M
Mark Rejhon
Guest
Pretty good credentials. The stuff I am talking about is not normallyThanks, I guess......
Sorry I omitted a detail, I REALLY thought since I specified the
(NTSC) it would be clear.
I have only designed displays since 1963, so I could use a LOT of
help.... I'm sure!
stuff you need to know in the 1960's, 1970's or even 1980's.
However, these days, it is important to be familiar with the fact that
a single TV frame can contain two different images in the two fields
(even scanlines and oddscanlines). It's common knowledge against 21st
century designers. It also complicates the design of line doublers,
which has to interpolate between the lines ("guess the pixels" so to
speak).
You've probably heard of Sony/Toshiba/Whatever's marketing
terminologies of things like DRC, PureProgressive, XBR, and whatnot.
Some of these technologies digitally processes NTSC material and
digitally upconverts the 480i image (525i, including vertical blanking
interval), to HDTV modes such as 720p and 1080i, and does all the
interpolation between scanlines. If you study the C/C++ source code
of one of the open-source versions of these interpolators (at least
480i -> 480p), http://www.dscaler.org , you suddenly realize things
are much more complicated today, since there's so much internal
processing done inside a typical HDTV set today. Don't get me
started on the digital equipment that cable companies and broadcasters
use these days, either...
These two-fields-in-one-frame is a necessary item of knowledge mantra
for upconverter designers, HDTV television, digital editing (since
you're importing interlaced video for a progressive-scan computer
display), and so on.
This knowledge is also mandatory among the chip designers of DVD
players -- where DVD is normally stored as interlaced video, and it
also gets more complex for these chip designers, when it is needed to
playback these interlaced video as progressive scan (in
progressive-scan DVD players).
Anyway, all NTSC TV's since the invention of NTSC have always been
capable of 60 distinct image per second via alternating fields (even
vs odd fields -- although some day high vs low fields, pick your own
preferred terminology), if the video camera captures a new image all
over again between fields (nearly all of those used for news and
sports TV do, at least). It's also why you notice fast motion is
typically much smoother during broadcast news/sports than via 24fps
film -- to the human eye, temporally, you're really comparing 24
versus 60 ... not 24 versus 30.
Thanks,
Mark Rejhon