Patent Grant
7558918
The present invention relates to streaming information. More particularly, the present invention relates to recording streaming information and retrieving the stored information for selective playback.
With advances in technology including faster processors, improved graphics, and so forth, a desktop computer can easily receive and present streaming information to the user. Common examples of streaming information include streaming video and audio delivered over a wide area network, such as the Internet. For instance, television broadcast signals that would otherwise be transmitted wirelessly using satellites, television transmitters, etc., are encoded and made available for transmission to remote computer users via the Internet. Upon request by the desktop computer user, the encoded data packets containing audio and video data are sent to the desktop computer user sequentially. Upon receipt, the data packets are decoded and processed by the desktop computer in order to render the streaming information to the user in as close to real time as possible. After rendering or presentation, the data packets are discarded.
Although processing streaming information in the manner described above is useful, there exist a number of shortcomings. Currently, streaming information is provided at the request of each desktop computer. Thus, each user must form a separate connection with the source of streaming information in order to receive the desired streaming information. Once initiated, the user is unable to control the manner in which streaming information is rendered. For instance, the user cannot temporarily “pause” the incoming streaming information in order to perform another task and then resume viewing when desired. Likewise, the user is unable to repeat a previously rendered portion since the data packets are discarded, or skip ahead since the data packets have not been received.
There thus is an ongoing need to improve the manner in which streaming information is rendered. Although described above with respect to a desktop computer and streaming information received from the Internet, the improved process should be applicable to other information appliances or computing devices and other forms of streaming information.
An information appliance receives streaming information. The information appliance includes a storage buffer and a writer module which receives the streaming information and writes the streaming information to the buffer. A plurality of reader modules are configurable to read the streaming information from the buffer independently of one another.
In accordance with another feature of the present invention, the streaming information includes at least one channel comprised of a plurality of streams of information. The reader modules each include a plurality of output pins, one of the plurality of streams being provided by the reader module on each output pin. An interface is associated with each of the plurality of reader modules. The interfaces are configured to identify the plurality of output pins corresponding to the reader module associated with the interface.
With reference to
Although the exemplary environment described herein employs a hard disk, a removable magnetic disk 29 and a removable optical disk 31, it should be appreciated by those skilled in the art that other types of computer readable medium which can store data that is accessible by a computer, such as magnetic cassettes, flash memory cards, digital video disks, Bernoulli cartridges, random access memories (RAMs), read only memory (ROM), and the like, may also be used in the exemplary operating environment.
A number of program modules may be stored on hard disk, magnetic disk 29, optical disk 31, ROM 24 or RAM 25, including an operating system 35, one or more application programs 36, other program modules 37, and program data 38. A user may enter commands and information into personal computer 20 through input devices such as a keyboard 40 and pointing device (mouse) 42. Other input devices (not shown) may include a microphone, joystick, game pad, satellite dish, scanner, or the like. These and other input devices are often connected to processing unit 21 through a serial port interface 46 that is coupled to system bus 23, but may be connected by other interfaces, such as a sound card, a parallel port, a game port or a universal serial bus (USB). A monitor 47 or other type of display device is also connected to system bus 23 via an interface, such as a video adapter 48. In addition to monitor 47, personal computers may typically include other peripheral output devices such as a speaker 49 connected to a sound card 57 and printers (not shown).
Personal computer 20 may operate in a networked environment using logic connections to one or more remote computers, such as a remote computer 49. Remote computer 49 may be another personal computer, a server, a router, a network PC, a peer device or other network node, and typically includes many or all of the elements described above relative to personal computer 20, although only a memory storage device 50 has been illustrated in
When used in a LAN networking environment, personal computer 20 is connected to local area network 51 through a network interface or adapter 53. When used in a WAN networking environment, personal computer 20 typically includes a modem 54 or other means for establishing communications over wide area network 52, such as the Internet. Modem 54, which may be internal or external, is connected to system bus 23 via serial port interface 46. In a network environment, program modules depicted relative to personal computer 20, or portions thereof, may be stored in the remote memory storage devices. It will be appreciated that the network connections shown are exemplary and other means of establishing a communications link between the computers may be used.
Streaming information can be received by computer 20 using a number of known methods and technologies. Commonly, a source of streaming information is a remote computer wherein computer 21 is connected to the remote computer using a wired or wireless modem. This technique is used often when streaming information is provided through an intranet or the Internet, typically in digital form. Digital streaming information can further comprise satellite signals received by a satellite receiver, dish or the like.
However, streaming information can also arrive as analog signals. For instance, the streaming information can also comprise broadcast radio or television signals. In such cases, computer 20 includes a radio tuner 60 and a television tuner 62 to receive the broadcast signals and convert the analog signals to digital form for transmission over system bus 23.
It should be understood that the present invention can be used in other computing devices besides computer 20, discussed above.
Memory 72 is implemented as non-volatile electronic memory such as random access memory (RAM) with a battery back-up module (not shown) such that information stored in memory 72 is not lost when the general power to mobile device 68 is shut down. A portion of memory 72 is preferably allocated as addressable memory for program execution, while another portion of memory 72 is preferably used for storage, such as to simulate storage on a disc drive.
Memory 72 includes an operating system 80, an application program 82 as well as an object store 84. During operation, operating system 80 is preferably executed by processor 70 from memory 72. Operating system 80, in one preferred embodiment, is a “WINDOWS CE” brand operating system commercially available from Microsoft Corporation. Operating system 80 is preferably designed for mobile devices, and implements database features which can be utilized by application 82 through a set of exposed application programming interfaces and methods. The objects in object store 84 are maintained by application 82 and operating system 80, at least partially in response to calls to the exposed application programming interfaces and methods.
Communication interface 76 represents numerous devices and technologies that allow mobile device 68 to receive streaming information. The devices are similar to those discussed above with respect to computer 20 and include wired and wireless modems, satellite receivers and broadcast tuners to name a few. Mobile device 68 can also be directly connected to computer 20 to exchange data therewith. In such cases, communication interface 76 can be an infrared transceiver or a serial or parallel communication connection, all of which are capable of transmitting streaming information.
Streaming information is provided to entertainment appliance 90 through a communications interface 98. Communications interface 98 can be any of the devices and technologies described above with respect to the two previous environments.
At this point, it should be noted that system 110 can be operated in any of the computing environments described above, or similar computing environments. Those skilled in the art will appreciate that delay filter 112, rendering device 116, encoder 118 and decoder 120 can be implemented in hardware, software, or combinations thereof. In one embodiment by way of example, delay filter 112 is embodied in the operating system. Higher level application programs or other portions of the operating systems can access functions of delay filter 112 using application program interfaces (APIs) as is well known in the art.
