Re: Hyperstereo threshold

[Stephen Kearney <slk@xxxxxxxxxx>]

Thanks for the replies. I guess what I really want to know is what kind of
aesthetic impact the 90mm separation will have. I've used the formulas and
understand the technical implications. Basically I want to know if my
pictures will look very 'hyper', with the apparent reduction in scale. The
effect is definitely noticable with my 4x5 stereo camera due to its 150mm
separation.

I don't remember Greg's pictures suffering from the undesirable effects of
hyperstereo. Perhaps I'm just remembering the Spud shots, or Greg only used
the twinned Ricohs for subjects like city skylines where hyperstereo would
be a benefit.

-Stephen

Joel Alpers wrote:

> john bercovitz wrote:
> >
> > > Only one warning! I am new at this, and I probably breaking all
> > > kind of rules but there is no one around to tell me that I am wrong.
> >
> > No rules are being broken here.  In fact, your observations go right
> > down the line with the rules.  The nominal rule is 7' for standard
> > separation but that's only nominal.  In most situations you can get
> > much closer.  The 7' would scale up linearly with the increased
> > stereobase if I remember correctly.
> >
> > John B
>
> Ha ha! Now here's a REAL switch - John B answered a question in English,
> and I provide the formula!!!!
>
> Here's the formula for stereo base:
>
>           D * S1 * S2
>     B = ---------------
>          F * (S2 - S1)
>
> Where:
>
>     B  = Base separation
>     D  = Deviation on film
>     F  = focal length of lens
>     S1 = distance to near object in picture
>     S2 = distance to far object in picture.
>
> For medium format, D is about 2.5mm. Re-arranging to get the near
> point, I get (check me!)
>
>
>           B * F * S2
>    S1 = ------------------
>         (D * S2) + (B * F)
>
> Usually you include infinity in the photo, so as S2 -> infinity,
> the (B*F) term becomes insignificant, the S2's cancel out, and
> we're left with the simpler:
>
>          B * F
>   S1 = ---------
>            D
>
> For the 80mm 124G lenses 90mm apart, I get  90*80/2.5 = 2.88 meters,
> or almost 9 1/2 feet.
>
> At 8 feet (with infinity in the photo)  on-film deviation goes up to
> just over 2.95mm, and at 7 feet, it goes to 3.37.  You can probably
> fudge with this -some-, but the larger you make it, the bigger a
> chunk you're cutting out of the people who will be able to view
> the slide without eyestrain. I break this "rule" at times, but when
> I do I'm careful who I show the slides to.
>
> Do some experimenting with your single camera! Perhaps a piece
> of angle iron with two marks 90mm apart -- take a photo of a
> scene (including infinity) and put a near object 9', 8', 7', 6'
> away. It would make a fun entry for the MF folio...
>
> Interesting that you bring this up right now, as yesterday for
> fun I was working on a CGI WEB program to allow someone to plug
> differnt numbers into the formula(s) above. Thought it might make
> a neat learning tool for the "tutorial" section of the RMM WEB page...
>
> Joel.



--
_____________________________________
 Stephen Kearney : mailto:slk@xxxxxxxxxx
http://jimkearney.ne.mediaone.net/~slk
           Oakland, California