Re: P is for Problems, I is for Insanity

[ADavidhazy <ANDPPH@xxxxxxxxxxxxxxxxxx>]

> Did you notice that every panoramist has problem with there camera.  
> ... Everybody as a story about banding.  What causes banding? "The slit"
> somebody told me. The slit? OK. But i'm sure there is more than that. Why 
> does the slit causes banding? Is the slit manufacturer the common cause of 
> all panoramist headaches? Could the film manufacturers one of the cause of 
> banding? ...

Well, in "Cirkut"-type cameras, banding that manifests itself as "lines" of
density extending perpendicular to the direction of motion of the film are due
to changes in the speed at which film moves past the exposure slit. Thise that
are oriented in the same direction as the film moves are due to uneven slit
edges.
                                               
Uneven speed of the film past the slit also may cause unsharpness or blurring
of the image recorded on the film. To minimize such blurring it is advisable to
make the exposure time, or the time over which the image can move with respect
to the film, short. This calls for narrow slit widths.

To minimize banding one can use large slit widths thus minimizing the effect of
a discrepancy in film speed as it moves from one side of the slit to the other
and consequently on the total exposure. Another way to do this is similar in
concept and that is to make the edges of the slit fuzzy by locating the slit a
significant distance from the film surface. 

As you can see, there is an direct relationship between lack of banding and
evennes of exposure but an inverse relationship with image sharpness. It is
impossible to make the film move exactly at a given speed. Slight fluctuations
in speed are always present. In the audio industry associated with transporting
tape past a recording head they talk about flutter, wow, etc. ... these are
factors that the moving-film type panoramic cameras also exhibit but in visual
form. To make really sharp panormaic pictures you must use a really narrow slit
witdth. If you do so, however, the likelihood of picking up banding increases.
So, what to do? - well, compromise! 

In the case of rotating lens cameras the variations in density paralell to the
sweeping slit are caused by uneven sweep rate caused by various factors. With
these cameras there is no image blur introduced by uneven sweeping (assuming
correct camera construction) as the image does not (is not supposed to!) move
while the lens rotates on its rear nodal point. A density band perpendicular to
the slit orientation is caused by uneven slit width. 

I thought that going into the digital realm the problems of banding would be
gone forever. Not so. Banding-type artifacts seem to be present there too. My
cheap, hand-scanner based panoramic camera project is coming along nicely. I
will be making a presentation at the next SPIE (Optical Engineering Society)
meeting in San Jose in late January. I am looking for a hand-scanner with a
direct USB connetion to a computer. Know of one?

regards,

Andrew  o o  0 0 o . o  Davidhazy, Imaging and Photo Tech
         \/\/\/\/\/\/          http://www.rit.edu/~andpph
__________|        |_____________________________________