In operation, streaming information source 114 provides an information stream to delay filter 112 (optionally through encoder 118). Generally, the streaming information comprises digital data representing one or more channels of content information. For instance, streaming information source 114 can comprise an Intranet or the Internet available through the communication interfaces described above. Likewise, streaming information source 114 can comprise an analog or digital television tuner wherein separate audio, video and data (e.g. closed captioning) information streams comprise a single channel. Other sources of streaming information include, but are not limited to, audio tuners, satellite receivers and the like.
In the embodiment illustrated, encoder 118 receives the streaming information and encodes or compresses the streaming information into a known format such as “MPEG”, “AVI”, “MOV” (Apple®QuickTime®) and “WAV”, although, if used, the present invention is not limited to any one particular encoding format.
Generally, as discussed below, delay filter 112 includes a writer module 122, a circular buffer 124 and one or more reader modules 126. Writer module 122 receives the streaming information provided by streaming information source 114 and writes the streaming information into circular buffer 124. Circular buffer 124 can comprise any of the storage devices described above, for example, hard disk 27 or RAM memory. Reader module 126 accesses circular buffer 124 to retrieve the streaming information when the streaming information is to be rendered. If the streaming information stored in circular buffer 124 is encoded or compressed, decoder 120 decodes or uncompresses the streaming information, which is then provided to rendering device 116.
Streaming information and synchronization point information is provided to writer module 122. In one embodiment, as illustrated, writer module 122 comprises a mux writer that receives multi-stream streaming information for storage in circular buffer 124. As indicated above, one or more reader modules 126 (herein labeled as 1261, 1262 and 1263) are provided to read the streaming information from circular buffer 124 for rendering. Writer module 122 stores synchronization information in an index 132. Reader modules 126 may access index 132 in order to locate a particular portion of the streaming information and properly render the streaming information. Operation of writer module 122, circular buffer 124, reader modules 126 and index 132 are discussed in detail below.
In the embodiment illustrated, two separate playback reader modules 1261, and 1262 are illustrated having outputs that provide streaming information to separate video, audio and data decoders 120 and rendering devices 116. In general, this illustrates that separate reader modules 126 can be reading streaming information from circular buffer 124 at different points in circular buffer 124, and thus, represents separate individuals accessing the data stored therein. In addition, other reader modules 126, such as indicated at 136, can be implemented to archive and store the streaming information in circular buffer 124 for later viewing. Generally, in archiving system 136, reader module 1263 provides streaming information to a mux formatter 138 that, in turn, provides the information to a writer module 140 for storage in any of the storage devices indicated above such as hard disk 27.
Referring back to circular buffer 124, circular buffer 124 has “floating” beginning and ending points, which are referred to as a logical “head” 150 and “tail” 152. Head 150 corresponds to the logical head of valid data in circular buffer 124, and tail 152 corresponds to the logical tail of valid data in circular buffer 124. Writer module 122 always writes to the head of buffer 124, which moves circularly through the buffer in the direction of arrow 154. Buffer 124 therefore always has a fixed maximum time quantum of data available for reading. For example, when time shifting multimedia (e.g., audio and video) content, writer module 122 receives the streaming multimedia information and stores the information in circular buffer 124. The user views the stored multimedia content through one of the readers modules 126. The circular structure of buffer 124 allows some portion of the streaming information to be available for the user to “instant replay” or “pause” on demand, for example, without allowing the buffer to “fill-up” with “time-shifted” data. Circular buffer 124 can be implemented in volatile or non-volatile memory, such as random access memory (RAM), a hard disk, a floppy disk or an optical disk. In one embodiment, circular buffer 124 is implemented in hard disk drive 27.
Since writer module 122 and reader modules 1261-126N can operate independently of one another and at different data rates, buffer IO layer synchronizes writer module 122 and reader modules 1261-126N to maintain a predetermined temporal order between writing and reading. In one embodiment, buffer IO layer 200 prevents any reader 1261-126N from reading data that is not yet logically available and prevents writer module 122 from overwriting data that is in the process of being read by one or more of the reader modules 1261-126N In circular buffer 124, a given physical position corresponds to multiple logical positions. Without synchronization, a reader module that is trailing writer module 122 by a distance that is close to the buffer size may be reading from the same physical area to which writer module 122 is writing. Buffer IO layer 200 also allows reader modules 1261-126N to follow writer module 122 as close as possible to minimize latency.
Buffer IO layer 200 implements a synchronization algorithm for writer module 122 and reader modules 1261-126N. Each time writer module 122 wants to pass data to buffer IO layer 200, its corresponding application calls the synchronization algorithm. Similarly, each time one of the reader modules 1261126N wants to read data from buffer IO layer 200, its corresponding application calls the synchronization algorithm. The synchronization algorithm can be implemented in hardware, software, or a combination of both, as desired.
The synchronization algorithm uses “blocking” to (1) block a reader module 1261-126N that is trying to read data which has not yet been written and/or (2) block writer module 122 if it is trying to write to an area of circular buffer 124 from which one of the reader modules 1261-126N is currently reading. In both cases, one component is blocked until another component has completed the operation necessary to remove the offending condition. For example, if writer module 122 is blocked, it remains blocked until all of the reader modules 1261-126N, which who are reading from the area to be written, have completed their reads. When a reader module is blocked, it remains blocked until writer module 122 has written all of the data requested by that reader module.
The synchronization algorithm uses a plurality of shared variables. Each client has its own set of variables which are shared with the other clients. In
When writer module 122 issues a write command, it specifies an area in buffer 124 to be written, starting at a present write position 234. A “Writer Blocked On” integer variable 236 is used to identify the logical position corresponding to the end of the data to be written when all or part of the area to be written is blocked by one or more of the reader modules 1261-126N. The Writer Blocked On variable is set when one of the reader modules 1261-126N is currently reading from that area. For example, a reader module may be presently reading from logical position 235, which is in the area between positions 234 and 236. A “zero” value for the “Writer Blocked On” variable indicates that writer module 122 is not currently blocked by any of the reader modules 1261-126N.
A “Currently Reading From” integer variable is maintained for each reader module 1261-126N. The Currently Reading From variable is used to indicate that the reader module is currently performing a read that starts at this logical position in circular buffer 124. For example, in
A “Writer Unblocked Event” variable is used to “wake-up” writer module 122 when it can proceed with its desired write command. For example, the Writer Unblock Event variable is set to an active state when writer module 122 is not blocked, and is reset to an inactive state when writer module 122 is blocked. The Writer Unblock Event can be implemented as a Windows® event (a “Win32” event) or any other similar synchronization mechanism familiar to those skilled in the art.
A “Reader Unblock Event” variable is used for each reader module 1261-126N for “waking-up” the reader module when the data it is requesting is available. For example, the Reader Unblock Event variable is set to an active state when the corresponding reader module 1261-126N is not blocked and is reset to an inactive state when the corresponding reader module 1261-126N is blocked.
A “Critical Section” variable is used to protect access to each of the above shared variables. For example, a “Win32” Critical Section may be used or any other similar synchronization mechanism such as a mutual exclusion “mutex” object, as is known to those skilled in the art.
