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

FWD>RE>a few issues from Da

Mail*Link(r) SMTP               FWD>RE>a few issues from Dave Corbitt

Kevin, In reply to some of your statements,  I would disagree with some of it.
 Below are my comments for your perusal.
Dave Corbitt

Kevin wrote....
>Date: 5/17/95 1:56 AM

>Lets talk about noise. I would tend to agree that a well tuned Mk 111 can
>quieter than a raw Ursa. Although I have to say many of the noise problems I
>have encountered have been at least in part attributable to poor grading

What Grading practices cause noise?  It's my experience that noise is
generated in Photomultipliers that have insufficient light input to produce a
clean signal, thus the higher noise on the Blue PEC than the Green. And
cleaner signals at Higher Beam Current than at Lower Beam Current.  It appears
after installing Tube Enhancement's replacemant CRT for URSA with its 1.8
times greater light output than the Rank Brimar CRT that better Signal to
Noise can be achieved by increasing the amount of light available.  But there
is still a noise factor accounting for less of an improvement than expected if
noise were only being generated in the PEC's.  My best guess now is that the A
to D converters are in need of major reworking to get a true 14 bit conversion
rather than the double flash conversion that probably generates a lot of the
remaining noise.

>Nevertheless, it seems that the cost of more color separation and
>sharper pictures is more noise. 

The color separation filters and optics in the Cell Box for URSA are identical
to a MK3 so color separation of the two machines should be the same. I don't
think URSA is sharper than a MK3, in fact the roll off used in the URSA
Digital color Channel actually reduces resolution at the higher frequencies
which is then "compensated" for by a fixed amount of aperture correction that
can not be turned off by the user without pulling out IC's. Maybe GOLD doesn't
do this but the Gold machine has the same A to D's as the standard URSA.

>Some people advocate defocusing the tube to
>reduce noise, but then you do not have sharp pictures anymore either. 

Noise shouldn't be dependent on focus of the CRT unless we are confusing the
resolving of film grain with electronic noise. Or the decoding of tiny detail
such as textural qualities of fabrics, etc. by a composite monitor as "Cross
Color" on the monitor.

>Also as
>our colour correction capabilities expand, we have more ways to amplify and
>exaggerate noise through the system. I reckon you can match a mk 111 with an
>Ursa but not the other way round.

>More to the point though,(sorry I'm rambling) I would refuse to use an Ursa
>without a noise reducer. You 've seen the recent discussions on artifacts. I
>use a DVNR 4x4 (which I recommend, in spite of giving Michael a hard time)
>believe I can remove any noise left in the picture without loss of picture
>information, or artifacts. I agree it needs setting up, there's no magic
>switch, :)
Wouldn't it be preferral to see the scanned film image with no electronic
noise added by the telecine at least to the limits of the particular signal
system? That is my hope for the near future. Then "Noise Reduction" would
really be "Grain Reduction". Sorry Kevin, I'm not getting on your case, I just
feel it's time to clean up the URSA signal path. I feel noise shouldn't be a
part of the signal anymore and I view it as a challenge to tackle and resolve.
Best Regards, Dave Corbitt/Manhattan Transfer/NYC