Panasonic’s 720p 23.98 – a fake format

[Shane Ross]

[aaronowen] “Finally some specific info! Thanks so much!! I will be testing this over the weekend and will post the results…”

That is why these discussions work. LOTS of voices saying things, and more voices to interpret this information and take ideas and add their own. So Gary gave the answer, but it took more people to interpret the answer and say it in a way you understood.

Or sometimes tempers can flare and we brush over the useful information. Or they say SO much that you gloss over it…been there…

Shane

Littlefrog Post
http://www.lfhd.net

[Matt Silverman]

Duh.

[Charles Caillouet]

aaronowen,

I am sorry that your thread turned sour and i hope that you got your problem solved. I have not been monitoring the forum for a while but i ran across this discussion today and thought that it deserved some clarification. So here are some comments on the misunderstandings running rampant in this and other threads about 720p24.

1) 720P23.98 as a fake format
It is a real format in Final Cut Pro and in Quicktime files.
It is, in fact defined along with 720P24 in the ATSC Table A3 (which was not approved by the FCC).
It is not defined across firewire or HDSDI or in any SMPTE standard and no video recorders can handle it.
You just have to get it around in other ways than the two above, or convert it to some other format.

2) 720P23.98 or 720P24 (the common terminology for 720P23.98, which corresponds to 24 when the camera is running at a system frequency of 59.94) do not use advanced pull down or 2,3,3,2. That is a cadence used for interlaced video to make it easier for non-linear editors to pull out the duplicate fields, assuming that they know how do deal with it. Some don’t.
Also, advanced pull down should never be used for distribution, only from a camera to the editor, and only then, with care.
At any rate, it has nothing to do with progressive recordings so that is a red herring in this discussion.

By the way, most recordings are made at 59.94 or 23.98 to simplify editing and audio layback, even if a film out process is anticipated. The conversion to 24.00 is made at the very end of the chain. When i write 24, 30 or 60, i mean 23.98(actually 23.976), 29.97 and 59.94.

3) The metadata that you refer to is simply a string of bits in the user data part of the SMPTE time code associated with the clip in question. It is carried with the time code numbers but the editor needs to be cadence-aware to know to put it in a time code stream.

4) When you drop a clip onto a sequence track, you retain a link to the original time code for edit decisions but you assign a new time code for the finished sequence. If the sequence is 60 (or 59.94) fps, then a cadence is implied for all 24 fps material included and it is incumbent on the editor to keep track of the 2,3 cadence. It is possible to do this on a 60 fps sequence but it is not trivial. The accepted standard for doing this is to keep track of “A” frames, the start of the 2,3 cadence.
Many editors and software utilities know about A frames and can extract a 24P sequence when the cadence is intact (no edits which change the cadence,) even without other flags. This is commonly referred to as “reverse telecine” and has been used for film transfers through “telecines”. A smart application can even look at the images and detect the cadence by finding repeated frames.
If you convert the clip to 24 (or 23.98), through any number of methods, then you can drop the clip on a 24 fps sequence track without regard for cadences. All cadence information is lost in a 24 fps sequence because there is no cadence. Even the off-speed effects clips have been converted before they are dropped into the sequence.

5) If you want to export the sequence at 24 fps, you have several options.
One is to keep it in file form as 720p24, without cadence issues. You can move this over a network or on a drive to another system or application that knows about that format and the compression scheme that is used to represent it. This works for compressed or uncompressed clips. The receiver just has to know how to decode it.
Another is to export it as a 1080p24 clip to a recorder that can handle that format. The most obvious are HDCam and D5. Note that not all HDCam and D5 decks can record and play 24 fps. 1080P24 does not contain a pull down cadence, only the unique frames. It may be actually recorded and displayed as segmented frames for operational reasons, but there are no duplicate fields or frames.
You can also chose to recreate a cadence and export the clip as 720p60, 1080i30 or 480i30. When you do this, you are recreating a new cadence, not retaining the original one.

6) The comment about many workflows is very true. There are as many as there are users and edit systems. Panasonic will suggest some options based on your configuration but when you mix systems, you are required to learn something about the capabilities of the systems and the signals that can move among them.

Ranting on either side is not helpful.

cheers,

Vision Unlimited/LA
Prairieville, LA
HD production technical support since 1987
…searching for the right tools for the job…