Referring back to
At step 304, algorithm 290 sets the “Writer Blocked On” variable to the value of the “Tail Pointer” variable 250. This indicates the largest logical position in circular buffer 124 at which the reader modules 1261-126N must be clear of before writer module 122 can write the data. Algorithm 290 then blocks writer module 122 by resetting the “Writer Unblock” event to the inactive state, at step 305. Since writer module 122 is blocked, the Critical Sections are unlocked at step 306. At step 307, algorithm 290 waits for the “Writer Unblock” event variable to be activated by the synchronization algorithm for the conflicting reader that is accessing the area to be written. When the conflicting reader module activates the “Writer Unblock” event, algorithm 290 locks the Critical Section at step 308 and then resets the “Writer Blocked On” variable to zero, at step 309. Algorithm 290 again unlocks the Critical Section at step 310 and proceeds to step 311 (shown in
At step 311, buffer IO layer 200 converts the starting logical address to a circular or physical address and writes the data in circular buffer 124, beginning at that physical address. Once the data has been written, algorithm 290 locks the Critical Section for the variables used in steps 313-315 and advances the “Head Pointer” variable to the logical position in circular buffer 124 that corresponds to the end of the data written by writer module 122. In the example shown in
At step 314, algorithm 290 determines whether any of the “Reader Blocked On” variables for reader modules 1261-126N is less than the advanced “Head Pointer” variable 274. If not, none of the reader modules 1261-126N were waiting for the data written by writer module 122, and algorithm 290 proceeds to step 316. If so, one or more of the reader modules 1261-126N were waiting for writer module 122, which has now validated the desired logical positions. Algorithm 290 sets any such “Reader Unblock” event variables at step 315 to unblock the corresponding reader module.
In the example shown in
Referring back to
At step 360, algorithm 350 determines whether the logical position corresponding to the beginning of the requested data is before the logical position of the “Tail Pointer” variable. If so, the requested data is invalid since it has already been overwritten by writer module 122. In this case, algorithm 350 unlocks the Critical Section at step 361 and fails the requested read operation at step 362. The failure can be passed to the corresponding reader module through a variety of mechanisms, such as a status variable.
If the requested data starts at a logical position that is not before the “Tail Pointer” variable, algorithm 350 proceeds to step 363 (shown in
At step 368, algorithm 350 determines whether the “Writer Blocked On” variable points to a logical position in circular buffer 124 that is greater than the read position. The read position is the logical position in circular buffer 124 corresponding to the beginning of the data to be read. If the “Writer Blocked On” variable is not greater than this logical position, then this reader module was not blocking writing module 122. Algorithm 350 then unlocks the Critical Section at step 369 and completes its function at step 370.
If the “Writer Blocked On” variable is greater than this position, then this reader module was blocking writing module 122, and algorithm 350 proceeds to step 371 to determine whether any other reader module is blocking writer module 122. If not, algorithm 350 sets the “Writer Unblock” event variable to unblock writer module 122, at step 372. If there are other reader modules that are blocking writer module 122, then algorithm 350 proceeds to step 369 without setting the “Writer Unblock” event variable.
The algorithm shown in
In addition, algorithm 290 can be modified to perform steps 360-362 (validating the read operation) either before or after the read is performed at step 365, or both. Validating the read operation before the actual read operation avoids a wasteful read. However, steps 360-362 can be performed after step 366 if desired. Steps 360-362 can also be performed both before and after read step 365.
The synchronization algorithm shown in
The synchronization algorithm shown in
In one embodiment, delay filter 112 (shown in
Certain user operations might be permitted only in certain states of the delay filter. Use of a prohibited operation might lead to inconsistent states in the delay filter and user interface. For example, assume that a reader module that is performing “time-shifting” of multimedia content is paused at point X in the circular buffer 124. In the meantime, writer module 122 is still writing data into circular buffer 124. Depending on the size of the circular buffer 124 and the time for which the reader module has been paused, there is a possibility that writer module 122 can catch up to the point X where the reader module is paused and try to overwrite the data. In this case, delay filter 112 can forcibly “unpause” the reader module and send a notification to the application that the paused reader module has been unpaused. There is a small time window between the time at which the reader is “unpaused” and the time at which the application receives the notification. During this time window, the application might issue a prohibited user command, such as a forward scan. The reader module will now be performing the forward scan at the time the application receives the notification that the reader module has been unpaused and is playing at normal speed. The application and the delay filter 112 are therefore in inconsistent states.
To avoid this problem, delay filter 112 allows the application to check the user operation bits whenever a user operation is issued to check for validity of the operation. Also, whenever the user operation bits change, delay filter 112 reports this change to the application. The application can therefore update its user interface to ensure that invalid user operations are not issued to the delay filter 112.
Another broad inventive aspect illustrated by the exemplary embodiment includes the concept of “time shifting” when streaming information is rendered. Generally, an information appliance for receiving streaming information implements time shifting by including a buffer (in a one embodiment circular buffer 124), writer module 122 which receives blocks of streaming information and writes blocks to the buffer, and at least one reader module 126 which selectively reads the blocks from the buffer.
Although all signal processing equipment inherently includes signal propagation delay, “time shifting,” as used herein, is distinguishable for the reasons discussed below separately or in any combination. In the first instance, the amount of time shifting (i.e., the relative position of any one reader module 126 with respect to the writer module 122 in circular buffer 124) is selective and adjustable. In a first mode of operation, the user can “rewind”, “pause”, “fast forward” and “play” in any desired order, thereby changing the relative reading position of an associated reader module 126 with respect to a position of writer module 122 in circular buffer 124. Hence, rendered streaming information (which begins with a reader module 126) is delayed in time from the streaming information that is written by writer module 122.
In another mode of operation, logic is provided to adjust the amount of “time shifting” based upon operation of the system and without user intervention. For instance, if a reader module has been “paused” by a user, thus maintaining the current position of reader module 126 in circular buffer 124, at some time it may be necessary to begin advancing the reader position in order that writer module 122 can store new streaming information in circular buffer 124. This situation may require reader module 126 to begin moving forward within circular buffer 124 at a rate greater than or at least equal to the rate at which streaming information is being written to circular buffer 124.
Automatic adjustment of the position of a reader module 126 in circular buffer 124 can also be initiated by user commands not directly associated with movement of a position of reader module 122, such as “play”, “pause”, “rewind” and “fast forward”. As probably best illustrated by example, the source of streaming information 114 to delay filter 112 can be a multiple broadcast channel device such as a television tuner, or the like. If the user begins rendering information pursuant to a selected broadcast channel, pauses, and then again resumes rendering information on the same broadcast channel, the associated position of reader module 126 will change position accordingly, thereby increasing the relative distance of the position of reader module 126 and the position of writer module 122 in circular buffer 124. If the user then selects a different broadcast channel to be rendered, for example, a different television channel, writer module 122 will then record the streaming information associated with the new broadcast channel in circular buffer 124. However, since the position of reader module 126 is “time shifted” from the position of writer module 122 by a delay in proportion to the amount of time the user had paused, the user will not notice the new broadcast channel selection until reader module 126 reaches the new streaming information in circular buffer 124 provided by the change in broadcast channel. In order to obviate the problems presented by the foregoing, the position of reader module 126 is automatically advanced to a position in circular buffer 124 adjacent the position of writer module 122. This technique reduces the amount of delay so that rendered streaming information substantially follows broadcast channel selection in a multiple broadcast channel environment.
