HK1051435B - Method for editing to a picture a group of pictures in a media without decoding and encoding - Google Patents

Description

Method for editing pictures in a group of pictures in a medium without decoding and encoding

Technical Field

The present invention relates generally to a method and apparatus for providing advanced operational functionality for audio programs, video programs and programs made up of both audio and video recorded on disk media such as recordable digital video disks, hard disks, and magneto-optical disks.

Background

A variety of devices have been developed to enable customers to record video and/or audio programs for later playback. Such devices include tape recorders, video tape recorders, recordable compact discs, and most commonly available today as recordable Digital Video Discs (DVDs). Hard disks and magneto-optical disks have also been used.

A DVD, which is recordable only once, is essentially a DVD read only memory, given its acronym DVD-R. The acronym DVD-R may also generally refer to write-once, or record-once, technology. There are several formats available for recording, erasing or re-recording of DVDs; i.e. flushing overwriting or rewriting. These are indicated by the acronyms DVD-RAM, DVD-RW and DVD + RW. No uniform industry standard has been adopted so far. The acronyms DVD-RAM, DVD-RW and DVD + RW are commonly used to refer to various rewritable techniques. The rewritable DVD technology, apparatus and methods referenced herein are intended to generally encompass all standards that are in use now and that may be developed in the future.

In many cases, programs are recorded when the viewer and/or listener are absent for later playback at a more appropriate time. This is called a time-shifted program. At another time, the program is viewed and/or listened to without recording the program, and there is no interest in the recording, but the attention of the viewer and/or listener is interrupted by, for example, a telephone call or an unexpected visitor. A television program may be recorded if the viewer and/or listener is watching the program and has a cassette in a VCR (video cassette recorder), for example, or the viewer and/or listener can quickly retrieve and load such a cassette. However, the viewer and/or listener cannot view and/or listen to the entire program and the program in the proper time sequence until the recording is complete. The time for completion of recording can be short or long depending on the length of the program.

Although rewritable DVD technology is generally available, operation is limited to basic functions such as play, record, fast forward, fast reverse, and stop. The pause function is available, but only as a function corresponding to the pause operation of the VCR, such as interrupting the playing of a pre-recorded program or interrupting the recording of a viewed program to remove commercials from the recording. Unlike computer hard disks, recordable DVD devices have a very significant additional function of playing pre-recorded DVDs. Therefore, there is a financial drive to develop rewritable DVD technology, including methods and apparatus, which can be used to replace computer hard disks. It is challenging to provide a device with such improved, advantageous features, regardless of the goals of compromising reduced cost and increased sales.

It is desirable to have some editing capabilities for any recordable DVD device. Customers need these functions in order to be able to make their own home videos, record special events such as weddings, birthday parties, etc. These functions are also useful after the TV program has been recorded. For example, a customer may wish to delete commercials from certain programs, such as sporting events and movies.

In an MPEG (moving picture experts group) bitstream, the smallest access unit is a group of pictures or GOP. In DVD, the smallest access unit is VOBU. Each VOBU nominally comprises a group of pictures (GOP). Also, normal MPEG editing only edits one GOP without decoding and re-encoding the MPEG encoded signal at the picture level. Therefore, most editors today are functionally limited in terms of resolution for minimum GOP editing, which is about 0.5 seconds of resolution. In these types of conventional systems, the user cannot arbitrarily define the start and end points at the picture level resolution to mark the bitstream segments on the editing operations that are performed. In any case, the result is that the edits can be displaced (displace) over at most 0.5 seconds from the user-selected start and end points.

In general, the above limitations result in the complexity associated with decoding and re-encoding DVD video data to a single picture level. DVD video is highly compressed in order to make the most use of the disc space possible. The most commonly used compression techniques involve storing a single picture with substantially full detail and then only the changes associated with subsequent pictures. This technique uses three pictures, commonly referred to as intra pictures (I pictures), predictive pictures (P pictures), and bi-directional pictures (B pictures). An I picture is a reference picture in which the entire picture content is compressed and stored. An I picture is considered as a reference picture that can assemble other pictures. I pictures can be accessed directly within the video stream because they contain all the data needed to render the picture. By comparison, P pictures alone cannot be used to reproduce a picture. Instead, such a picture contains motion vectors describing a difference from the nearest preceding I-picture or P-picture. Finally, B pictures are encoded by looking forward or backward at the next I or P picture to match the macroblocks that form these I and P pictures.

