >on this I wouldn't mind the references.
- >thanks,
- >-ms
- >--
- >Marcell Stoer Internet : mstoer@sol.uvic.ca
-
-
- I always thought the camera tried to maximize high spatial frequencies
- in the 2d fourier transform of the image. For vertical patterns this
- should only require 2 diodes.
-
-