Another distinguishing feature between “time shifting” as used herein and typical signal propagation delay is the medium used to record streaming information. In particular, as discussed above, circular buffer 124 can be embodied using any of the storage devices described above such as RAM memory, hard disk 27, or the like. In addition, the amount of memory present in circular buffer 124 is sufficient to provide a perceptible delay to the user, if desired. In one embodiment, the extent of circular buffer 124 provides at least five minutes of delay between writing and rendering of streaming information. In a further embodiment, the extent of circular buffer 124 is sufficient to provide at least 30 minutes of delay between writing and rendering streaming information. In yet a further embodiment, the extent of circular buffer 124 is sufficient to provide at least one hour of delay.
A further distinguishing feature of “time shifting” as used herein includes the presence of a single writer module 122 and a plurality of independent reader modules 1261-126N.
In one embodiment, any information received from streaming source 114 is first recorded in circular buffer 124 prior to rendering the streaming information to the user. In other words, all information to be rendered to the user is read with an associated reader module 126 from circular buffer 124, wherein there exists no direct connection from streaming source 114 to rendering devices 116. Although recording and reading from circular buffer 124 prior to rendering streaming information, may slightly increase propagation delay when the position of reader module 126 in circular buffer 124 is substantially adjacent to the position of writer module 122, advantages of always writing to and reading from circular buffer 124 include simplification in system design and operation, which enhances system stability. Although a switching mechanism could be provided in the system to switch from substantially “live” streaming information and time shifted streaming information as provided by delay filter 112, the switching mechanism whether software, hardware or a combination thereof, nevertheless necessitates another component in this system that must be designed and tested to operate with the remaining components of the system. Furthermore, by eliminating a direct connection of the source of streaming information 114 and the rendering devices 116 and always reading from circular buffer 124 prior to rendering, the quality of rendered information is consistent. In other words, there is no change in quality (e.g., video quality or audio quality) of rendered streaming information since all information must pass through the same system components. Furthermore, by always writing to and reading from circular buffer 124, the user need not remember to execute a particular command so that streaming information is recorded, for example, in order to provide “instant replay.” In this manner, the user is also able to retrieve and save an entire segment of streaming information (such as a television show) even when the user has already rendered some of the streaming information. Thus, if the user is viewing a show and then later decides to save the show, he can invoke archive system 136 to begin at the start of the show in circular buffer 124 and transfer a copy of the streaming information to a more permanent file. The latter can be performed while still viewing the show with a separate reader module 126. However, the user must initiate saving or archiving prior to writer module 122 writing over any of the desired streaming information, but in a preferred embodiment, circular buffer 124 is of sufficient length to provide many minutes if not hours of stored information.
In one embodiment, a field 512 is also provided in header portion 502. Field 512 provides other information relevant to the streaming information. As used herein, field 512 is referred to as “prefix field” in view of its relation to the actual data or payload 502. In a one embodiment, prefix field 512 is of variable length, thus the information need not be length restricted. If desired, a field 513 is also provided, indicating the size of the prefix field 512. Typically, field 513 is also of fixed length, like fields 508-511.
Prefix field 512 can be used to store information such as synchronization indicators, discussed below, a pointer to the next data block pertaining to the same stream, or other information as may be desired. As appreciate by those skilled in the art, a format is specified for information to be contained in prefix field 512 in order that relevant information can be decoded easily. However, since prefix field 512 can be of any length, additional information can be stored as necessary, thereby making data block 500 extensible.
It should also be noted that prefix field 512 can also include a selected amount of “fill” data. The fill data is used to adjust or increase the size of data block 500 to a predetermined length, which may be desirable for processing or storage in circular buffer 124.
Data or payload field 504 stores digital data received from the source of streaming information 114, which will be rendered.
In the embodiment illustrated, tail or end portion 506 includes two fields 520 and 521. Field 520, herein also referred to as “suffix field” is similar to prefix field 512 in that it can be used to store information about data blocks that are successive or preceding in the same stream. In one embodiment, suffix field 520 is at a known location from the end of data block 500 and stores a pointer indicating the location of the preceding data block corresponding to the same channel. This information is particularly useful when it may be necessary to perform a “rewind” operation upon the streaming information contained in circular buffer 124. Suffix field 520 can also include other information present in the data block 500 such as data associated with field 511 and field 513. In a manner similar to prefix field 512, a format for the information contained in suffix field 520 is predetermined and known to the reader module in order that relevant information can be obtained therefrom. Suffix field 520 can be of fixed length, or can be of variable length, wherein it may be desired to provide an additional field similar to field 513 that indicates the length of suffix field 520. Suffix field 520 can also be used to store “fill data”, if desired.
Field 521 provides the size of data block 500 and, as such, is identical to field 508 in header portion 502. Field 521 is useful when a reader module progresses backward through the circular buffer 124. In this manner, the reader module can ascertain the size of a data block 500 by reading the value contained in field 521, at which point, the reader module has determined where the end of the preceding block is located. Without the presence of field 521, the reader module would otherwise have to scan the data block 500 in order to determine its size or beginning location, which can consume processing time.
As indicated above, prefix fields 512 and suffix fields 520 can be used to store pointers indicating preceding and succeeding data blocks in each of the stream streams. In the illustration of
It should be noted that field 509 of each data block can be used to identify padding data blocks from other streaming information data blocks wherein writer module 122 can generate padding data blocks as necessary during processing and formation of streaming information data blocks. Furthermore, padding data blocks can be used in combination with “fill” data provided in prefix field 512 or suffix field 520, if desired. Mux formatter 138 and writer module 140 can also implement the use of padding data blocks, and/or fill data contained in prefix field 512 or suffix field 520.
It may be desirable, in some instances, to allow a plurality of users to access the streaming information at any one time. This can be accomplished in any number of different ways. For example, the streaming information could be duplicated and one reader module can be provided for each copy of the streaming information. However, this requires a great deal of storage, and may also require increased processing power for making multiple copies of the streaming information and for accessing the duplicate copies. Also, once the predetermined number of copies of the streaming information are in use, it can become very difficult to add new users.
