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

DVD



Ok seems to be my day at the keyboard.

With consideration  that my experience in this particular field is 10months
outdated I can add the following.

Keep in mind that DVD is only the carrier of coded information, thus it is the
coding process (perhaps the decoding process is playing a small role here) that
is the focus of the discussion. 

The percieved resulting quality greatly varies with hardware/software used even
though the tools may be using to the same coding standard. It also varies
greatly where and by whom the coding was made. Let me explain this. I dare say
that  no fully automated process or device (real or non realttime) exists that
can receive telecine graded source material and out put optimally coded DVD. In
fact, it may even be so that in the forseeable future, coding for DVD is going
to be a process for specialists much like fmporal only, spatial or colorimetric or???
 
Mike Reichel 

ocate for coding beeing carried out in the telecine
domain/facility as it is there the infrastructure exists. There is another very
good reason for this, if DVD is going to be a commercial success, the quality
must be very good therfore the coding machinery today REALLY needs good
operators (compressionists) to acheive this. What better would a colourists be
with a paid eye for quality. 

Returning to the original question posed by Rob plus deduction made from above,
a significant variable can be found in the complexity of source material. A big
difference is simply found in whether it is film or videooriginated (brings us
back to Rich Torpeys request for a field correlator). The latters higher
temporal information density makes it potentially more demanding in coding
whereas a letterboxed film with long scenes and little motion is not (Mr Bergman
may be pleased).

It shall be mentioned that a real booster to quality in coding is carried out in
the process of just  before actual coding, a sort of preprocessing
where the source material is deliberately limited in bandwith/detail and in
random artifacts (different types of noise). This is done dynamically  scene by
scene or when requiredon a  frame by frame basis.

So my bet is that DVD is as good as the effort provided in the coding process
and the material used. Be sure that any demo you see will have carefully
selected material. 

What we have to look forward to are colorists increasing their level of skill
plus newer and smarter preprocessing equipment that is better integrated into
todays suites. Maybe will larger DVD recording densities (comming up next) allow
a slightly more relaxed degree of coding thus allowing for higher quality, wider
aspect ratios etc...

If it is better than D1? Never in my mind. Give any D1 recording the same care
to detail as you must to DVD coding and you will boost D1 too!  
Mike Reichel