[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: IBC Itch & Urge



At 4:58 PM 9/19/96, Mikael Reichel wrote:

>.....................Spectrally the light source provides wider bandwidth
>than halogens or CRT's,  it
>does so more spectrally uniformly thus reducing heavy masking in the post
>PEC/CCD domain when it is, strictly speaking, too late (from a scientific point
>of view).
>..............Obviously, this allows to
>adjust light rather than (by the conventional method) electrical gain (and
>noise
>with it)  in the post film domain. An intelligent eye-like colour/detail
>concept
>can thus be designed without taking into account  the constraint of an
>inadequate light source,


Mike R. has brought up some valid points, but missed mentioning some others.
The broad spectrum of a Xenon source is much closer to a daylight or
projector lamp viewing situation than a telcine tube can ever hope to be.
It's also very bright. However, the telecine tubes these days are bright
enough to blind photomultipliers, as we all know. The real issue is not the
light output, or so much the spectral characteristic.

The fact of the matter is there needs to be no burn and very little shading
compensation on the CCD machine. Even more importantly, *there is no
afterglow compensation and little or no compensation for spot or aperture
size*.

There is an enormous noise penalty in compensating for these tube-based
drawbacks, quite apart from the problem (as I understand it) that the
telecine cannot presently *correctly* perform the de-convolution necessary
to compensate for spot size and afterglow. The area where photomultipliers
score is in the several orders of magnitude linear characteristic, which is
achieved at the expense of any knowledge about the spatial details of the
photons falling on the tube photocathode. In other words, its all a blur as
far as the PMT is concerned.

So, we gotta scan somehow to make use of all this PMT dynamic range, hence
tubes or lasers, neither of which are ideal solutions. Tubes lag, have
grain, burn, falloff, are bulky, heavy, expensive, are sensitive to
magnetic fields etc, etc. Lasers are *really* expensive, are very noisy,
fragile (even diode lasers are electrically fragile) and overwhelmingly
monochromatic and specular. Both these solutions for a film scanner produce
a machine that could become living space for the homeless guy at the back
of Editel/LA. (He is presently living in the crate that one of our Onyxs
came in).

Dissolve to black

Scene ends

New scene: Mike Orton's "State-of-the-craft" address:

Fade up to a post house somewhere in Hollywood.

It is the present day, in the background can be heard grunts as telecine

engineers struggle to adjust run/still compensation on a Rank URSA.


By way of contrast, two guys walked into our facility today, and set up a
tabletop scanner measuring 3ft x 2ft x 18ins (without reel boxes), capable
of scanning 8000 by 6000 pixels  at  12 linear bit depth via a tri-linear CCD
array (Kodak again). The damn thing frightened the sh*t out of me, and
served as a grim reminder of why I got the hell out of telecine. Oh, and by
the way, it costs one-fifth of what a nice shiny URSA Gold would run, and
pin-registration is thrown in for free!!

Ladies and gentlemen, this is the future of our precious "craft".
Its time to get used to it.

Mike "trying to avoid the train wreck" Orton





|---------------------------------------------------------|
| "I do not distinguish by the eye, but by the mind,      |
|   which is the proper judge of the man."                |
|         Seneca. 8 B.C.- 65 A.D.                         |
|---------------------------------------------------------|