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

Re: MARKETING HELP.. the real scoop



At 10:23 AM 1/9/98, David Fenton wrote:
 Because my
>understanding - and that of Peter and Terry, and of respected people
>who contribute to the TIG like Mike Orton and Howard Lukk to name just
>the ones who have recently commented - is that the Kodak CCD head used
>in Spirit scans colour at 960 pixels per line and luminance at 1,920 pixels
>per line.
>
OK, David, since I'm mentioned in your somewhat contentious posting, my
kneejerk response mechanism has already begun to execute its synaptically
enhanced reply program:

First of all. Leave the sarcasm to those best able to deliver it. There's
been enough Transatlantic flak here so that even the most sheltered of my
American colleagues here know it when they read it. In fact they can very
well distinguish clever sarcasm from the lesser common-or-garden variety.

Second. Just because I poke fun at Philips/BTS folks' disingenuity (sic),
when they refer to specifications for resolution, doesn't mean I
automatically absent myself from their list of potential purchasers, which
I believe you may have inferred. I spent many happy years bitching and
moaning at Cintel, as everyone is of course well aware. Didn't stop me
buying the product, and making a living out of operating and maintaining
them.

The fact of the matter is that those days are *over*. Machines are (or
should be) reliable and good enough to not need endless tweaks and bolt-ons
just to make them work properly. A Post-house could have spent upwards of
$200,000 on improvements on the URSA Gold model, if you take
pin-registration into account. Philips just realized this a bit earlier
than the rest of us. Note that I differentiate between those tweaks, and
local in-house competitive-edge modifications.
On the point of resolution, I'll be the first to concede that a
"22:11:11:11" YRGB situation yields a pretty complex MTF for the output
RGB. It is neither 960 pixels (lines?) of RGB resolution, nor is it 1920
pixels of RGB resolution. It does mean that the output will be more
tailored towards delivering HDTV type images, but certainly *does not
preclude* the output images being used in the 2k "data" world. We've known
for years that filmmakers will accept 2k by 1.5k scanned images for film
effects, and many movies even have 601-resolution images "rez'ed up" and
filmed out, fairly seamlessly integrated with the rest of the images.

Now as to your later comment:
>
>PS: Can anyone tell me what you can do with real-time 2K data?  Can
>anyone realistically justify the colossal hardware investment required
>to handle such vast quantities of data per second in the commercial
>environment of post production?
>
The short answer to the latter half of the question is yes.

That real-time 2k data represents only about 1.8 times (less than a third
of an order of magnitude, for physicists!) the data rate of an HDTV 1080i
system:

Calculation: 2048 x 1536 x 10 x 3 x 24 bits/sec for data = 2.265GBit/sec
             1920 x 1080 x 10 x 2 x 30 bits/sec for HDTV = 1.244GBit/sec

On this side of the Atlantic, many people are acting and planning like HDTV
is a done deal. In Post, there is always a potential justification to be
able to perform a given task faster. It's not that much of a leap.

I do not see the Cintel/Philips competition as a duel, with the best man
winning; rather I welcome the competition as being nothing but good for us
poor sap consumers, the net result of which is to advance the art. I have
no doubt that the appearance of the Spirit has spurred Cintel on to better
solutions and engineering.

Best to everyone

Mike O


|---------------------------------------------------------|
| "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.                         |
|---------------------------------------------------------|



---
thanks to Rod Wayne and SanLab Systems for support in 1998
no product marketing allowed on the main TIG.  Contact rob at alegria.com
TIG subscriber count is 907 on Sun Jan 18 17:28:29 PST 1998
complete information on the TIG website http://www.alegria.com/tig3/