Automatic memory sizing for a digital video recorder.

0001The invention relates to automatic memory sizing for a digital video recorder.

0002With the rapid advancements in digital circuit technology, digital video recorders are now a reality. A digital video recorder differs from the popular video cassette recorder ("VCR") in that it records video information in a digitised, or binary encoded, format. This difference is critical to applications involving image processing because digital image processing usually provides image processing capabilities far more sophisticated than its analog image processing counterpart.

0003An example of where digital image processing plays a key role is that of a television receiver capable of providing a picture-in-a-picture ("PIP") video display. While viewing a normal video display with such a television receiver, a viewer can selectively introduce a second video display within a smaller, predefined area within the original video display. One use for this is where the television viewer, while watching one channel's programming as the main display, can selectively view another channel's programming via the smaller PIP display. Further background information on such television receivers can be found in Hakamada, U.S. Patents nos. 4,725,888, 4,746,983 and 4,761,688, and Hakamada et al., U.S Patents nos. 4,729,027, 4,774,582 and 4,777,531.

0004One way to provide a PIP video display is to use a digital video recorder. The circuitry constituting the digital video recorder receives two analog video signals, where one constitutes the primary signal intended to be the main display and the other constitutes the secondary signal intended to be the PIP display. The digital video recorder digitises these signals and stores them within its digital memory. This stored, digitised video information is subsequently read out in the appropriate manner to be converted back to analog video information and produce both a main display and the PIP display.

0005Depending upon such factors as the desired resolution of the reproduced video displays and/or the potential time delays to be introduced in displaying the stored video information, more or less video information will need to be stored. For example, an NTSC standard video signal consists of successive "frames" of 525 lines made up of two interlaced "fields" of 262½ lines each. Furthermore, an NTSC standard video signal consists of 30 frames per second. The equivalent figures for a PAL system are 625 lines per frame, 312½ lines per field and 25 frames per second. Thus, as will be appreciated by one of ordinary skill in the art, depending upon the desired resolutions of the video displays and/or the potential time delays to be introduced, a greater or lesser number of frames will need to be stored. This in turn, translates to a potential need for a selectively variable storage capacity within the digital memory circuits within the digital video recorder.

0006One way to provide for this selective variability in storage capacity is to use multiple, interchangeable circuit card assemblies ("CCAs") to hold the digital memory devices used for storing the video information. By installing the appropriate CCAs, which in turn contain the appropriate digital memory devices, into the digital video recorder, selectively variable amounts of video information (e.g., video frames) can be stored and read out as desired.

0007Generally, a video data memory CCA uses a single memory address bus for writing and reading video data into and out from its memory, respectively. To have and use more than one memory address bus on a video data memory CCA would greatly increase the complexity of the wiring backplane and programming (i.e., software) necessary to handle dual memory address buses, particularly at the speeds at which the video data must be transferred.

0008Therefore, when video data is being written into memory locations in memory devices located on one video data memory CCA, other video data being read out simultaneously for a video display (or storage elsewhere) must be read out from corresponding memory locations in the memory devices located on a separate video data memory CCA. Thus, at least two video data memory CCAs are needed to allow incoming video data to be stored while previously stored video data is being read out for simultaneous, albeit time delayed, display.

0009For example, a digital video recorder can use four video data memory CCAs, each having a storage capacity of up to 32 video frames, to store its video data. These four video data memory CCAs are paired together, with each pair used to simultaneously store and read out, as discussed above, video data representing up to double the number of video frames of which each video data memory CCA is capable of storing.

0010Of course, implicit in this is the requirement, or virtual requirement, that all four video data memory CCAs, or at least both within each pair, provide equal data storage capacities. If their storage capacities are not equal, then any additional storage capacity of the larger capacity video data memory CCA beyond that of the smaller capacity video data memory CCA will be unused, and therefore wasted. For example, if one video data memory CCA is capable or storing up to 16 video frames, while the other (or at least its mate) are capable of storing only up to eight video frames, then only eight video frames worth of its storage capacity will be used because of the simultaneous write and read operations, as discussed above.

0011In accordance with a first aspect of the invention, there is provided a digital video recorder comprising a plurality of memory units wherein each individual memory unit has its own respective data storage capacity, and a controller for establishing a hierarchical pairing of the individual memory units based upon their respective relative data storage capacities, wherein the pairings each comprise portions of the paired individual memory units, the pairings being established by the controller allocating and storing in separate allocation memory means the corresponding addresses for the paired portions of the memory units, the controller further simultaneously writing and reading digital data into and out from, respectively, the paired portions of the memory units according to the allocated addresses.

0012Thus the invention provides a digital video recorder which allows for the installation and memory efficient use of multiple video data memory CCAs without regard for their respective data storage capacities.

