[MF3D.FORUM:306] Re: 1/30 voodoo

[Tloc54452@xxxxxxx]

Tom wrote:

> The 1/30 "rule" is sophistry masquerading as fact. You may as well use dice.
>
> It works in *one* certain situation, but not in lots of others.  [..]

Oleg wrote:

> [...]
>
> If I’ve got it wrong, please set me straight.

I don't know if "wrong" is the right way to put it.  There are a lot of
ways of interpreting the 1 in 30 rule.  They derive from the fact that
if you view from the correct perspective point*, you can tolerate an
on-screen disparity of 1/30 of the focal length of the lenses.  If you
substitute infinity for one of the distances, you get the conventional 
1 in 30 rule.  See Ferwerda or Waack for more detail and how to calculate
or see my website or Tom's for how to do calculations following their work.

You can beat this 1 in 30 rule if you have a gentle change of distance -
foreground objects not lying directly in front of background objects.

* The correct perspective point (see Kingslake) is the focal length of
the taking lens times the enlargement.  So if your format is ~1" high
and you blow it up 50 times when you put it on the screen, then you
should sit 50 times the focal length away from the screen to sit at the
perspective point.  So if your taking lens was 50 mm, you want to sit
2.5 m from the screen.  We sit much farther from the screen at OCC, Oleg,
hence the noticeable degree of stretch in many of the images (like Barrie's
nudes).  You can see stretch in any image of something which you know the
correct shape of and the reconstructed image is close enough to you, the 
viewer.

Enough of my blither; I don't want to bore you nice poeple.

John B