|
|
Posted by Jukka Aho on 11/13/06 20:13
Teeafit wrote:
> It looks like, without paying even MORE than I did for the
> standards conversion in London (with each field lovingly
> hand-painted by elves) I'm going to be stuck with an inferior
> product for the NTSC countries.
Snell & Wilcox "Alchemist Ph.C" was already suggested to you. I have not
seen any material processed by that device myself - or I _might_ have
seen without knowing about it - but at least the company has managed to
market the product as some sort of a "gold standard" of standards
conversion.
The stated difference between Alchemist Ph.C and the old-fashioned
field-blending/field-dropping/field-duplication standards conversion
methods is that Alchemist Ph.C analyzes the material for moving objects
or overall motion, such as pans, constructs a set of motion vectors, and
synthesizes new frames out of thin air. In these synthesized frames,
which were never actually shot with a camera, moving objects reside in
positions that are in-between of their positions in the original frames.
There's no discontinuity in the motion, and there are no several frames
blended to each other. Or so the theory goes, at least.
The technical term for this is "motion compensation". You can read more
about it here:
<http://www.iki.fi/znark/video/>
See the PDFs named "The Engineer's Guide to Standards Conversion" (page
46 and onwards) and "The Engineer's Guide to Motion Compensation".
If I were you, I'd at least get a quote of how much a motion-compensated
standards conversion would cost you when compared to the old-fashioned
(blurry and juddery) field-dropping and field-blending based standards
conversion methods.
--
znark
Navigation:
[Reply to this message]
|