0013The invention also provides a controller for determining within a plurality of digital memory circuits the availability of data storage capacity for storage of digital data therein and for storing the digital data within the digital memory circuits, wherein the controller comprises control logic for causing: <ul id="ul0001" list-style="dash"><li>electronic sensing of the data storage capacity available within each individual memory circuit within the plurality of memory circuits;</li><li>electronic pairing of the individual memory circuits hierarchically according to their respective relative data storage capacities by electronically storing in a separate digital memory the addresses corresponding to the memory locations within the paired memory circuits, thereby creating an address allocation table containing addresses for addressing the individual memory circuits according to the hierarchy; and</li><li>electronic writing and reading of the digital data into and out from, respectively, the individual memory circuits according to the address allocation table.</li></ul>

0014Thus the invention provides a means for enabling the configuring and addressing of digital memory circuits used, for example, to store binary encoded video information, so as to accommodate the use of multiple memory circuits having dissimilar storage capacities and to maximise the efficiency of such usage. The invention thereby provides a means, which, in combination with a digital video recorder, enables the digital video recorder to recognise and compensate for substitutions of video data memory CCAs having data storage capacities different from the remaining video data memory CCAs, while minimising the amount of unused or wasted data storage capacity otherwise caused by such differences in relative data storage capacities.

0015In accordance with a second aspect of the invention, there is provided a method for determining electronically within a plurality of digital memory circuits the availability of data storage capacity for storage of digital data therein and for storing the digital data within the digital memory circuits, the method comprising the steps of: <ul id="ul0002" list-style="none"><li>(a) electronically sensing the data storage capacity available within each individual memory circuit within the plurality of memory circuits;</li><li>(b) electronically pairing the individual memory circuits hierarchically according to their respective relative data storage capacities by electronically storing in a separate digital memory the addresses corresponding to the memory locations within the paired memory circuits, thereby creating an address allocation table containing addresses for addressing the individual memory circuits according to the hierarchy; and</li><li>(c) electronically writing and reading the digital data into and out from, respectively, the individual memory circuits according to the address allocation table.</li></ul>

0016A preferred embodiment of the invention described below provides an automatic memory sizer for a digital video recorder which programmably examines and configures available memory space within its video data memory circuit card assemblies ("CCAs") for efficient storage and retrieval of digital video data, regardless of whether the memory CCAs have similar data storage capacities. By checking for selectively placed conductive jumpers on the memory CCAs containing the memory devices, or by systematically writing test data into and reading it back out of the memory CCAs, the sizer electronically determines the presence or absence of available memory CCAs and their respective relative data storage capacities.

0017The sizer then establishes a hierarchy of the available memory CCAs by pairing portions thereof, based upon their respective relative data storage capacities. The sizer creates a memory address allocation table based upon such pairings for efficient addressing thereof by storing, in look-up table form, addressing information by simultaneously addressing corresponding memory locations in the paired video data memory CCAs for simultaneously writing and reading incoming and outgoing video data, respectively.

0018The invention will now be further described by way of illustrative and non-limiting example, with reference to the accompanying drawings, in which corresponding elements therein are designated by like references throughout and in which: <ul id="ul0003" list-style="none"><li>Figure 1 illustrates, in simplified, functional block diagram form, a digital video recorder providing automatic memory sizing in accordance with the invention;</li><li>Figure 2 illustrates, in simplified, perspective pictorial form, the video data memory circuit card assemblies for use in a digital video recorder providing automatic memory sizing in accordance with the invention;</li><li>Figures 3A-3D illustrate, in simplified, conceptual form, examples of hierarchial pairing of memory space within the video data memory circuit card assemblies of a digital video recorder providing automatic memory sizing in accordance with the invention;</li><li>Figure 4 illustrates, in flow chart form, the basic operational steps or control logic of a controller providing automatic memory sizing in accordance with the invention to determine the number of and memory capacities of its video data memory circuit card assemblies;</li><li>Figure 5 illustrates, in flow chart form, the basic operational steps or control logic of a controller, providing automatic memory sizing in accordance with the invention to sort its video data memory circuit card assemblies according to their memory capacities;</li><li>Figures 6A-6B illustrate, in flow chart form, the basic operational steps, or control logic of a controller providing automatic memory sizing in accordance with the invention to construct a memory address allocation table for addressing its video data memory circuit card assemblies.</li></ul>

0019Figure 1 shows the basic elements for a digital video recorder 10 providing automatic memory sizing in accordance with the present invention, namely: an input buffer 12; a user interface 14; a memory and system controller 16; memory circuit card assemblies 18; and an output buffer 20.