In the illustrative embodiment shown in
Again, it should be noted that
In one illustrative embodiment, reader modules 614-620 are implemented as objects which conform to the COM object-oriented programming model. Each reader module or “reader object” has independent access to the buffered streaming information. Each reader module 614-620 is depicted as accessing the streaming information at a different location on circular buffer 124. This illustrates that reader modules 614-620 can access the buffered information at different times in the buffered information stream. This allows the multiple users 630-636 to independently employ all the features of time shifting mentioned above, without interfering with the other users accessing the buffered information stream. One user, for instance, can archive the streaming information for later viewing, while at the same time another user can view the information.
Similarly, the archival user can store the buffered streaming information from a time or location in the data buffer which is different than the time or location in the data buffer which the viewer is accessing. For instance, the archival user may simply be reading the information, as it is recorded in circular buffer 124, and storing it in archival storage. By contrast, the viewer may be viewing the data, but intermittently pausing to take breaks or perform other tasks. Therefore, the reader module associated with the viewer may well be at a temporally displaced location in buffer 124 than the reader module associated with the archival process.
Thus,
Since the streaming information written in circular buffer 124 can be formed of one or more channels each having a plurality of streams (e.g., audio, video, closed captioning, etc.), the output from each of the reader modules 614-620 will likely have a plurality of effective output pins, each pin carrying one of the streams of a given channel in the streaming information. Since system 610 illustrates that a plurality of different reader modules can be used to access the same buffered streaming information, this can present some obstacles which must be overcome.
For example, in some streaming architectures, the need may arise to group the pins associated with each reader module 614-620 so that the user, application, or whatever other component is receiving the data from the reader module, knows which pins are associated with that reader module. An example of a streaming architecture which can be used in accordance with the present invention is an architecture known as DirectShow services.
DirectShow services is an architecture which is commercially available and well known. However, for a better understanding of certain aspects of the present invention, a brief description of the DirectShow services system may be beneficial.
DirectShow services refers to a modular system of pluggable components known as filters, arranged in a configuration known as a filter graph. A component referred to as a filter graph manager oversees the connection of these filters and controls the flow of the streaming information therethrough. An exemplary filter graph is composed of a collection of filters of different types.
Most filters can be categorized into one of three types:
1. Source filters take data from a source, such as a disk file, camcorders, satellite feed, internet server, or VCR, and introduce that data into the filter graph;
2. Transform filters process data and pass it along to other portions of the filter graph, or out of the filter graph; and
3. Rendering filters render data to a hardware device or to any location that accepts media input (such as memory or a disk file).
In addition to these three types of filters, there are other kinds of filters also. Examples of other filters include effect filters, which add effects without changing the data type, and parser filters, which are configured to understand the format of the source data and know how to read correct bytes, create time stamps, and perform seek operations.
Further, it is quite possible for some filters to represent a combination of filter types, or functions. In the DirectShow architecture, a filter is said to pass streaming information “downstream” to a next subsequent filter. An “upstream filter” refers to the filter which passes data to the downstream filter, and a “downstream filter” refers to the next filter in line to receive the data. In one illustrative embodiment, the filters are program modules written in any language which can generate objects adhering to component object model (COM) programming. Of course, COM programming refers to the object-oriented programming model which defines how objects interact with a single application or between applications. In COM, client software accesses an object through a pointer to an interface (e.g. API) which has a related set of functions, called methods, on the object.
By way of example only, a filter graph 640, the purpose of which is to play back MPEG-compressed video information from a file may take the form set out in
It can thus be seen that, when using certain streaming architectures, such as the DirectShow architecture, it can be important that applications or other programming modules which are to receive streaming information from any of reader modules 614-620 know which output pins correspond to which of the reader modules. For example, since the streaming information being read by each of the reader modules 614-620 can include a channel having a plurality of streams, it is important that a user or application program receiving information from any given reader receive all streams associated with that reader, and no other reader. In other words, it would be undesirable for an application program (or a rendering filter, for example) to receive an audio stream from reader module 614 but a video stream and closed caption stream from reader module 616. Reader module 616 may well be reading the stream information from a different location in buffer 124 than reader module 614. Therefore, the audio and closed captioning streams would not correspond to the video stream.
In order to address this obstacle, one feature of the present invention includes a new COM interface which allows an object outside of reader modules 614-620 to enumerate the output pins which belong to that particular reader module. In other words, in the embodiment in which reader modules 614-620 are implemented as objects, those objects are configured to expose methods through associated interfaces 622-628 to the application (or other external components which have contact with the reader modules) which allow manipulation of the object. One exemplary method exposed by interfaces 622-628 is referred to as IenumChannelStreamPin, which, when invoked by an outside object, yields a value which represents an enumeration of the particular output pins belonging to the particular reader module which was queried. In this way, an outside object can quickly and easily obtain an enumeration of the particular pins belonging to any of the given reader modules 614-620, through its associated interface 622-628.
Interfaces 622-628 also expose another method which allows any specific pin output by any of reader modules 614-620 to be queried for its group's enumerator object. In other words, an outside object can query any pin with which it has contact to obtain the identity of the particular enumerator corresponding to the group which includes that pin. In this way, a user or application program which has only recently gained access to an output pin, can query that output pin to find the enumerator associated with that output pin. The external component can then invoke the method exposed by the enumerator to obtain a complete enumeration of the pins corresponding to that channel. The external component can thus quickly and accurately obtain the identity of the pins associated with any given reader module, or channel, being output by system 610.
As discussed above, indexer 132 generates an index which contains index entries that are used for seeking. In other words, each of reader modules 614-620 can seek to any point in buffer 124 which has a corresponding entry in the index. In some instances, depending on the nature of the streaming information, substantially any point in buffer 124 can be indexed. However, for other types of streaming information, it may be desirable to index only certain points within the streaming information. Such points are referred to herein as sync points. In other words, the nature of the streaming information may render it unreasonable or undesirable to seek to certain points within the streaming information and begin rendering the data at that point.
For example, some video streaming information works on the well known MPEG2 video format. Such a format includes a number of different types of frames which are referred to as I frames, B frames, and P frames. I frames need no other information in order to be rendered, while B and P frames are dependent frames which require information in a preceding I frame in order to be rendered. In such an example, it may be undesirable to seek to a point within buffer 124 which begins with a B or P frame. In fact, many conventional video decoders may not even be able to decode data which is provided from a B or P frame, without the necessary information contained in the preceding I frame. Therefore, it may be highly desirable to index only I frames in such a video stream stored on buffer 124.
Similarly, where the streaming information contains a stream, such as written text for closed captioning, it may be desirable to allow a user to seek to points which correspond to the beginning or ending of words, the beginning or ending of sentences, or the beginning or ending of paragraphs. The same is true for an audio stream. In other words, it may be desirable to allow a user to only seek to certain spots in the audio stream. Such spots may correspond to sentence or word boundaries, etc.