The following is a typical GOP structure of a 15 picture GOP without detelecine in display order: b is₀B₁I₂B₃B₄P₅B₆B₇P₈B₉B₁₀P₁₁B₁₂B₁₃B₁₄. However, other GOP structures are possible and may be used. Regardless of the exact GOP structure used, B pictures cannot be accessed at will because such B pictures cannot be properly decoded without adjacent I and P pictures. Similarly, a P picture cannot be properly decoded without reference to the immediately preceding I picture in the bitstream. This difficulty creates a particular problem in using editing functions in recordable DVDs, since it means that some pictures within the bitstream cannot be executable apart from adjacent pictures.

Disclosure of Invention

A method of editing pictures in a group of pictures (GOP) in a DVD medium without decoding and encoding an MPEG encoded signal, the method comprising the steps of: marking at least the user-selected ending picture to identify the selected segment of the encoded signal to be edited; if the user-selected end picture is not one of the at least I picture and the P picture, automatically identifying an actual end picture of the selected segment based on a position of the user-selected end picture relative to at least one other picture within the encoded signal by designating the actual end picture as the one of the at least I picture and the P picture that is closest to the user-selected picture; performing an editing function on an actual segment defined by an actual end picture, without decoding and encoding the MPEG encoded signal, wherein the editing function can be performed on the picture; and blanking the actual end picture if the user-selected picture is not one of at least an I picture and a P picture.

The method further comprises the steps of: any B pictures that appear between the user-selected end picture and the actual end picture are deleted. According to one aspect, the step of identifying the actual end picture includes designating the actual end picture as the user-selected end picture if the user-selected end picture is one of at least an I picture and a P picture.

The method may further comprise the steps of: marking a user-selected starting picture to identify a selected segment of the encoded signal to be edited; and automatically identifying an actual starting picture of the selected segment based on a position of a user-selected starting picture relative to at least one other picture within the encoded signal. The identifying step for the actual start picture may include: if the picture selected by the user is a B picture or a P picture, the actual start picture is designated as an I picture immediately preceding the picture selected by the user. Furthermore, the method may comprise the steps of: any B pictures that appear between the user-selected picture and the actual start picture are deleted. In this regard, the method may further comprise the steps of: blanking any I pictures and any P pictures that occur between the user-selected picture and the actual starting picture.

According to one aspect, the identifying step for the actual start picture may comprise: if the user-selected start picture is an I picture, the actual start picture is designated as the user-selected start picture. According to another aspect, the identifying step for the actual start picture may comprise: if the picture selected by the user is within the last two pictures of the GOP containing the selected picture, the actual starting picture is designated as an I picture in the subsequent GOP.

In an alternative embodiment, the present invention may include a rewritable DVD device for editing pictures in a group of pictures (GOP) in a medium without decoding and encoding an MPEG encoded signal. The apparatus comprises: means responsive to user input for marking at least the end picture selected by the user to identify the selected segment of the encoded signal to be edited; processing means for automatically identifying an actual end picture of the selected segment based on a position of a user-selected end picture relative to at least one other picture within the encoded signal, including means for designating the actual end picture as one of at least an I picture and a P picture that is closest to the user-selected picture if the user-selected end picture is not one of at least a 1 picture and a P picture; editing means for performing an editing function on an actual segment defined by an actual end picture, which can be performed on a picture, without decoding and encoding the MPEG-encoded signal; and means for blanking the actual end picture.

The apparatus can delete any B picture that appears between the user-selected end picture and the actual end picture. The system may also designate the actual end picture as the user-selected end picture if the user-selected end picture is one of an I picture and a P picture.

The apparatus may also mark a user-selected start picture to identify a selected segment of the encoded signal to be edited in response to a user input; and automatically identifying an actual starting picture of the selected segment based on a position of a user-selected starting picture relative to at least one other picture within the encoded signal. If the picture selected by the user is a B-picture or a P-picture, the apparatus may designate the actual starting picture as an I-picture immediately preceding the picture selected by the user. The apparatus can also delete any B pictures that appear between the user-selected picture and the actual start picture. The device may also blank any I pictures and any P pictures that occur between the user-selected picture and the actual starting picture.

According to one aspect, if the user-selected start picture is an I-picture, the rewritable DVD device can specify the actual start picture as the user-selected start picture. Further, if the picture selected by the user is within the last two pictures of the GOP containing the selected picture, the rewritable DVD device can specify the actual starting picture as an I picture in the subsequent GOP.