0020Video information is inputted to the input buffer 12 as a video input signal 22. This video input signal 22 can consist of a composite video signal, analog red, green and blue ("RGB") video signals, or digital RGB video signals. If the video input signal 22 is a composite video signal, the input buffer 12 includes circuits for extracting the vertical and horizontal synchronisation signals and converting the composite video signal to its RGB equivalents. Once the composite video signal has been converted to its RGB equivalents, or if the video input signal 22 was originally analog RGB video signals, these analog RGB video signals are digitised within the input buffer 12, and thereby converted to their digital RGB video signal equivalents.

0021The digitised video signal 24 is outputted from the input buffer 12 onto the input video data bus 26. As explained more fully below, this digitised input video data is stored within the video data memory circuit card assemblies ("CCAs") 18. This stored video data is subsequently read out onto the output video data bus 28 for transfer to the output buffer 20.

0022The output buffer 20 outputs this video data as a video output signal 30. The video output signal 30 can be the digital RGB video signals, analog RGB video signals, or a composite video signal. If the video output signal 30 is desired as an analog signal, the output buffer 20 includes circuits, which are well known in the art, appropriate for making this conversion.

0023The memory and system controller 16 contains several types of digital circuits which are well known in the art, such as a microprocessor, read only memory ("ROM"), random access memory ("RAM"), digital registers and digital counters. In this embodiment of the invention, automatic memory sizing is achieved through the combination of the programming of the microprocessor and its concomitant interaction with the aforementioned circuits within the memory and system controller 16. This programming of the microprocessor within the memory and system controller 16 can be accomplished by inserting the appropriately programmed ROM or by inputting the appropriate program steps via the user input signal 32 and user interface 14.

0024The memory and system controller 16 is programmed to configure the available memory space within the memory CCAs 18. The memory space within the memory CCAs 18 is configured such that two memory CCAs 18, or portions of each thereof, are paired with one another such that input video data 24 is written into one of the input video data bus 26 while, simultaneously, output video data is read out from the other on the output video data bus 28.

0025For example, the first memory CCA 18a and the second memory CCA 18b can be paired together, e.g., as "primary" 18a and "secondary" 18b memory CCAs, respectively. The video data writing and reading operations will alternate between these two memory CCAs 18a, 18b such that as two lines of new video input data 24 are written in the primary memory CCA 18a, the previously stored two lines of video data in the secondary memory CCA 18b are read out, and vice versa. This alternating, simultaneous writing and reading of video data in and from the memory CCAs 18a, 18b is programmably controlled by the memory and system controller 16.

0026The controller 16 outputs an appropriate bit pattern on the write address bus 34 and a correspondingly appropriate bit pattern on the read address bus 36. Decoding circuits located on the primary memory CCA 18a decode the bit pattern on the write address bus 34. Decoder circuits on the secondary memory CCA 18b decode the bit pattern on the read address bus 36. The controller 16 then strobes these memory CCAs 18a, 18b with individual strobe signals 40a, 40b, thereby activating the write and read operations, respectively, on the memory CCAs 18a, 18b.

0027In a preferred embodiment of the present invention, the addresses placed upon the write 34 and read 36 address buses consist of eight bits each, e.g., four "board" bits and four "frame" bits. The four board bits indicate which half of which memory CCA 18 is addressed, and the four frame bits indicate which frame therein is addressed. In other words, each memory CCA 18 is addressed as if it were two assemblies or "boards", having up to 16 frames' worth of data storage capacity in each "board". Thus, the four board address bits indicate which memory CCA 18 and which half thereof is to be written into or read from. Similarly, the four frame address bits indicate which block, or frame's worth, of memory space within the designated board (see description for Figures 3A-3C below) is to be written into or read from.

0028After the primary memory CCA 18a has written two lines of input video data and the secondary memory CCA 18B has read out two lines of input video data, the write and read addresses's bit patterns are interchanged on the write 34 and read 36 address busses, thereby causing the next two lines of incoming video input data to be written into the secondary memory CCA 18b, while the previously stored two lines of video data are read out from the primary memory CCA 18a.

0029The video data written into the primary memory CCA 18a via the input video data bus 26 is written into sequential memory locations within the primary memory CCA 18a. Correspondingly, video data read out from the secondary memory CCA 18b with which the primary 18a is paired, is read out from sequential memory locations which correspond to those in the primary memory CCA 18a in which video data is written. This is accomplished by sequentially addressing the memory locations within the memory CCAs 18a, 18b into which the video data is written into or read out from.

0030This sequential addressing is done by the memory and system controller 16 which places a sequentially incremented bit pattern on the pixel address bus 38. The address bit pattern placed on the pixel address bus 38 begins at a value which corresponds to the first memory location into which video data is to be either written into or read out from. Digital counters located in the memory and system controller 16 simply count up or down, thereby sequentially incrementing or decrementing, the pixel address bit pattern.

