P3D Re: Starosta's site Metrics

[john bercovitz <bercov@xxxxxxxxxxx>]

Dr. T writes:

> How did you figure out the enlargement factor of eight?

I'm sorry, I didn't make myself very clear, did I?  I meant
that only as an example.  

> What happens to the ortho-perpective calculations when I
> cross-view?

You get keystone error and a bulging reconstructed image center; 
see Ferwerda.


While I'm on line, I must put in a vote for 3D portraits.
I really like them but they're kind of an interesting case.
Here's my experience.  If I don't know the person in the 
picture, my brain will correct the picture to an average so 
it is tough to know the correct distance to view from.  If 
it is someone I know, I immediately recognize a much smaller 
amount of non-ortho of the perspective type.  I am thinking 
here of a pair I did of a bust of George Gordon for NSA and 
another pair of my daughter.  The pair of GG was somewhat 
difficult to recognize as being out of whack even with a 
large perspective error.  The pair of my daughter is easy to 
see the error in with as little as 25% perspective error.  
This is funny because the human face is what the brain corrects 
best.  In fact, it will even correct a pseudo pair, usually.

John B


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