Drawings

Fig. 1 is a block diagram of a rewritable DVD device capable of providing one or more advanced operational functions according to an aspect of the present invention.

Fig. 2 is a schematic diagram for explaining a spiral track of a rewritable DVD.

Fig. 3 is a schematic diagram for explaining the structure of a video disc.

Fig. 3A is a typical GOP structure of 15-picture GOPs in display order for explaining picture processing editing.

Fig. 4 is a table of ending screen transition for editing by sub-explanation screen processing.

Fig. 5 is an end screen transition table for explaining another embodiment of screen processing editing.

Fig. 6 is a start picture transition table for explaining picture processing editing of an open (open) GOP.

Fig. 7 is a flowchart for explaining screen processing editing.

Detailed Description

Recordable DVD device

An apparatus 100 implementing various high-level operational functions according to an aspect of the present invention exemplified herein uses a rewritable disc medium 102 according to an aspect of the present invention as depicted in the block diagram of fig. 1. The rewritable disc medium 102 is used as a rewritable DVD in the illustrated embodiment. In many of the examples described later, the rewritable disc medium may also be, for example, a hard disk or a magneto-optical disc (MOD). An example of a MOD is a small disc. In many instances, aspects of the present invention may be applied to video, or audio, or both video and audio.

The device 100 is capable of writing to and reading from a disc medium, in this example a rewritable DVD 102. The device includes a mechanical assembly 104, a control portion 120, a video/audio input processing path 140, and a video/audio output processing path 170. The arrangement of most of the blocks into different parts or paths is self-evident, as the arrangement of these blocks is for convenience and not critical to understanding the operation of the device.

The mechanical assembly 104 includes a motor 106 for rotating the DVD102 and a pickup assembly 108 adapted to move over the rotating disk. A laser on the pickup assembly burns a spot on a spiral track on the disc or illuminates a spot already burned on the track to record and playback video and/or audio program material. For the purpose of understanding the invention, it is irrelevant whether the disc is single-sided or double-sided recordable, or in the case of double-sided recording, whether double-sided recording or subsequent reading of the disc is performed from the same side of the disc or from both sides. The pickup and motor are controlled by a servo 110. The servo system 110 also receives as a first input a playback signal of data read from a spiral track of the disc 102. The playback signal is also input to an error correction circuit 130, which error correction circuit 130 may be considered part of the video/audio output processing path or part of the control section.

The control section 120 includes a control Central Processing Unit (CPU)122 and a navigation data generating circuit 126. Control CPU 122 provides a first input signal to navigation data generation circuit 126 and servo 110 provides a second input signal to navigation data generation circuit 126. The servo system may also be considered as an integral part of the control section. The navigation data generation circuit 126 provides a first input signal to a Multiplexer (MUX)154, the MUX 154 forming part of the video/audio input processing path 140. The output of MUX 154 is the input to error correction coding circuit 128. The output of the error correction coding circuit 128 is a recordable input signal that is provided to the pickup 108 that will be "burned" onto the spiral track of the disc 102 using a laser.

The control CPU 122 also preferably accesses the data contained in the track buffer 172 and the recording buffer 152 shown in fig. 1. To implement aspects of the present invention, CPU 122 may delete, modify, and reformat the video data stored in track buffer 172 and record buffer 152. For conventional operations performed by control CPU 122, suitable software or firmware is provided in memory. Further, program routines of the high level functions 134 are provided to control the CPU 122 in accordance with the present invention, which will be described in greater detail below.

The control buffer 132 for the viewer activated function (active function) indicates those functions that are currently available, i.e., play, record, rewind, fast forward, pause/play, and stop. The pause function is a function corresponding to a pause operation of the VCR, such as interrupting playback of a pre-recorded program or interrupting recording of a viewed program to remove commercials in the recording. A separate buffer 136 is provided to receive commands for implementing aspects of the invention described herein.

Video/audio input processing path 140 is a signal processing circuit for converting a conventional television signal, such as NTSC or PAL, into digitized packet data, such as MPEG-1 or MPEG-2, for digital recording by apparatus 100. Input path 140 includes an NTSC decoder 142 and a video encoder 144, e.g., MPEG-1 or MPEG-2, for video, and includes an audio analog-to-digital converter (A/D)146 and an audio encoder 148, e.g., for MPEG-1 or MPEG-2. A sub-picture encoder 188 is provided for generating sub-picture information to be inserted into the data stream. The digitized signals are combined in multiplexer 150 and stored in a record buffer 152 until the entire packet is established. As each packet is created, each packet is merged with the output of the navigation data generation circuit in the MUX 154 and sent to the error correction coding circuit 128. The error correction coding circuit 128 may also be considered part of the input path 140.