0031In the preferred embodiment of the present invention, the pixel address consists of 16 bits. These 16 bits indicate the corresponding memory locations within the "boards" and "frames" indicated by the write and read addresses, as described above.

0032This process whereby two lines of video data are written into a primary memory CCA 18a, while the previously stored two lines of video data are read out from a secondary memory CCA 18b, and vice versa, is repeated until an entire frame of video data has been written and read out. In turn, this is repeated until as many frames of video data have been written and read out as of which the memory CCA 18 with the smaller storage capacity is capable.

0033In other words, if one memory CCA 18a has a video data storage capacity of 32 frames and is paired with a memory CCA 18b having a video data storage capacity of 16 frames, this process is repeated until 16 frames of video data have been written and read out. This must be so, because the smaller memory CCA 18b has no more storage capacity with which the remaining 16 frames of storage capacity in the larger memory CCA 18a can be paired.

0034However, as explained more fully below, the remaining 16 frames of video data storage capacity in the larger memory CCA 18a can be paired with video data storage capacities existing within the other memory CCAs 18c, 18d. If this portion of the larger memory CCA 18a is paired with any of the remaining memory CCAs 18c, 18d, the write and read operations are controlled as described above by the memory and system controller 16 via the write address bus 34, read address bus 36, pixel address bus 38 and strobe signals 40a, 40c, 40d.

0035Configuration of the available video data memory space, i.e., paring of the memory CCAs 18, is done by the memory and system controller 16. When pairing the memory CCAs 18, the controller 16 first determines the respective storage capacities of the memory CCAs 18 via a status bus 42. Each memory CCA 18 makes available to the controller 16, via the status bus 42, a status signal which indicates its video data storage capacity. Once the storage capacities are known (as explained below for Figure 2), the controller 16 pairs the memory CCAs 18, or portions of each thereof, to maximise the usage of all available video data storage capacity (as explained below for Figures 3A-3C).

0036Figure 2 shows the video data memory CCAs 18, which consist of printed circuit boards on which edge connectors 50 can be installed to allow the memory CCAs 18 to be easily replaced. These edge connectors 50 have multiple conductive contacts 52 which provide for electrically coupling to the aforementioned data 26, address 34, 36, 38 and status 42 buses, as well as the strobe signals 40.

0037Installed upon each of the memory CCAs 18 are multiple memory devices, e.g., dynamic random access memories ("DRAMs") 54. These memory devices 54 provide the memory locations into which incoming video data is written and from which outgoing video data is read. Also installed upon the memory CCAs 18 are status registers 56. The status registers 56 are simply programmable digital registers which are programmed with status bits indicating the size, i.e., the data storage capacity, of their respective memory CCAs 18. These status registers 56 drive the status bus 42, thereby informing the memory and system controller 16 as to the data storage capacities of the memory CCAs 18, as described above.

0038These status registers 56 are programmed by installing, or not installing as the case can be, conductive jumpers 58, 60. When a conductive jumper 58, 60 is installed, it electrically couples two conductive pads on the memory CCA 18. By any of several means which are well known in the art, this electrical coupling of conductive pads can be made to represent a logical one or a logic zero which in turn can be used as a status bit with which the corresponding status register 56 is programmed. Similarly, the non-installation of a jumper 58, 60 can also, by several means well known in the art, be used to represent a logical zero or a logical one for programming the corresponding status register 56.

0039For the jumper 58, 60 configurations illustrated on the exemplary memory CCAs 18 in Figure 2, the status bit patterns programmed by the jumpers 58, 60 for the four memory CCAs 18 can be represented by the bit patterns "11", "10", "01" and "00". The first status bit pattern "11" indicates that the first memory CCA 18a contains sufficient memory devices 54 to store up to 32 frames of video data. The second status bit pattern "10" indicates that the second memory CCA 18b contains sufficient memory devices 54 to store up to 16 frames of video data. The third status bit pattern "01" indicates that the third memory CCA 18c contains sufficient memory devices 54 to store up to 8 frames of video data. The last status bit pattern "00" indicates that the last memory CCA 18d contains sufficient memory devices 54 to store up to 4 frames of video data.

0040As will be appreciated by one of ordinary skill in the art, all four memory CCAs 18 can be outfitted with sufficient memory devices 54 to store up to 32, 16, 8 or 4 frames of video data, as desired, with the appropriate Jumpers 58, 60 installed.

0041Means for determining the data storage capacities of the memory CCAs 18 other than electrical jumpers 58, 60 programming status registers 56 can be used. For example, test video data can be inputted to a memory CCA 18 via the input video data bus 26 and read back out via the output video data bus 28 to the memory and system controller 16 (see Figure 1). By writing into and reading out from sequentially higher addressed memory locations within the memory CCA 18, the controller 16 can determine the maximum data storage capacity of the memory CCA 18.

