[Tfug] OT: Optics

[Adrian]

On Saturday 16 December 2006 15:20, Bexley Hall wrote:
[snip] 
> I was intrigued by the number of *mirrors* inside
> it!  And, soon realized they must be concave mirrors
> cascaded to act as a focusing lens (i.e. to take
> the ~9" wide scan line and map it onto the ~0.5" CCD.
[snip]
> And, I also have a vague recollection (from a high
> school physics class 794 years ago) about lenses
> causing color distortion (?)... "fringe effect"
> (sorry to be vague but those neurons haven't been
> fired in many years  :<  )
> 
> Can someone clarify this?  I.e. aside from the
> practical aspects of using mirrors being more
> mechanically manufacturable in this sort of a
> package, what *other* advantages they might have
> over the naive approach of a lens?
> 

Yep... many many scanners use a linear CCD or CMOS array instead of a full 
size imaging array. Not only were such things not available just a few years 
ago... a minimum 2500*3300 (8.5"x11"@300dpi = 8Mpix) would have been insanely 
expensive. On the other hand, a couple 1x1024 linear CMOS arrays (like here: 
http://cgi.ebay.com/Panasonic-CCD-MN3674-linear-imaging-sensor-10-chips_W0QQitemZ300059842384QQihZ020QQcategoryZ67003QQssPageNameZWDVWQQrdZ1QQcmdZViewItem),  
were (are) pretty cheap. And it only took a mechanical sweep with a single 
1x1024 array to scan 10" wide and infinately long at 100dpi resolution.

My understanding of optics is fairly fundamental (I'm sure there are probably 
several UofA people on the list studing/doing it every day), but overall you 
are correct. The problem with a simple lens is refraction of light passing 
through it. That is, as light passes through the interface between two 
materials, it bends (an angle of refraction), and different wavelengths of 
light (colors) bend at slightly different angles (i.e. it also acts as a 
prism). A "perfect" mirror, on the other hand, reflects all the light at the 
same angle (the angle of incidence), instead of refracts, so the colors are 
not separated. 

And so ends this evenings armchair physics lesson, no doubt complete with 
gross mischaracterizations of basic optical science priciples...

Adrian