The output processing path 170 includes an error correction block 130 and a track buffer or output buffer 172 where data read from the disk is assembled into packets for future processing. These packets are processed by conditional access circuitry 174 which controls packet propagation through a demultiplexer 176 and onto various paths for video and audio processing. Thus, it may also occur that the track buffer 172 only needs to store enough data to correspond to a storage capacity of about 0.5 seconds of audio and video program material.

Decoder 178 decodes video, such as MPEG-1 or MPEG-2 video, and is encoded in TV encoder 180 as a conventional television signal, such as NTSC or PAL. The circuit 182 decodes audio, such as MPEG-1 or MPEG-2 audio, and is converted to analog form by an audio digital-to-analog (D/A) converter 184. Finally, the circuit 186 decodes the sub-picture information that has been read out in the disc. Output processing path 170 may be considered to include error correction circuit 130, as shown.

DVD medium

For purposes of illustrating aspects of the invention, program material may be recorded on and played back from a rewritable DVD. The rewritable DVD10 shown in fig. 2 is suitable for use as a disc 102 in the device 100. The tray 10 is formed from an oblate plastic dish-shaped component 12 with a hole 14 in the middle. The on-track recording direction, from the smaller radius portion of the spiral to the larger radius portion of the spiral, is typically outward along the spiral track 16. Several track sections, not shown in the sequence diagram, consisting of three large dots (- - - -). As a result, the beginning of the spiral track is considered to be proximate the hole 14 and is represented by block 18. The end of the spiral is considered to be near the edge, indicated by diamond 20. The definition of start and end of the illustrated helix is generally accepted by those of ordinary skill in the art. Some advanced functions of the solution according to the invention use backward recording, i.e. from a larger radius part of the spiral to a smaller radius part of the spiral. The track may also have a side-to-side wobble (not shown) to accommodate media type retrieval (index). Due to scale, only parts of the track 16 are shown, and these are shown in greatly enlarged scale.

The radially concentric portion of each spiral that approximates a circle is sometimes referred to as a track, but this term is not generally accepted as having a particular meaning. In CD-ROM, for example, the term track is used to refer to a spiral track, or other portion, containing a single audio song, and for DVD, the term may or may not become common.

Various modifications of the apparatus shown in fig. 1 and the disc medium shown in fig. 2 may be used together to implement the high-level functions of the solution according to the invention exemplified herein. It will be appreciated that the advanced functionality exemplified herein may be applied to other kinds of disc media and disc media players and recorders.

As shown in fig. 3, each DVD includes a video manager 26 and a Video Title Set (VTS) 28. The VTS comprises Video Title Set Information (VTSI)27, a set of selectable video objects 29 for the menu and one or more VOBS 30 for the title which contain the actual title content. Each VOBS 30 comprises a plurality of video objects 32. Each video object 32 includes a plurality of slices (cells) 34. Each VOBS comprises a set of pointers to slices. Thus, the VOBS data links the pieces together and indicates in which order the program or pieces are played. The slices in a particular VOBS can be marked for playback in any desired order. For example, they may be played sequentially or randomly.

Each slice includes a plurality of VOBUs 36. Within each VOBU 36 resides video content typically containing 0.4 to 1.0 seconds of the disc material being displayed. Each VOBU is a sequence of data packets in the recording order. Each VOBU starts with exactly one navigation pack (NV PCK)38 and can include all the following pack types, including a video pack (V _ PCK)39, an audio pack (a _ PCK)40, and a sub-picture pack (SP _ PCK) 42.

The NV _ PCK 38 includes display control information and data search information. Data search information is useful in "trick" modes of performing playback, i.e., modes in which a subset of the entire picture is displayed in order to achieve faster display of material. One aspect of the Data Search Information (DSI) is that it helps the decoder find the reference pictures within the VOBU corresponding to the current NV _ PCK. Another aspect of the DSI is that it helps the decoder find VOBUs that are related to the current VOBU as far as the future or past of the presentation. The past VOBU related to the display of the current VOBU is referred to in the NV _ PCK field known as BWDI (backward info). The future VOBU related to the current VPBU display is referred to in the NV _ PCK field known as FWDI (forward info).