Therefore, one illustrative feature of the present invention includes a stream analyzer 654, such as that illustrated in
In one illustrative embodiment, stream analyzer 654 embodies knowledge of the organization of the streaming information provided by source 656. Analyzer 654 also illustratively embodies knowledge regarding the detection of logical boundaries in the information in order to obtain the location of those boundaries (sync points). The location of the sync points can then be presented to any other software or hardware component which may desire the information in order to enable skipping forward or backward through the streaming information more practicable.
Stream analyzer 654 illustratively provides an indication of the sync points in one of two ways. In a first embodiment, stream analyzer 654 embeds the location of the sync points within the streaming information itself, as it is provided to information sink 658. This is referred to as in-band communication of derived sync point information.
In an alternative embodiment, stream analyzer 654 can provide the sync point information as indicated by dashed arrow 660, separately from the streaming information provided to sink 658. This is referred to as out-of-band communication of the derived sync point information.
In one illustrative embodiment, stream analyzer 654 is implemented in the C++ programming language using classes and communicating with other software objects using COM interfaces. However, it should be noted that such an implementation is illustrative only and the present invention is not limited to this particular method of implementation.
The information sink 658 can, of course, be a downstream filter (downstream of stream analyzer 654) an application program, a rendering filter or other program component, etc. Sink 658 can use the derived sync point information itself, or simply pass it on to other components which may wish to use the derived sync point information. In one illustrative embodiment, the derived sync point information is provided to delay filter 112 which uses it in generating an index, as is described in greater detail below.
In another illustrative embodiment, not only does stream analyzer 654 analyze the incoming streaming information for logical boundaries which can be used as sync points, but stream analyzer 654 also analyzes the incoming streaming information for points of interest which may be desirable sync points. Such points of interest can be substantially any points which are deemed to possibly be of interest to the user. Therefore, in such an embodiment, stream analyzer 654 embodies knowledge of the type and format of information in the incoming stream, as well as knowledge of the types of events of interest to the user. Similarly, stream analyzer 654 is configured to contain knowledge of how to detect these points of interest in the incoming streaming information, and is provided with a mechanism to report the location in the data stream which corresponds to these points of interest.
In
Parsing component 670 receives the streaming information at its input. In the embodiment illustrated, the streaming information may include a plurality of streams (although only a single stream may be processed as well), such as a written textual stream (e.g., closed caption information), a video stream and an audio stream. Parsing component 670 parses the incoming streaming information into its respective stream components and provides those components to appropriate analysis blocks 672, 674 and 676.
Analysis components 672, 674 and 676 analyze the incoming streams, identify potential points of interest and generate information indicative of the location of the those points of interest. The streaming information, as well as the sync point information, is then provided from each analysis component 672-676, to integration component 678. In the embodiment illustrated, integration component 678 re-integrates the streams into the original streaming information, and also embeds the sync point information at appropriate locations within the streaming information. Thus, the in-band sync point information is provided in the data stream at the output of integration component 678.
It should be noted that the analysis components 672-676 can look for substantially any desired points of interest. For example, written language component 672 can be configured to look for sentence boundaries or word boundaries. In looking for sentence boundaries, the analysis component can simply be configured to look for periods in the incoming stream. To look for word boundaries, component 672 can be configured to simply look for spaces in the incoming stream. It should also be noted that language analysis component 672 can be a more sophisticated analysis component, such as a natural language processing or natural language analysis component, in which case component 672 can be configured to identify certain types of clauses, surnames, parts of speech, etc. In any case, information indicative of the written language points of interest to be identified by component 672 is provided to component 672 such that the desired points of interest can be identified in the incoming stream.
Similarly, video analysis component 674 can be used to identify any number of different points of interest. For example, component 674 can be configured to identify commercial breaks, as well as to identify the I, B and P frames mentioned above. In identifying commercial breaks in the incoming video stream, video analysis component 674 can simply be configured to look for a black screen which exists for a certain predetermined amount of time. Similarly, in the MPEG2 video format, the I, B and P frames include markers identifying them as such, or are preceded by a header identifying the frames. Therefore, video analysis component 674 can be configured to look for the identifying information in the stream.
Audio analysis component 676 can also be configured to look for a wide range of points of interest. By way of example, if the incoming stream is from a horror genre film, a war genre film, or an action genre film or television program, audio analysis component can be configured to look for loud noises, such as screams, gun shots, or car chase noises. In that instance, component 676 can simply be configured to look for amplitude information which exceeds a predetermined threshold. Similarly, component 676 can be a more advanced component, such as a speech recognition or voice recognition component. In the instance in which component 676 includes a speech recognition component, it can be configured to look for sentence or word boundaries, or it can be used to look for particular words or phrases. For example, component 676 can be used to identify words of profanity which can be marked for later deletion. Similarly, where component 676 includes a voice recognition component, it can be configured to identify the voice of certain actors or actresses, or any other entity which can be so identified.
It should again be mentioned that the point of interest analysis component 662 can be configured to identify the location of substantially any point of interest which can be identified. Component 662 simply needs to be provided with the information indicative of points of interest to be analyzed, as well as information indicative of the type of information and format of information which will be received in the streams provided thereto. The specific embodiments mentioned above are simply provided as examples, and are not to be viewed as limiting the application of this feature of the present invention.
Once the point of interest information is derived from the streams, the streams are then reintegrated and the point of interest information is provided either in-band or out-out-band, along with the integrated streams. This is indicated by block 688. The data stream and the point of interest information is then transmitted to a downstream component (such as delay filter 112) for further processing. Again, it should be noted that the point of interest information can either be provided in-band or out-out-band. This is indicated by block 690 in
As described above, reader modules 614-620 in system 610 are configured such that they can seek to any indexed point, in buffer 124. For information streams which do not have sync points, it is reasonable for the reader module to seek to substantially any sample within buffer 124. Therefore, in such information streams, any sample can be indexed. In such an embodiment, indexer 132 generates index entries which simply correspond to the desired granularity of the indexer. For instance, it may be desirable in certain streams to index points which are temporally spaced by approximately 0.25 seconds. With other information, it may be desirable to index points which are spaced by two seconds or more. In information streams which have no sync points, indexer 132 simply generates index entries identifying locations in buffer 124 which are separated by the desired granularity of the indexer.
However, as discussed in the previous section, many data streams will be provided with sync points. In those cases, it may be desirable for indexer 132 to generate index entries which correspond only to sync points. It will, of course, be noted that if sync points occur more frequently than the desired granularity of the index generator, every single sync point may not be indexed. However, if sync points do not occur more frequently than the desired granularity of the index generator, substantially every sync point may be indexed.
Under such an arrangement, a problem can arise. For example, in a given streaming architecture, stream analyzer 654 described above may not exist, or it may not yet know whether the streaming information it is receiving contains any sync points. That being the case, indexer 132 may not know, at the time it begins receiving streaming information, whether it must simply begin indexing samples according to its granularity, or whether it is to wait to index only sync points.
