1. Field of the Invention
This invention relates to the transmission of entertainment programming, and more specifically to the merging of multiple encoded audio and video streams into a single program in the transport stream to enable playback of any audio stream with any video stream.
2. Description of the Related Art
Television programs are distributed to viewers by a variety of broadcasting methods. These methods include traditional analog broadcast television (National Television Systems Committee or “NTSC” standard), the digital broadcast television “Advanced Television Systems Committee” or “ATSC” standard), cable television (both analog and digital), satellite broadcasting (both analog and digital), as well as other methods. These methods allow audio and video streams for television programming to be encoded multiplexed into a transport stream that is transmitted over a common transmission medium.
As shown in
To view a television program on a TV 28, a subscriber may have to subscribe to a service package offered by a pay-TV service/transmission provider such as a direct broadcast satellite (DBS) operator (e.g., DIRECTV) or a cable company. Such a pay-TV service provider may require a subscriber to utilize an integrated receiver decoder (IRD) 30 that enables the descrambling or decryption of the transmission downloaded from an antenna 32. The IRD may be configured to allow the viewing of one or more particular channels, programs, etc. based on a subscriber's payment or subscription.
As shown in
As shown in
Current generation manufacturers such as Divicom, Thomson, and Motorola now produce MPEG Encoder/Program Mux boxes 60 that operate independently of one another as shown in
The present invention provides a system and method for merging of multiple encoded audio and video streams into a single program in the transport stream to enable playback of any audio stream with any video stream.
This is accomplished by first locking the frequency of audio and video input streams for a plurality of different programs to a source clock, independently encoding each program with its own PCR and PTS into a program stream, multiplexing the program streams into a transport stream, and then synchronizing the PCR and PTS of the program streams to a declared master PCR. The A/V streams can be frequency locked either by using frame synchronization or by providing a common source clock. The PCRs are synchronized by declaring one program and its PCR to be the master and comparing each slave PCR value to an interpolated value of the master PCR. The slave PCR is set to the interpolated value and the offset added to each occurrence of the PTS (or DTS). The offset may be time averaged to improve resolution in high jitter encoders. This approach can be used with currently existing 2nd generation hardware to enable playback of an audio stream from one program and encoder box with a video stream from another program and encoder box. Although this approach does require frequency locking of the input streams, it does not require a single encoder reference clock nor does it require collocation of the encoder boxes.
These and other features and advantages of the invention will be apparent to those skilled in the art from the following detailed description of preferred embodiments, taken together with the accompanying drawings, in which:
Satellite and cable providers continue to expand television programming to include some special events such as NASCAR, NFL, concerts, etc. that include multiple video and audio inputs. In a NASCAR event, the view from the announcer, pit crew, driver and field and audio of the announcer, pit boss, driver and crowd may be available. Typically, the director cuts the various audio and video feeds to create the program that is broadcast. Using currently available technology, the A/V inputs could be made available to a consumer in assigned pairs as different programs. For example, the view from the car and driver commentary would be one program, the view from the pit and pit boss commentary would be a second program, and the director's cut of the available audio and video inputs would be the broadcast program. Each of these A/V inputs would be a separate program allowing the consumer to simply surf from one channel to the next. Neither frequency locking nor clock synchronization is required.
However, in some programming it may be desirable to provide the consumer with the flexibility to mix and match the audio and video inputs, permitting the viewer to be his or her own director. For example, the consumer could select the “view from the car” & “pit boss commentary”. In order to do this, all of the sources must be both “source clock frequency locked” and “encoder (MPEG PCR) clock synced”. Otherwise the consumer IRD cannot reliably decode and display the selected audio and video streams at the proper times. The variation in frame frequency may cause the audio and video streams from different programs to drift apart and could cause overflow or underflow in the buffer. A discrepancy in the PCR would cause the IRD to try to buffer the audio to play at the wrong time.
The currently installed encoder systems do not provide this type of synchronization. Notably, the original CLI hardware could be used to provide this mix and match capability. However, as noted above CLI's common MPEG clock approach was not well accepted by customers and was abandoned by the industry. Furthermore such an approach is not compatible with the existing installed hardware.
As shown in
The current invention builds on the existing hardware platform to frequency lock and clock synchronize the program streams 76 to effectively merge them into a single program in the transport stream 78. This allows a consumer, and the existing base of installed IRDs, to mix and match audio and video streams as if they were encoded using a single N-input encoder/transport Mux.
The audio and video streams can be source clock frequency locked in one of two ways. As shown in
In this example, the first PCR for program X has a value of 1,000,000 ticks and the PTS for the video and audio streams are 2,000,000 and 1,901,000, respectively. The first PCR for program Y has a value of 10,000,000 ticks due to the different epochs in the separate encoders and the PTS for the video and audio streams are 12,000,000 and 11,860,000, respectively.