Each VOBS can include navigation commands that allow branching or other interactive functions. These navigation commands may be included as part of the VOBS as a set of pre-commands. These pre-commands may be followed by optional navigation commands identified within the NV _ PCK 38 of the slice 34, which are executed after the slice is displayed. Finally in the VOBS these can be followed by an optional set of post commands. All these commands are stored in a table within the VOBS and can be referenced by numbers so that they can be reused. The navigation commands may instruct control CPU 122 to perform various operations related to branching or interactive functions. For example, they may provide flow control using commands such as goto, link, jump, exit, etc.

Picture editing without decoding and re-encoding MPEG bit streams for recordable DVD

The method according to the aspect of the present invention provides a method of picture editing in a group of pictures (GOP) and pictures within a VOBU on a DVD disc. Fig. 3A is a typical GOP of a GOP of 15 pictures in display order without detelecine. The present invention is described in relation to a typical 15 picture GOP structure. However, it is apparent that other GOP structures such as a 12 picture GOP (M ═ 1 or 2) may be used with this process.

To perform video editing, a user typically needs to identify segments of the bitstream once an editing operation (e.g., deletion) is to be performed. This presents a particular problem because certain pictures, such as B pictures, cannot be decoded without reference to the I and P pictures that occur immediately before and after the B picture in the bitstream. Similarly, a P picture cannot be decoded without a previous I picture.

The solution of the present invention provides a unique solution to the above-mentioned problems. The process begins when the user marks the beginning and ending points of a segment defining the editing operation to be performed on. The exact picture start and end positions in the display that have been marked by the user may correspond to B pictures or P pictures that cannot be decoded independently of adjacent pictures. In that case, the system automatically selects a new start, and the start picture is required when the actual segment is defined when the editing operation is to be performed. Such a selection transformation is illustrated in fig. 4-6.

Fig. 7 is a flow chart illustrating a process according to the present invention with reference to the scheme of fig. 3A and fig. 4-6. The process begins when a user selects a start picture and an end picture that define a video segment on which an editing operation is to be performed in step 700. The selection process may be performed by suitable means that allow the user to mark as segments of the video display. In step 702, if it is determined that the end picture of the editing segment selected by the user is an I picture or a P picture, the system will proceed to step 704, and the end picture selected by the user will be used as an actual end picture. However, if the end picture is a B picture, the system automatically selects the new last picture as an I picture or a P picture immediately before or after the user-selected B picture. Fig. 4 shows a conversion table in which the system automatically selects the new last picture as the I-picture or P-picture that is the closest before or after the user-selected B-picture. Fig. 5 shows a conversion table where the system automatically selects the new last picture as the I-picture or P-picture that is the closest after the user-selected B-picture. If the translation table of FIG. 4 is used, the process can proceed directly from step 706 to step 712. If the translation table of FIG. 5 is used, additional processing is required at steps 708 and 710.

If the system is configured to use an I picture or a P picture that appears after the user-selected B picture in step 706, the system preferably adds a sub-picture to the picture in step 708, thereby blanking the subsequent I picture or P picture during display or playback. In fig. 5, blanked I and P pictures are shown underlined. In step 710, any B pictures that appear between the selected end picture and the actual end picture are deleted. For example, in fig. 5, when the last screen selected by the user is B3, the B4 screen is deleted.

In step 712, if the first picture defining the user-selected segment is not a B picture or a P picture, it must be an I picture. Thus, in step 714, the system defines the actual start picture as the I picture that was selected as the start picture. Processing then proceeds to step 722.

In step 713, if the user-selected starting picture for marking the start of the editing section falls in a later position of the GOP (e.g., B12, B13), it is preferable to skip the GOP entirely. Thus, the system will proceed to step 715 and preferably treat the actual starting picture as the first I picture of the next GOP. Processing then proceeds to step 722.

If the user-selected start picture determined in step 713 does not fall in a later position of the GOP, the system proceeds to step 716. In step 716, the system automatically picks the immediately preceding I picture that is the closest to the user-selected B picture as the new starting picture. This new start picture is then considered as the actual start picture defining the start of the segment.