0042Once the memory and system controller 16 has determined the respective data storage capacities of the memory CCAs 18 via their status registers 56 and the status bus 42, the controller 16 programmably constructs a memory hierarchy by hierarchically pairing portions of the memory CCAs 18 (described more fully below). Figures 3A-3C illustrate conceptually how this memory hierarchy is constructed.

0043Figure 3A illustrates the conceptual equivalents of the four memory CCAs 18a, 18b, 18c, 18d, designated as CCAs "A", "B", "C" and "D" with data storage capacities of 32, 16, 8, and 4 video frames, respectively, as illustrated in Figure 2 and discussed above. The first memory CCA 18a is illustrated as a group of 32 blocks, or squares. Each block represents one video frame's worth of data storage capacity, i.e., one "frame" as described above. Similarly, the remaining memory CCAs 18b, 18C, 18d are represented by groups of 16, 8 and 4 blocks, respectively, each block representing one frame.

0044As discussed more fully below the hierarchical pairings of the various portions of the memory CCAs 18 are done as "overlapped" pairings. The interconnecting lines indicate conceptually how the various portions of the memory CCAs 18 are paired in an overlapping manner to support the simultaneous write and read operations. As shown, the overlapped primary-secondary pairings of the memory CCAs 18 are as follows in Table 1: <tables id="tabl0001" num="0001"><img file="EP0459800A2_D0001.tif" /></tables>

0045As stated above, the hierarchial pairings of the various portions of the memory CCAs 18 are done as "overlapped" pairings. This overlapped manner of pairing the memory CCAs 18 is illustrated in Figure 3D for the exemplary memory pairings shown in Figure 3A. The first frame A1 represents the first frame of the first memory CCA 18a, i.e., CCA "A". Similarly, the second B1 and third A2 frames represent the first and second frames of the second 18b and first 18a memory CCAs, i.e., CCAs "B" and "A" respectively. Each of the numbered areas (e.g., 1, 2, 3, . . .) within each of these frames A1, B1, A2 represents one video line's worth of data storage capacity. Thus, the first two areas A1 (1,2) within the first frame A1 are used to store the first two lines of a video frame.

0046For the example of Figures 3A and 3D, the first frame A1 of the first memory CCA 18a and the first frame B1 of the second memory CCA 18b are paired as "primary" and "secondary" frames, respectively. This "primary-secondary" designation means that as an incoming video frame is being stored within these memory CCAs 18a, 18b, alternating pairs of its lines are stored in corresponding memory locations, beginning with the first, i.e., "primary", frame A1. In other words, its first two lines are stored in the first two lines A1(1,2) of the first CCA 18a, its third and fourth lines are stored in the third and fourth lines B1(3,4) of the second CCA 18b, its fifth and sixth lines are stored in the fifth and sixth lines A1(5,6) of the first CCA 18a, and so on, as illustrated in Figure 3D.

0047Similarly, the next video frame is stored in the first B1 and second A2 frames of the second 18b and first 18a memory CCAs, respectively, with the first frame B1 being the primary frame and the second frame A2 being the secondary frame. This overlapped manner of pairing primary and secondary "frames" and "lines" within the memory CCAs 18 is continued in accordance with the pairings listed in the Table 1 above.

0048As will be recognised by one of ordinary skill in the art, the memory locations used to store incoming video lines need not correspond exactly to the video lines themselves. In other words, the first two video lines need not be actually stored in the first two lines' worth of memory, as defined by whatever addressing scheme is used, but instead can be stored in any area of memory. For example, the first actual two lines' worth of memory can be reserved for other uses, such as temporary storage of video test pattern data. Thus, the above references to the "first", "second", "third" and "fourth" lines' worth of memory are used as relative, and not absolute, terms.

0049Figure 3B illustrate the example where memory CCA A has a storage capacity of 32 frames and each of the remaining memory CCAs B, C, D has a data storage capacity of eight frames. As shown, the overlapped primary-secondary pairings of the memory CCAs 18 are as follows in Table 2: <tables id="tabl0002" num="0002"><img file="EP0459800A2_D0002.tif" /></tables>

0050Figure 3C illustrates the example where memory CCA A has a storage capacity of eight frames and each of the remaining memory CCAs B, C, D has a data storage capacity of four frames. As discussed more fully below, since in this case no memory CCA 18 has sufficient memory capacity to be paired with all other memory CCAs 18, "staggered" overlapped pairing is used. As shown, the staggered overlapped primary-secondary pairings of the memory CCAs 18 are as follows in Table 3: <tables id="tabl0003" num="0003"><img file="EP0459800A2_D0003.tif" /></tables>

