Motion video signal processing for recording or reproducing – Local trick play processing – With randomly accessible medium
Reexamination Certificate
2001-04-27
2002-11-26
Tran, Thai (Department: 2615)
Motion video signal processing for recording or reproducing
Local trick play processing
With randomly accessible medium
C386S349000
Reexamination Certificate
active
06487361
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to video post-production editing systems, and in particular to editing systems which are used to edit program material taken from standard 24-frame-per-second film.
2. Related Art
Movies and videotapes give the illusion of motion by playing back rapidly a sequence of still images called “frames.” Commercial movies are designed to be played back at 24 frames per second. In contrast, under the American NTSC television standard, videotapes are designed to be played back at 29.97 frames per second. Under high-definition video standard SMPTE-240M, high-definition videotapes are designed to be played back at 30 frames per second. In Europe, under the PAL and SECAM television standards, videotapes are designed to be played back at 25 frames per second.
Besides the difference in frame rates, video playback differs in another important way from movie playback. In video, the screen image is divided up into a large number of horizontal scan lines. The American NTSC standard provides for 525 scan lines on the screen (not all of which are visible), whereas the European PAL standard provides for 625 scan lines. In videotapes, the scan lines that correspond to each frame are divided into two “fields,” the first containing the even-numbered scan lines and the second the odd-numbered scan lines. On videotape, to record a frame, one records first one of the fields and then the other field. Thus, a video frame contains two fields, and an SMPTE-240M videotape is designed to be played back at 60 fields per second, corresponding to 30 frames per second.
It is common to move program material from motion picture film to videotape. A machine which can perform this function is called a “telecine.” For further background on telecines, see U.S. Pat. No. 5,428,387, issued to the present inventors. Nowadays, it is also common to move program material from videotape to film, particularly inasmuch as high definition video equipment allows this operation to be carried out without degradation in image quality. Either process requires a way of bridging the gap between the 24 frames per second used in film and the 25, 29.97, or 30 frames per second used in videotape.
When converting 24 frame per second film to 29.97/30 frame per second videotape, the standard process used to bridge the frame rate gap is called “3-2 pulldown.” In this process, even-numbered frames of the film are recorded as two successive fields, which are identical (except that they correspond to different sets of scanlines). Odd-numbered frames of the film are recorded as three successive fields, which are also identical (except that they correspond to different sets of scanlines). Thus, every pair of frames in the film, which takes {fraction (1/12)}th of a second to play back at the film playback rate of 24 frames per second, is converted into five video fields, which also take {fraction (1/12)}th of a second to play back at a video playback rate of 60 fields per second.
As depicted in
FIG. 1
, videotapes derived from a film
100
by means of 3-2 pulldown consist of a sequence of groups of fields such as
125
,
130
,
135
and
140
, each field group corresponding to one film frame such as
105
. There are actually four different types of field groups, which we call A, B, C, and D. An A field group consists of two fields corresponding to one videotape frame and to one film frame. A B field group consists of three fields, the first two belonging to one videotape frame and the third to the next videotape frame, with the three fields corresponding to one film frame. A C field group consists of two fields, corresponding to two different videotape frames but only one film frame. A D field group consists of three videotape fields, the first belonging to one videotape frame and the other two belonging to the next videotape frame, but all corresponding to one film frame.
It is important to note the following difference between A and C field groups. Even though both consist of two fields corresponding to one film frame, the first field of an A field group is the first field of a video frame, and thus it corresponds to even-numbered scanlines. In contrast, the first field of a C field group is the second field of a video frame, and thus it corresponds to odd-numbered scanlines. There is a similar difference between B and D field groups. The first field of a B field group is the first field of a video frame, and thus it corresponds to even-numbered scanlines. In contrast, the first field of a D field group is the second field of a video frame, and thus it corresponds to odd-numbered scanlines.
A videotape derived from film by means of 3-2 pulldown thus consists of a sequence of field groups in the order ABCD, ABCD, ABCD, etc. Notably, the first two fields of the videotape will be an A field group, and thus it will be possible to know, just from the location of a field within the videotape, whether it belongs to an A, B, C, or D field group. In modern videotape editing, the frames are referred to by “time codes,” which take the form hh:mm:ss:ff, where hh denotes hours, mm minutes, ss seconds, and ff frames (not fields). Because there are six sets of A, B, C, and D field groups in each second's worth of videotape, each set of field groups encompassing five video frames, it is possible to know just from the video frame number ff of the time code whether the fields of that video frame belong to A, B, C, or D field groups. In particular, the first video frame of each second consists of two fields forming an A field group, the second video frame consists of two fields that make up part of a B field group, the third video frame consists of one field from a B field group and one from a C field group, and so forth.
It is common, in the process of editing videotape, to cut, wipe, or dissolve between two or more sources of video. Post-production video editing systems of the types known in the art permit this cutting, wiping, or dissolving to be performed at any video frame boundary.
Consider a typical editing scenario, depicted in FIG.
2
. The user of the editing system has two source videotapes, tape X and tape Y, derived from film by means of 3-2 pulldown. The user wishes to copy the first six video frames, numbered
0
through
5
in
FIG. 2
, from tape X onto the output, followed by five video frames, numbered
0
through
4
, from tape Y. Because both tapes were obtained by 3-2 pulldown, tape X's video frames
0
through
5
come from five film frames and have an ABCDA sequence, while tape X's video frames
0
through
4
come from four film frames and have an ABCD sequence. If this operation were carried out on a prior art editing system, first the six video frames
0
through
5
in tape X will be copied to the output, which will give the output tape an initial ABCDA sequence. Then the five video frames
0
through
4
from tape Y will be copied to the output. This additional copy operation will result in the output having an undesirable irregular ABCDAABCD sequence, as shown in FIG.
2
.
Suppose now that the operation depicted in
FIG. 2
were a dissolve rather than a cut, so that contents of the two tapes are blended, with an increasing proportion of tape Y content being used as time advances; It is readily seen that the output video frame numbered
7
consists of two fields. The first field of output frame
7
is derived from tape X's B film frame (spanning the video frames numbered
6
and
7
) blended with tape Y's B film frame (spanning video frames numbered
1
and
2
). The second field of output frame
7
, however, is derived from a different film frame on tape X, the C film frame spanning video frames
7
and
8
. This is also undesirable.
As this example shows, even if the source videotapes X and Y are derived from film via 3-2 pulldown and have perfect ABCD sequences, an editing operation can potentially result in irregularities in the ABCD sequence of the output videotape. These irregularities can give rise to visible artifacts when the out
Galt John J.
Pearman James B.
Haverstock & Owens LLP
Omaku Christopher
Sony Corporation
Tran Thai
LandOfFree
Method and apparatus for changing phase of video program... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for changing phase of video program..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for changing phase of video program... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2990415