Fig. 6 illustrates the foregoing process for selecting a new starting picture in step 716 in relation to the exemplary GOP structure shown in fig. 3A. According to a preferred embodiment, any intermediate B pictures that appear between the user-selected picture and the actual start picture are deleted in step 718. These pictures can be deleted using one of the following methods: (1) deleting the MPEG bitstream data; (2) replacing MPEG data with stuff bytes such as 00 or FF; (3) a stream start code (000001 BE) using padding and a 2-byte padding data length; alternatively, PES _ packet _ length is set to zero if the entire packet is to be deleted. In step 720, the I-picture and any intervening P-pictures are preferably blanked during the addition of the appropriate configuration sprite for display. In fig. 6, the blanked picture is underlined. It is noted that the foregoing method allows the display of the first picture selected by the user without having to display the preceding I and P pictures that would otherwise be necessary to decode the B picture. At step 722, an editing function can be performed on the video segments defined by the actual start picture and the actual end picture. Processing is complete at step 724.

The process of picture blanking described herein may be accomplished in a variety of different ways. For example, the main picture can be blanked with the sub-picture by setting the size of the sub-picture equal to the size of the main picture, setting the gray scale blanking of the sub-picture, setting the color of the sub-picture to colorless, setting the contrast to 16 (entity), and setting the display start time of the sub-picture to be the same as the display time of the main picture. If there is more than one sub-picture to be blanked, the display time is set to the sum of the display times of all the pictures to be blanked.

To edit the film structure, it must be determined that the interleaved upper and lower edges of the film match. If they do not match, they can be matched by adding a duplicate upper edge or deleting a duplicate edge, depending on the actual start picture or the actual end picture.

The audio may or may not be related to the edited video. If the audio is related to the video, a lip sync problem will occur. To solve the lip-sync problem, some audio pictures have to be deleted or some or all of the associated audio packets in the VOBU have to be muted by setting audio _ mute _ flag.

E.g., closed caption line 21 data, is associated with the display field of the GOP. The line 21 data is in the GOP header user data field. Because some fields have been deleted, some closed-form data may not be valid. To prevent closed caption errors, the line 21 data needs to be changed or deleted.

It will be readily appreciated by those of ordinary skill in the art that the inventive arrangements for editing GOPs to a single picture level described above are not limited to the precise embodiments disclosed herein. In fact, the solution of the invention can be used not only with DVD media, but also with any other media that makes use of the MPEG compression method.

Claims (9)

1. A method of editing pictures in a group of pictures (GOP) in a medium without the need to decode and encode an MPEG encoded signal, the method comprising the steps of:

marking at least the user-selected end picture to identify the selected segment of the encoded signal to be edited;

automatically identifying an actual end picture of the selected segment based on a position of the user-selected end picture relative to at least one other picture within the encoded signal by designating the actual end picture as one of at least an I picture and a P picture that is closest to the user-selected picture if the user-selected end picture is not one of at least an I picture and a P picture;

performing an editing function on an actual segment defined by said actual end picture without decoding and encoding the MPEG encoded signal, wherein the editing function can be performed on a picture; and

blanking the actual end picture if the user-selected picture is not one of at least an I picture and a P picture.

2. The method of claim 1, further comprising the step of deleting any B pictures that appear between the user-selected end picture and the actual end picture.

3. The method of claim 1, wherein the identifying for the actual end picture comprises:

designating the actual end picture as the user-selected end picture if the user-selected end picture is one of at least an I picture or a P picture.

4. The method of claim 1, further comprising marking a user-selected starting picture to identify a selected segment of the encoded signal to be edited; and

automatically identifying an actual starting picture of the selected segment based on a position of the user-selected starting picture relative to at least one other picture within the encoded signal.

5. The method of claim 4, wherein the identifying for the actual starting picture comprises:

if the picture selected by the user is a B-picture or a P-picture, the actual start picture is designated as an I-picture immediately preceding the picture selected by the user.

6. The method of claim 4, further comprising the steps of: deleting any B pictures that appear between the user-selected picture and the actual start picture.

7. The method of claim 6, further comprising the steps of: blanking any I pictures and any P pictures that occur between the user-selected picture and the actual starting picture.

8. The method of claim 7, wherein the identifying for the actual starting picture comprises: designating the actual starting picture as the user-selected starting picture if the user-selected starting picture is an I picture.

9. The method of claim 4, wherein the identifying for the actual starting picture comprises: and if the picture selected by the user is in the last two pictures of the GOP containing the selected picture, designating the actual starting picture as an I picture in a subsequent GOP.