0051In the preferred embodiment of the present invention in which the write and read addresses each consists of eight bits (e.g., four "board" bits and four "frame" bits), as described above, each "board" consists of up to 16 "frames". Thus, in the example of Figure 3A, the first memory CCA 18a contains two boards having 16 frames each; the second memory CCA 18b contains one board having 16 frames; the third memory CCA 18c contains one board having eight frames; and the fourth memory CCA 18d contains one board having four frames. In the example of Figure 3B, the first memory CCA 18a contains two boards having 16 frames each, and the second 18b, third 18c and fourth 18d memory CCAs each contain one board having eight frames each. In the example of Figure 3C, the first memory CCA 18a contains one board having eight frames, and the second 18b, third 18c and fourth 18d memory CCAs each contain one board having four frames each.

0052Referring to Figure 3A, when establishing the memory hierarchy and one of the memory CCAs 18 contains sufficient memory capacity to be paired with the memory capacities of all other memory CCAs 18 combined, the memory and system controller 16 begins with the largest capacity memory CCA 18a and pairs it with the second largest capacity memory CCA 18b. This results in the memory capacity (B1-B16) of the second largest memory CCA 18b being used in conjunction with a corresponding amount of memory capacity (A1-A16) within the larger memory CCA 18a to support the simultaneous write and read operations.

0053The controller 16 then pairs the next smaller capacity memory CCA 18c with the first memory CCA 18a. Therefore again, the storage capacity (C1-C8) of this smaller capacity memory CCA 18c is used in conjunction with a corresponding capacity (A17-A24) within the larger capacity memory CCA 18a for supporting the simultaneous write and read operations.

0054The fourth, and final, memory CCA 18d is then pired with the first memory CCA 18a. Their respective and corresponding storage capacities (D1-D4, A25-A28) are used together to support the simultaneous write and read operations. Since the total memory capacity of the smaller memory CCAs 18b, 18c, 18d is less than that of the largest memory CCA 18a, some memory capacity (A29-A32) goes unused.

0055For the example illustrated in Figure 3B, the first memory CCA 18a is again the largest in terms of data storage capacity. Therefore, as before, the controller 16 begins with this memory CCA 18a. However, in the example illustrated in Figure 3B, the remaining memory CCAs 18b, 18c, 18d are of equal data storage capacities. Therefore, they can be paired in any arbitrary order with similarly sized portions of the first memory CCA 18a, as illustrated in Figure 3B. As in the case of Figure 3A, a portion (A25-A32) of the first memory CCA 18a goes unused due to a lack of further data storage capacity on the other memory CCAs 18b, 18c, 18d.

0056For the example illustrated in Figure 3C, the pairings of the memory CCAs 18 cannot be accomplished as simply. As will be recognised by referring to Figure 3C, if each of the memory CCAs 18 were merely paired with only one other memory CCA 18, e.g., CCAs A-B and CCAs C-D, although all available memory capacity on CCAs C and D would be used, fully half of the memory capacity of CCA A would go unused. Therefore, "staggered" overlapped pairing of the memory CCAs 18 is done. As shown for the example of Figure 3C, this results in CCA A being paired with three CCAs (B, C, D), CCA B being paired with two CCAs (A, D) and CCA C being paired with only one CCA (A). Such staggered pairing provides for the pairing of all available memory capacity, and is possible because of the fact that the memory capacities of the CCAs 18 (i.e., 32, 16, 8 and 4 frames) are a series of binary multiples.

0057As will be appreciated by one of ordinary skill in the art, numerous permutations of possible data storage capacities exist among the four memory CCAs 18. However, by programming the memory and system controller 16 to begin with the largest capacity memory CCA 18 and pairing it, or portions thereof, in overlapped pairing with successively smaller capacity memory CCAs 18, or alternatively, pairing the memory CCAs 18 in staggered overlapped pairings, efficient pairings of data storage capacities are accomplished, while supporting the simultaneous video data write and read operations and minimising the amount of unused, and therefore wasted, video data storage capacity.

0058This hierarchical pairing of the memory CCAs 18 by the memory and system controller 16 is preferably done immediately following initial power-up of the digital video recorder 10. By programming the controller 16 to perform this hierarchical pairing immediately upon system power-up, performance of the digital video recorder 10 will be optimised immediately. In other words, even if some or all of the memory CCAs 18 have been replaced with other memory CCAs 18 having different or smaller, data storage capacities, by performing the hierarchical pairing of the memory CCAs 18 immediately upon system power-up, the memory and system controller 16 optimises performance of the digital video recorder based upon the data storage capacities of the memory CCAs 18 presently installed.

