re:Digital vs Analog - Part 2

[P3D Larry Berlin <lberlin@xxxxxxxxx>]

.............. Continued from previous post....................
the fact that in the *multiple image capture* scenario, the separate capture
devices would not be in the same identical spot. This would pose no problem
to software processing and insure there is a spread of data possible between
analagous points in any two samples. It also means that you have more than
one vantage point and therefore could synthesize any desired number of
stereo pairs with any desired stereo base, inside of actual sampled points
and a certain spread of range outside the sampled points. The spread
outwards would at some point have a decreasing percentage of accuracy.
Another advantage to this would be that instead of telephoto for obtaining
larger images, you could synthesize from the available spread of images
(assuming a cone-like area-shape based on a circular spread of imaging
devices), the actual perspective of some other point closer to the subject.
This combined with enlargement would provide more than the conventional
camera system could offer.

In this scenario, more image capture devices in one camera unit simply
equates to a higher resolution, a bigger total *eye*. A compound eye.  The
diameter of the arrangement of capture devices determines the possible
stereo base range and the possible perspective range adjustment. If these
devices are in a circle, they would also provide usable disparities for
vertical displacement as well as horizontal. You could for example
synthesize the effect of turning the camera on it's side or any angle other
than horizontal. It would be useful for helping to determine precise pixel
quantities and qualities in whatever resolution the user has chosen for the
output. This basic technology would supply as a normal standard, resolutions
far in excess of current film and you could routinely work in any chosen
format with the same camera. (35mm, medium format, etc.)

Price hasn't been a part of this consideration yet but once the processing
steps are worked out and made a part of an on-board processor, these cameras
could be as cheap as today's better digital watches and clocks and
calculators. The cheapest ones wouldn't offer all the functions perhaps but
even the cheapest could exceed 35mm resolution standards. The other ignored
part of the equation is data storage costs. By the time this camera gets
developed maybe a more dense and less costly alternative will be available.

>
>Consider the situation where a picture is taken and the background is
>*COMPLETELY* white.   If *one* pixel-width circle of the image is only 
>*slightly* darker... say 1% darker.  Will that film faithfully resolve
>that one pixel with the correct 99%-brighness value?  The digital camera
>could.  And even more-bits-per-pixel are available now in scanners so
>the one-pixel dot could be even less different.

*************  See my comments above about how digital pixels approximate
the in-between and surrounding areas instead of just their own location.
Film at least is totally faithful to each recorded bit and no other. Digital
uses this process to approximate a higher resolution and it works to a
degree... at the least it helps partially to correct for the *neat little
rows and columns* effect. 

>
>I can't see that the comparison of pixels is one-for-one.  I'd guess that
>each digital bit  is "worth" perhaps eight film-bits, 

*************** if film bits were binary,... but since they aren't, then
each of those eight film bits would be 24 bit (or more), making film a lot
higher than digital, depending on the size of your pixel. For viewing in a
monitor or as defined by the current single-element capture process, this is
a fixed size. For a graphic file, it is whatever you define it to be
(virtual space is infinite) and is usually determined by what is do-able
with the processing and storage capacity available. Ultimately it comes down
to a comparison of which printing process gets closest to the desired film
grain resolution. I think it's the infinite nature of digital (virtual)
space that provides the excitement and drive behind it's popularity, not the
direct comparison of today's commonly available resolution options. 

>so using your numbers, 
>35mm film is  equivalent to about 4 million 24-bit pixels.  That's
>just a SWAG, but I think it's less than the full 30 million.
>
>Or so it seems.   Correct my logic error (or terminology error
>which is often if not usually the case)  if there is one!
>
>Mike K.
>
>

********  I don't know which numbers for film are most accurate, but I know
that digital has the potential to go beyond film eventually. I most disagree
with the industry when it assumes the customer's demand is the lowest
possible configuration and chooses only to move in that direction. That
particular quality is supposed to define your innermost core of design
parameters, comprising only a part of your total design goal, not it's outer
limits. 

The last bit - Regarding the *analog vs digital* topic line,....... 
however we derive the information our perception process is analog but
highly multi-dimensional. Digital is a useful method of interfacing our
analog with the infinite. On and Off, the basic digital building block, are
symbolic of the everflowing Yin and Yang. The basic analog action of all
existing matter. So at it's root, digital is, despite it's stepped nature, a
kind of analog after all.

Larry Berlin

Email: lberlin@xxxxxxxxx
http://www.sonic.net/~lberlin/
http://3dzine.simplenet.com/


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

End of PHOTO-3D Digest 1754
***************************
***************************
 Trouble? Send e-mail to 
 wier@xxxxxxxxxxxxxxx 
 To unsubscribe select one of the following,
 place it in the BODY of a message and send it to:
 listserv@xxxxxxxxxxxxxxx 
   unsubscribe photo-3d
   unsubscribe mc68hc11
   unsubscribe overland-trails
   unsubscribe icom
 ***************************