First, indexer 132 simply executes a query against stream analyzer 654 to determine whether stream analyzer 654 can provide information as to whether the incoming streaming information contains sync points. This is indicated by block 702. Stream analyzer 654 may not even provide a response, or it may provide a response indicating that it does not yet know whether the streaming information contains sync points. If analyzer 654 provides a response indicating that information indicative of whether sync points exist is available, a flag is set. The flag is referred to in
Next, if information as to the presence or absence of sync points is available, it is determined whether any sync points are present. This is indicated by block 706. If stream analyzer 654 has already identified sync points in the incoming streaming information, it provides indexer 132 with a response indicating that sync points do exist. In that case, indexer 132 sets a sync point flag to a value which indicates that the incoming streaming data does, in fact, contain sync points. This is indicated by block 707. However, if stream analyzer 654 does not respond, or it has not yet located sync points in the incoming streaming information, and provides an output indicative of that to indexer 132, indexer 132 assumes, for the moment, that there are no sync points in the incoming streaming information, and resets the sync point flag. This is indicated by block 708. Indexer 132 then receives a sample of the streaming information as indicated by block 710.
Upon receipt of the sample, indexer 132 analyzes the sample to see if the sample has been marked as a sync point. This is indicated by blocks 712 and 714. If the sample has not been marked as a sync point, indexer 132 examines the sync point flag to see whether the sync point flag is set. This is indicated by block 716. If, at block 716, it is determined that the sync point flag is indeed set, that indicates that the streaming information being processed does contain sync points, and the present sample is not marked as a sync point. Therefore, indexer 132 does not index the sample under analysis but simply returns to processing at block 710.
If, however, at block 716, it is determined that the sync point flag is not set, indexer 132 is still assuming that no sync points exist in the incoming streaming information. Therefore, the indexer simply determines whether it should index the present sample under analysis, based on the desired granularity of indexer 132. In other words, if indexer 132 is to index points no more often than every one half second (for example), indexer 132 determines whether the present sample is temporally removed from the previously indexed sample by at least one half second. If so, the present sample is indexed. If not, processing simply returns to block 710. This is indicated by block 718.
If, at block 714, indexer 132 determines that the present sample is marked as a sync point, indexer 132 then determines whether the sync point flag is currently set. This is indicated at block 720. If the sync point flag is currently set, indexer 132 simply continues processing at block 718 and determines whether it has reached sufficient temporal displacement from the previous index entry to index the present sync point.
However, if, at block 720 it is determined that, even though the present sample is marked as a sync point, the sync point flag is not set, then indexer 132 determines whether the “Sync Point Info Is Authoritative” flag is set, at block 721. If not, then indexer 132 realizes that it has been assuming that no points exist in the streaming information under analysis, but it also realizes that sync points do actually exist. Therefore, indexer 132 discards all the previous index entries (since they corresponded to non-sync point samples) as indicated at block 722 and sets the sync point flag as indicated by block 724. Processing then again continues with respect to block 718. Similarly, if at block 721 it is determined that the “Sync Point Info Is Authoritative” flag is set, processing continues at block 718.
Therefore, it can be seen that by using the algorithm illustrated in
As described above, the streaming information stored in buffer 124 can include multiple streams. As is also describe above, it may be desirable to allow reader modules 614-620 (illustrated in
For example, the streams can be out of sync within the file located in buffer 124. In other words, audio information can be stored before or after video information, and in different size storage blocks, such that the two do not directly coincide. Also, the reader module may request to seek to a position in buffer 12 (to a sample with a particular time stamp) for which no sync point has been indexed. In other words, as set out above, indexer 132 does not index every single sample. Instead, it only indexes based on its own granularity, or based on sync points. Therefore, a reader module may request to seek to a time stamp for which no index entry is available. Further, some streams may contain sync points while others may not. In addition, not all sync points may be indexed, since they may occur more frequently than the desired granularity of the index. One illustrative feature of the present invention is directed to addressing these obstacles in seeking through a channel having multiple streams.
In order to address the obstacles identified above, one illustrative embodiment of the present invention is a seek algorithm which is described with respect to
More specifically,
Next, a variable FILEOFFSET is set to infinity (or a very large number). This is indicated by block 742.
The first of the plurality of streams is then chosen for analysis. This is indicated by block 744.
For the first stream, a time variable T1 is set to the index entry with the highest time stamp before the desired seek time T. In
Next, a time period T2 is set to an index entry for the selected stream with the lowest time stamp after the desired seek time T. It can be seen that sample 734D corresponds to a sample in the selected stream which is indexed, and which has the lowest time stamp to the right of time T (e.g., after time T). If no such index entry exists, the presently selected stream is ignored for purposes of computing the file offset, and a new stream is selected. This is indicated by block 748.
Thus, after processing at block 748, it can be seen that the two index entries which lie on either side of time T, and which are closest to time T, have been identified. Those index entries correspond to times T1 and T2. The temporal displacement between the desired time T and the previously indexed time T1 is represented by X=T−T1. The temporal displacement between the desired seek time T and the subsequent index entry T2 is represented by Y=T2−T.
If both X and Y are greater than a predetermined time interval (e.g., five seconds) that indicates that both of the indexed samples T1 and T2 are too far from the desired seek position T to be helpful. This determination is indicated by block 750. If it is the case that X and Y are greater than the predetermined time interval, then the present stream is simply ignored in computing FILEOFFSET. This is indicated by block 752. Processing then continues at block 754 where it is determined whether additional streams exist. If so, a next stream is selected at block 756 and processing continues with respect to block 746. If no additional streams exist, a seek is performed at block 758, which is described in greater detail below.
If, at block 750, it is determined that both values X and Y are not greater than the predetermined time interval, then it is determined whether either X or Y are greater than the predetermined time interval. For example, at block 760, it is determined whether value X (which corresponds to T−T1) is greater than five seconds. If so, then the value Y must not be greater than five seconds, and a variable FILE_OFFSET_THIS_STREAM is set to index value T2. This is indicated by block 762.
If, at block 760, it is determined that the value of X is not greater than five seconds, then it is determined whether the value Y is greater than five seconds. If so, then the value T2 can be ignored and the variable FILE_OFFSET_THIS_STREAM is set to value T1. This is indicated by blocks 764 and 766.
Basically, blocks 760-766 determine whether either index values T1 or T2 are too far separated from the desired seek time T to be of help. If they are, they are eliminated from further computing.
If, however, at block 764, it is determined that the value Y is not greater than five seconds, that means that both samples corresponding to values T1 and T2 are close enough to desired seek time T to be of interest. That being the case, it is determined whether other sync points between T1 and T2 may exist. In other words, if a distance between T1 and T2 is so small that no other sync points could have been indexed between them (based on the desired granularity of the indexer) then additional sync points may exist between samples T1 and T2 which are closer to the desired seek position T, but which have simply been omitted from the index because the granularity of the indexer was not sufficient to accommodate an additional index entry.