0059Figures 4, 5 and 6A-6B illustrate, in flow chart form, the basic operational steps or control logic of a controller for performing automatic memory sizing in accordance with the invention. Referring to Figure 4, the first subroutine 1000 determines the number of and data storage capacities of the memory CCAs 18. The first step 1002 is to initialise some program constants. The maximum number BM of boards, i.e., memory CCAs 18, is initialised at four and the maximum number FM of frames is initialised at 128. A loop counter C is initialised at zero and a board count variable BC is initialised at zero.

0060As discussed above, a preferred embodiment of a digital video recorder 10 providing automatic memory sizing in accordance with the present invention has a maximum of four memory CCAs 18. With a maximum frame capacity of 32 frames per memory CCA 18 in the preferred embodiment, the maximum frame storage capacity is 128 frames. Therefore, in such a preferred embodiment, the maximum number BM of boards is four and the maximum number of frames FM is 128. However, as will be recognised by one of ordinary skill in the art, these maximums BM, FM are not mandatory and can be adjusted as desired.

0061The next step 1004 is to determine whether or not the loop counter C is less than the maximum number BM of boards. Initially the answer to this question will be yes and operation will continue with the following step 1006. This step 1006 sets the board address B_CA equal to the value of the loop counter C and sets the board hold flag B_CH (discussed more fully below) in its false state for this board C, i.e., for the memory CCA 18 having board address "C".

0062The next step 1008 is to address and read the size, i.e., the data storage capacity in frames, of this board C. As described above, this is done by looking to see which conductive Jumpers 58, 60, if any, have been installed on the board C. The board size parameter B_CS for this board C is set equal to the measured value of its data storage capacity. The board index B_CI for this board C is also set equal to this value. As described above, the resulting board size parameter B_CS (and index B_CI) will be one of the series of binary multiples 32, 16, 8 or 4.

0063Collectively, the board address parameter B_CA, board hold flag B_CH, board size parameter B_CS and board index B_CI make up a board parameter vector B_C which characterises the board C for the purposes of hierarchically pairing and addressing the memory CCAs 18.

0064The next step 1010 is to increment the loop counter C by one and increment the board count variable BC by one. However, this incrementing of the board count variable BC is done only if the board size parameter B_CS represents a proper board size, in this case that being among the binary multiples of 32, 16, 8 or 4 frames.

0065The foregoing steps 1004, 1006, 1008, 1010 are repeated until the loop counter C is no longer less than the maximum number BM of boards. At that point, the next step 1012 is to determine whether the board count variable BC is less than two. If it is, this is an improper condition and the next step 1014 is to output an error signal of some form to indicate that an insufficient number of memory CCAs 18 has been installed into the digital video recorder 10.

0066However, if the board count variable BC is equal to or greater than two, the next step 1016 is to perform the board sorting subroutine, as shown in Figure 5. Upon completing the board sorting subroutine 1016, the next step 1018 is to perform the allocation table subroutine, as shown in Figures 6A-6B.

0067As explained more fully below, the board sorting subroutine 1016 sorts the memory CCAs 18 according to their respective data storage capacities so that the memory CCA 18 having the largest data storage capacity occupies the highest position within the memory hierarchy as defined by the allocation table, with successively smaller capacity memory CCAs 18 occupying successively less significant hierarchy positions. As explained more fully below, the allocation table subroutine 1018 constructs the memory address allocation table used for hierarchically pairing and addressing the memory CCAs 18.

0068Referring to Figure 5, the board sorting subroutine 1016 beings with the step 1020 of initialising a loop counter C at a value of one less than the maximum number BM of boards. Additionally, a change flag FC is set to its false state (discussed more fully below).

0069The next step 1022 is to determine whether or not the board size B_CS for this board C is greater than the board size B_C-1S for the immediately adjacent board C-1. If the answer is yes, the next step 1024 is to exchange the corresponding board parameter vectors B_C, B_C-1 and set the change flag FC to its true state. The setting of the change flag FC to its true state serves to indicate that board parameter vectors B_C, B_C-1 have been exchanged. If the answer is no, this step 1024 is skipped.

0070The following step 1026 is to decrement the loop counter C by one.

0071The next step 1028 is to determine whether the loop counter C is still greater than zero. If the answer is yes, operation resumes with the step 1022 of comparing the board sizes B_CS, B_C-1S of the board C and its adjacent board C-1. If the answer is no, the next step 1030 is to determine whether the change flag FC has been set to its true state to ascertain whether any board parameter vectors B_C, B_C-1 have been exchanged.

