Re: Auto-focusing and IR focus shift

[Daniel Cardish <dcardish@xxxxxxxxxxxxx>]

At 10:12 PM 14/9/96 +0100, you wrote:
>At 11:23 AM 9/14/96, Willem-Jan Markerink wrote:
>>On 13 Sep 96 at 21:17, Daniel Cardish wrote:
>
>>> Not only that, but the camera seemed to adjust the focus point for the lens
>>> with the filter in place.  I think that since the autofocus system is
>>> sensitive to IR light, then if I allow only IR light to pass through the
>>> lens, then the focus shift should be handled automatically.
>>
>>Tricky subject. I did notice the same thing about a year ago, with a
>>EOS 100/2.8 and EOS-1 and at that time I thought it didn't refocus
>>for IR, it seemed to hit the same focus point over and over again.
>>This was with a #87C btw. It has also been mentioned in FotoMagazin
>>8/95, a German mag....
>
>
>I fail to see how this could be considered anything except a really serious
>bug.  If the focus point changes as IR content changes, the focus point
>should change, regardless of film, with any filter that changes the IR to
>non-IR ratio.  How would the camera know the difference between Tri-X with
>a 25 filter and HIE with a 25 filter?  In other words, as filters get
>progressively darker, the focus point will shift regardless of the
>desirability of that shift.  Even a yellow filter could cause some focus
>shift.
>
>
This is not a problem with normal chromatically corrected lens.  There
should be no focus shift within the visible spectrum (or at least only
minimal). However, most lens are not corrected for IR, hence the IR
correction mark on most lens.  If the only light passing through the lens is
infrared (due to a filter such as the #87 being on the lens), then the
autofocus mechanism will tell the lens how to focus the image based on where
the lens would normally focus IR light.  This focus point will be different
than with visible light.  This would normally not be noticed for films such
as TRI-X since they are not sensitive to frequencies that will focus at a
different point than where the lens has been corrected.  Presumably, TRI-X
will not record any image with an optically opaque filter such as the
Wratten #87.

I guess the only point I am not sure of is whether or not the #89B  (B+W
#92) is sufficiently opaque to visible light to make the above effect real.  

>
>


Dan Cardish  
<dcardish@xxxxxxxxxxxxx>

<http://www.spherenet.com/dcardish/photo.htm>
        


------------------------------

Topic No. 23