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

Digital Betacam and noise



Dave meant to send this to the list, I'm hereby forwarding it.

--- Forwarded mail from "Dave Corbitt" <dc at mte.com>

        Reply to:   RE>Digital Betacam and noisy composite encoding

I was cleaning out my e-mailbox and came across this from October.
On 10/27/95 7:12 PM gary shaw <gshaw at direct.ca> wrote--->
>  Subject: Digital Betacam and noisy composite encoding. 

>  Why is it that this type of noise is most apparent in blue and
>  particularily in bright white areas of the image. 16 mm neg seems to
>  be the largest culprit.Some one in a previous post mentioned rounding
>  via the renaissance 8:8:8.

Gary,
Noise is not usually an artifact of the Digital Betacam format but a problem
of the engineering compromises that had to be made in the design of Rank
telecines.  The Blue channel is particularly noisier than Green due to a
combination of the cut off wavelengths of the Blue color separation filters in
the cell box and the good job the orange mask of negative film does in
removing most of the blue light as it passes through the film.  It also
relates to the spectral distribution of energy from the CRT which has a big
broad peak in the central green area of the spectrum but drops off
considerably in Blue, (not so much in Red but still a big roll off of light
there). The resulting lack of light reaching the Blue PEC forces us to run the
Blue PEC at a very high sensitivity level and brings up the noise.  By
increasing the light that gets to the Blue PEC the Blue signal to noise can be
drasticallly improved.  This can be done a number of ways;
1.   Get a brighter CRT.
2.   Change the cut off wavelength of the Blue Trim Filter in the cell box to
pass more light to the PEC then correct the errors with a new masking matrix.
3.   Install a faster gate lens that passes more light and/or Rank's
"Highlight Kit".  16 mm gate lenses are usually slower than 35mm gate lenses
and further away from the CRT.  Both of these factors reduce the light
throughput and thus the signal to noise.  
(Note, increasing the Blue light throughput can cause the Blue PEC to
saturate, so the Blue PEC may have to be replaced with the same type used in
Green).

URSAs also have the additional problem of an overall background noise in the A
to D converters.  If the signal out of the Cell Box is not maximized before
the A-D conversion then the A-D noise becomes very objectionable.

The reason the images are noisier in bright areas of a picture derived from a
negative film image is that is the densest part of the film.  The densest area
is then inverted to create the highlights of a positive image  and the gamma
curves are inverted to create a useful tonal scale.  The highlights are
stretched in images derived from negative so the noise will be exaggerated the
most in burned out highlights unless the signal can be optimized by attempting
to do as much as possible of the  steps listed above.  

I hope this isn't too simplistic.
Dave Corbitt


--- End of forwarded message from "Dave Corbitt" <dc at mte.com>




-- 
Rob Lingelbach          | 2660 Hollyridge Dr LA CA 90068 213 464 6266 (voice) 
rob at xyzoom.alegria.com  | "I care not much for a man's religion whose dog or 
rob at sun.alegria.com	|  cat are not the better for it."  --Abraham Lincoln
rob at dagmar.alegria.com        KB6CUN	   http://www.alegria.com