0072If the change flag FC is true, operation resumes with the first step 1020, wherein the loop counter and change flag FC are initialised. If the change flag FC is false, thereby indicating that the memory CCAs 18 have been hierarchically positioned according to their respective board sizes B_CS, operation continues with the following step 1032. The next step 1032 is to compute a memory size offset OS by subtracting the value of the board size B₁S for board one from the board size B₀S for board zero. Additionally, the loop counter C is set at a value of two.

0073The next step 1034 is to determine whether the computed memory size offset OS is greater than zero. If the answer is no, no further steps are performed within this subroutine. If the answer is yes, the computed memory size offset OS is reduced by a value equal to the board size B_CS for this board C.

0074Additionally, the board hold flag B_CH is set to its true state. As discussed above, this causes the memory on this board C not to be allocated, i.e., paired with the memory on another board, until last.

0075The next step 1038 is to increment the loop counter C by one, followed by the step 1040 of determining whether the loop counter C is still less than the maximum number BM of boards. If not, execution of this subroutine ceases and operation returns to the main subroutine (see Figure 4). However, if the answer is yes, operation resumes with the step 1034 of determining whether the memory size offset OS is greater than zero.

0076Once execution of the board sorting subroutine 1016 has been completed, the allocation table subroutine 1018 is executed, as shown in Figures 4 and 6A.

0077Referring to Figure 6A, the first step 1042 is to initialise a first loop counter C1 at zero and a second loop counter C2 at zero, followed by the step 1044 by initialising a third loop counter C3 at zero.

0078The next step 1046 is to determine whether the board index B_C2I (for board C2) is greater than zero. If the answer is no, the next step 1048 is to set a temporary board address variable B equal to "FF" (hexadecimal) and a temporary frame number variable F to "FF" ( hexadecimal ).

0079However, if the board index B_C2I is greater than zero, the next step 1050 is to then determine whether the board hold flag B_C2H is set to its false state. If not, the next step 1048 is to set the temporary board address B and frame number F variables to "FF" (hexadecimal).

0080However, if the board hold flag B_C2H is false, the next step 1052 is to set the temporary board address variable B equal to the board address B_C2A.

0081The following step 1054 is to set a step variable S equal to four.

0082The next step 1056 is to determine whether the board size B_C2S is greater than or equal to 16 frames. If it is, the next step 1058 is to reset the step variable S to a value of 16 frames. If not, this step 1058 is skipped.

0083The next step 1060 is to determine whether the board index B_C2I is greater than the step variable S. If it is, the next step 1062 is to set the fourth bit in the temporary board address variable B by performing a logical "OR" operation between its current value and the value "08" (hexadecimal). If not, this step 1062 is skipped.

0084Referring to Figure 6B, the next step 1064 is to decrement the value of the board index B_C2I by one and set the temporary frame number variable F at a value equal to the remainder of the quotient (i.e., "residue") of the board index B_C2I divided by the step size S.

0085The next step 1066 is to determine whether the board index B_C2I is equal to zero. If it is not, the next step 1068 is to reset the third loop counter C3 equal to the maximum number BM of boards. However, if the board index B_C2I is equal to zero, the next step 1070 is to determine whether the second loop counter C2 is less than the value of the maximum number BM or boards decremented by one. If the answer is no, the next step 1068 is to reset the third loop counter C3. If the answer is yes, the next step 1072 is to increment the value of the second loop counter C2 by one and set the board hold flag B_C2H to its false state.

0086The next step 1068 is to set the third loop counter C3 equal to the maximum number BM of boards, followed by the step 1076 of incrementing the second loop counter C2 by one.

0087The next step 1078 is to determine whether the value of the second loop counter C2 is less than the maximum number BM of boards. If it is not, the next step 1080 is to reset the value of the second loop counter C2 at zero. If it is, this step is skipped.

0088The next step 1082 is to determine whether the value of the third loop counter C3 is less than the maximum number of BM of boards. If it is, operation resumes with the step 1046 (see Figure 6A) of determining whether the board index B_C2I is greater than zero.

0089However, if the value of the third loop counter C3 is not less than the maximum number BM of boards, the next step 1084 is to set the board address BP_C1 equal to the temporary board address variable B and the frame number FP_C1 equal to the temporary frame number variable F.

0090The next step 1086 is to increment the first loop counter C1 by one.

0091The last step 1088 is to determine whether the value of the first loop counter C1 is less than the maximum number FM of frames. If it is not, execution of this subroutine 1018 is complete. However, it is, operation resumes with the step 1044 (see Figure 6A) of setting the value of the third loop counter C3 equal to zero.

0092It should be understood that various alternatives to the embodiments of the present invention described herein can be employed in practising the present invention.

0093For example, although the preferred embodiment of the invention is described with the control logic in the form of software logic, it will be appreciated that it could equally be implemented by special purpose hardware logic.