Matrox Electronic Systems Matrox RT.X2 - Manual (Page 181)

173
Advanced 2:3:3:2 pulldown
Performing a standard reverse pulldown (2:3 cadence) reverses the process by
converting 29.97 fps interlaced video to 23.98 fps progressive video. This is
achieved by discarding a video frame (two video fields) from the sequence. From
the diagram above, a typical standard reverse pulldown would discard fields b3
and d1. This means that video frames C and D must be read by the application to
remove the additional fields. For standard reverse pulldown that is performed on
compressed video, the application must decompress frames in order to remove
the additional fields, and then it must recompress frames to complete the
conversion.
Matrox RT.X2 implementation of standard
reverse pulldown
In order for Matrox RT.X2 to properly perform a standard reverse pulldown, the
material that is to be converted from 29.97 fps interlaced video to 23.98 fps
progressive video must have been originally acquired at 23.98 fps. When Matrox
RT.X2 performs a standard reverse 2:3 pulldown, the process identifies the A
frames of a sequence as those frames that have time code ending in 0 or 5. For
example, a frame with the time code 00:00:20:20 or 00:00:20:25 would be
considered an A frame, and therefore would be used as the first frame in a
pulldown sequence.
In the case of batch captures, when the first frame of a clip that is to be captured
might not always be an A frame, Matrox RT.X2 advances to the next instance of
an A frame in the clip to start the pulldown sequence (if it’s required). In order
for Matrox RT.X2 to perform standard reverse 2:3 pulldown accurately, the clips
must have A frames appearing at time codes 0 and 5. The clips must also be
recorded to a tape that has continuous, non-drop frame time code.
Advanced 2:3:3:2 pulldown
Advanced 2:3:3:2 pulldown was developed as an alternative to the standard
pulldown method for making frame rate conversions in DV video only. It is used
to convert 486p @ 23.98 fps video to NTSC (486i @ 29.97 fps) video. This
conversion is used when video footage is acquired by a DV camera at 23.98 fps
and recorded to DV tape with the recorder set for advanced pulldown (called
24PA or 24P Advanced mode on some devices). Advanced 2:3:3:2 pulldown
differs from standard 2:3 pulldown in that the two middle frames of the video
sequence are each converted to three video fields.