Therefore, a search algorithm is conducted in which the samples of the selected stream which reside between samples T and T1 are searched to determine whether they are in fact sync points. If so, that would mean that those sync points are closer in time to the desired seek position T, and are prior in time to that seek position. If that is true, then the variable FILE_OFFSET_THIS_STREAM is set to the value corresponding to that identified sync point. This is indicated by blocks 768 and 770.
Therefore, at this point in the processing, the desired file offset for the stream under analysis has either been chosen to be T1, T2, or a value between T1 and T. Then, for the stream under analysis, the seek algorithm requests subsequent filters in the streaming process to ignore any data samples which are read out of buffer 12 and have time stamps prior to that identified by the variable FILE_OFFSET_THIS_STREAM. This is indicated by block 772.
Once the variable FILE_OFFSET_THIS_STREAM has been derived, it is determined whether that value is less than the current value of the variable FILEOFFSET. If this is the first selected stream, the value will be less than the current value of FILEOFFSET, which has been set to infinity. However, if this was a second, third, or later selected stream, then the file offset for the present stream under analysis will be used as the value for the variable FILEOFFSET only if it is further to the left (or earlier in time) than the current value for the variable FILEOFFSET. This has the effect of moving the seek point which will actually be used in seeking in buffer 124 to a point far enough left in the buffer to satisfy every stream's individual seeking requirements. This is indicated by block 774.
Next, processing continues at block 754 where it is determined whether additional streams need to be analyzed. If so, processing continues at block 756. If not, processing continues at block 758. In block 758, the selected reader module is commanded to seek to the sample identified by the variable FILEOFFSET and start streaming data from buffer 124 at that point. Since in block 772, a value has been set for each stream indicating that streamed data prior to the desired sync point for that stream is to be ignored, the data streamed out of buffer 124 will not be rendered for a given stream unless that data is coincident with, or after, the earliest sync point identified for that given stream (and identified by the variable FILE_OFFSET_THIS_STREAM).
One embodiment of pseudocode corresponding to the flow diagram of
Thus, it can be seen that the algorithm illustrated in
Although the present invention has been described with reference to preferred embodiments, workers skilled in the art will recognize that changes may be made in form and detail without departing from the spirit and scope of the invention.
The present application is a continuation of and claims priority of U.S. patent application Ser. No. 10/827,619, filed on Apr. 19, 2004, which is a continuation of U.S. patent application Ser. No. 09/286,808, filed Apr. 6, 1999, now U.S. Pat. No. 6,748,481, issued Jun. 8, 2004; the contents of both aforementioned applications are hereby incorporated by reference in their entirety.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4901286 | Theus | Feb 1990 | A |
| 5136655 | Bronson | Aug 1992 | A |
| 5339418 | East et al. | Aug 1994 | A |
| 5349611 | Varian | Sep 1994 | A |
| 5351236 | Pawelski | Sep 1994 | A |
| 5371551 | Logan et al. | Dec 1994 | A |
| 5371842 | Easton et al. | Dec 1994 | A |
| 5414455 | Hooper et al. | May 1995 | A |
| 5463630 | Tooher | Oct 1995 | A |
| 5566301 | Koz et al. | Oct 1996 | A |
| 5568443 | Dixon et al. | Oct 1996 | A |
| 5596581 | Saeijs et al. | Jan 1997 | A |
| 5596582 | Sato | Jan 1997 | A |
| 5596645 | Fujimori | Jan 1997 | A |
| 5598276 | Cookson et al. | Jan 1997 | A |
| 5664226 | Czako et al. | Sep 1997 | A |
| 5692213 | Goldberg et al. | Nov 1997 | A |
| 5729471 | Jain et al. | Mar 1998 | A |
| 5754773 | Ozden et al. | May 1998 | A |
| 5758011 | Fujinami | May 1998 | A |
| 5761417 | Henley et al. | Jun 1998 | A |
| 5774643 | Lubbers et al. | Jun 1998 | A |
| 5784527 | Ort | Jul 1998 | A |
| 5802394 | Baird et al. | Sep 1998 | A |
| 5805821 | Saxena et al. | Sep 1998 | A |
| 5835667 | Wactlar et al. | Nov 1998 | A |
| 5838876 | Iwamura | Nov 1998 | A |
| 5850572 | Dierke | Dec 1998 | A |
| 5878280 | Lucht | Mar 1999 | A |
| 5915094 | Kouloheris et al. | Jun 1999 | A |
| 5930493 | Ottesen et al. | Jul 1999 | A |
| 5995091 | Near et al. | Nov 1999 | A |
| 6038636 | Brown, III et al. | Mar 2000 | A |
| 6055619 | North et al. | Apr 2000 | A |
| 6058106 | Cudak et al. | May 2000 | A |
| 6097380 | Crites et al. | Aug 2000 | A |
| 6134596 | Bolosky et al. | Oct 2000 | A |
| 6172927 | Taylor | Jan 2001 | B1 |
| 6195680 | Goldszmidt et al. | Feb 2001 | B1 |
| 6195701 | Kaiserswerth et al. | Feb 2001 | B1 |
| 6233389 | Barton et al. | May 2001 | B1 |
| 6233629 | Castellano | May 2001 | B1 |
| 6239793 | Barnert et al. | May 2001 | B1 |
| 6247069 | Smyers | Jun 2001 | B1 |
| 6249319 | Post | Jun 2001 | B1 |
| 6260156 | Garvin et al. | Jul 2001 | B1 |
| 6275899 | Savell et al. | Aug 2001 | B1 |
| 6304924 | Varma | Oct 2001 | B1 |
| 6314466 | Agarwal et al. | Nov 2001 | B1 |
| 6324235 | Savell et al. | Nov 2001 | B1 |
| 6407740 | Chan | Jun 2002 | B1 |
| 6493347 | Sindhu | Dec 2002 | B2 |
| 6813731 | Zahavi et al. | Nov 2004 | B2 |
| 6912637 | Herbst | Jun 2005 | B1 |
| 20030200548 | Baran et al. | Oct 2003 | A1 |
| 20050132101 | McEvilly et al. | Jun 2005 | A1 |
| 20060133456 | Ettorre et al. | Jun 2006 | A1 |
| Number | Date | Country |
|---|---|---|
| 0 633 694 | Jan 1995 | EP |
| 0 702 370 | Mar 1996 | EP |
| 0 713 341 | May 1996 | EP |
| 0 727 911 | Aug 1996 | EP |
| 0 756 281 | Jan 1997 | EP |
| 0 789 488 | Aug 1997 | EP |
| 0 812 108 | Dec 1997 | EP |
| WO 97 48049 | Dec 1997 | WO |
| WO 0060820 | Oct 2000 | WO |
| 03-007558 | Jan 2003 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20050246457 A1 | Nov 2005 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10827619 | Apr 2004 | US |
| Child | 11098018 | US | |
| Parent | 09286808 | Apr 1999 | US |
| Child | 10827619 | US |