To synchronize these program streams to the encoder clock, an encoder clock synchronizer 100 synchronizes the programs' PCR and PTS values in the transport stream off of a declared master PCR in one of the program streams to merge them into one program. As illustrated in
The synchronizer suitably processes the transport stream time sequentially a packet at a time. For example, the synchronizer will read a packet header and determine that the master PCR-X is included in the payload. The synchronizer reads and stores the PCR value e.g., 1,000,000 and then counts the number of packets until a next PCR or PTS is detected. In this case, the next time stamps are the PTS X-V (video) and PTS X-A (audio) values, which remain unchanged at 2,000,000 and 1,901,000 respectively. The synchronizer next encounters PCR Y four packets away from PCR X. The interpolated value for program X is 1,000,000+4 packets * 1,000 ticks/packet or 1,004,000 tics as shown in
The offset may be time averaged to improve resolution (step 118). The clock should advance at the exact consistent rate of 27,000,000 ticks per second. This is generally true, however, some manufacturers only keep to the 27 mega ticks per second rate as a long term average and in the short term, the PCR might vary a few thousand ticks variation from sample to sample. This is called PCR jitter. MPEG has a specification that permits PCR jitter to upwards to two seconds, however in practice; IRDs will not tolerate much jitter. By averaging the difference it makes it so the jitter from the main channel does not increase the jitter on the slave channel. Consequently, doing averaging is only important to make things work with ‘brand X’ encoders. This step is not at all necessary for properly designed low jitter encoders. The window size was selected to be approximately one second to match the behavior of ‘brand X’ encoders, however given a worse case MPEG encoder the window might go as high as two seconds. The typical value of the window size is zero, since most encoders are relatively jitter free PCRs. If a discontinuity is detected between the average and the current value of the offset, the time averaged offset is reset to the current offset (step 120). This may occur if, for example, an encoder is taken off-line and replaced with another encoder having a different EPOC.
As illustrated in
In the current example, the customer selects a channel corresponding to Program X video and Program Y audio. The IRD extracts the corresponding PID numbers and filters transport stream to extract only those packets. The IRD extracts a PCR value of 1,000,000 and loads it into its 27 MHz counter. The IRD decrypts and then decodes only the Program X video frame and Program Y audio segment and stores the data in a buffer. The IRD monitors the PCR via its internal counter and when it reaches 2,000,000 outputs the Program X video frame. Similarly when the counter reaches 2,864,000, the IRD outputs Program Y audio segment.
In an alternate embodiment, a customer may select more than one video stream and/or more than one audio stream for concurrent playback. For example, two or more videos may be simultaneously displayed using a picture-in-picture or montage functionality. A customer may select a primary audio stream for normal playback and a secondary audio stream that is intermittently played over the normal audio. Alternately, different audio streams could be directed to different speakers.
As mentioned previously this process can be applied to any programming content to provide a master PCR against which all the PTS are referenced. However, the process is particularly applicable to “live” event programming in which a director receives multiple “frame synced” feeds that he/she ordinarily cuts back & forth to mix and match the audio and video feeds to produce the one broadcast feed that a consumer ordinarily watches. As a result, the enhanced A/V feeds can be processed and delivered to consumers that have the proper subscriptions and hardware.
While several illustrative embodiments of the invention have been shown and described, numerous variations and alternate embodiments will occur to those skilled in the art. Such variations and alternate embodiments are contemplated, and can be made without departing from the spirit and scope of the invention as defined in the appended claims.
Number | Name | Date | Kind |
---|---|---|---|
5533021 | Branstad et al. | Jul 1996 | A |
5652615 | Bryant et al. | Jul 1997 | A |
5703877 | Nuber et al. | Dec 1997 | A |
5818512 | Fuller | Oct 1998 | A |
5874997 | Haigh | Feb 1999 | A |
5936968 | Lyons | Aug 1999 | A |
5982452 | Gregson et al. | Nov 1999 | A |
6002687 | Magee et al. | Dec 1999 | A |
6075556 | Urano et al. | Jun 2000 | A |
6519283 | Cheney et al. | Feb 2003 | B1 |
20020023267 | Hoang | Feb 2002 | A1 |
20020094025 | Hanamura et al. | Jul 2002 | A1 |
20020144265 | Connelly | Oct 2002 | A1 |
20030058268 | Loui et al. | Mar 2003 | A1 |
20030090591 | Concion et al. | May 2003 | A1 |
20030135861 | Sinz et al. | Jul 2003 | A1 |
20030185238 | Strasser et al. | Oct 2003 | A1 |
20040086041 | Ye et al. | May 2004 | A1 |
20040261127 | Freeman et al. | Dec 2004 | A1 |
20060031914 | Dakss et al. | Feb 2006 | A1 |
20060146815 | Tse | Jul 2006 | A1 |
Number | Date | Country |
---|---|---|
2430164 | Jul 2002 | CA |
2338383 | Dec 1999 | GB |
2001-313936 | Nov 2001 | JP |
02100110 | Dec 2002 | WO |
Number | Date | Country | |
---|---|---|---|
20060248559 A1 | Nov 2006 | US |