July 11th, 2003, 07:40 AM
Heres a question for you who can not go to sleep without reading SMPTE white papers all night long:
Why is it that the Panasonic DVX100 is a PAL or an NTSC camera when it is in fact shooting "p" material?
The very description of NTSC/PAL is an interlaced format. 50i and 60i.
It's so confusing describing something as PAL "p". Progressive formats are usually defined by resolution. PAL should be 576p and NTSC should be 480p. To be precise it should say 480p non square. The DVX100 is a 576p/480p 30p/25p/24p camcorder with the OPTION to shoot regular PAL or NTSC.
Am I making sense?
July 11th, 2003, 07:47 AM
Could it be that the final data stream recorded onto tape is infact interlaced PAL/NTSC just like the XLs frame mode. For it to be viewed on a standard TV it has to be interlaced and formatted. I'm not sure on how the camera works so I'm just taking a guess.
July 11th, 2003, 08:10 AM
Basically what they mean with p (in this case):
1) shot on (semi) progressive chips
2) stored as interlaced
DV only knows interlaced footage. What they simply do is eliminate
the small time difference inbetween the fields.
The signal is simply send to the TV one half at a time. Technically
this creates a small time seperation between the two halfs which
you could call inverse-interlaced or something.
What most people etc. are referring when talking about PAL and
NTSC is describing where the footage originated (america instead
of europe for example) and indiciating a resolution. Usually it
includes a framerate as well (and it definitely should in the correct
technical use), but how many people know the difference between
25 frames per second or 50 fields per second?
Most people seem to talk in frames and that is probably why it is
still called PAL and NTSC. Technically they are only 66% correct
(or something) in using that term....
July 11th, 2003, 10:16 AM
Aha! So we're not talking true progressive here. Keeping two fields might look good to the eye but it puts postproduction back to the interlaced stage which in fact doubles rendertimes in apps like AE. This was actually the only thing that got me interested in the DVX100 -- and now i discover it's still interlaced.
How can it be so hard to make a true progressive camera that delivers one clean image at a time?
July 11th, 2003, 10:46 AM
The image is completely clean and progressive in your computer,
Martin. Only when your sending a progressive signal out to a TV
does it become inversed-interlaced so to speak. Your progressive
image is stored in two seperate fields in DV, but that does NOT
make it interlaced!!
Perhaps my 2) stored as interlaced wasn't the best way to put
it... Perhaps: 2) stored as two fields would have been better.
It is just a different way of storing things, no quality loss or
other weird things.
July 12th, 2003, 03:40 AM
It's all getting clear to me now. Very helpfull, Rob. Thanks a bunch.
July 14th, 2003, 07:33 AM
Well I think that the sampling rates are different for the PAL and NTSC versions (4:2:0 vs. 4:1:1.) That's one difference.
July 14th, 2003, 07:36 AM
Correct, but that is only in regards to DV (they choose those