FIELD OF THE INVENTION

The present invention relates, in general, to data storage and archiving and, more specifically, to file formats for storing multiple tracks or streams of data.

BACKGROUND OF THE INVENTION

Thanks to its fidelity, digital video and audio have become increasingly popular in entertainment and information recording. For example, digital versatile disc or digital video disc (DVD) provides a format used to store movies, music, or software programs. A DVD movie often has multiple audio tracks for multilingual presentation of the movie and/or multiple video tracks for including special features such as interviews with the movie producer, movie trailers, etc. A DVD has a memory capacity of approximately six gigabytes (GB). In the standard format, a single sided DVD generally can store approximately two to three hours of video.

It would be advantageous to have a file format for storing digital data with a high compression rate. It would be desirable for the file format to be capable of storing data in multiple streams or tracks. It would also be desirable for the file format to be able to encode and archive video, audio, and text data on easily accessible streams or tracks. It would be of further advantage for the file format to be able to provide copyright protection for the digitized content.

DETAILED DESCRIPTION OF VARIOUS EMBODIMENTS

In accordance with preferred embodiments of the present invention, digitized data, e.g., digital video, audio, and/or text data, are encoded and stored in a multimedia file following a format that is compatible with a standard data coding and compression algorithm. The file is readable and/or executable by a processor, e.g., a specific or generic signal processor, a digital signal processor (DSP), a signal processor on an Application Specific Integrated Circuit (ASIC), an Advanced RISC Machine (ARM) microprocessor, etc.

In accordance with a specific embodiment of the present invention, the file format is based on the audio video interleave (AVI) multimedia format. The AVI file format is a Resource Interchange File Format (RIFF) file specification used with applications that capture, edit, and play back audio-video sequences.

RIFF, introduced in 1991 by Microsoft Corporation and IBM Corporation, is a format for storing tagged data structures. The structure and coding of RIFF can be found at the Microsoft Developer Network (http://msdn.microsoft.com/). The related information on the Microsoft Developer Network is incorporated herein by reference.

A RIFF file includes a RIFF header followed by zero or more lists and chunks. The RIFF header has the following form:

- â€˜RIFFâ€™ fileSize fileType (data)
  where â€˜RIFFâ€™ is a four character code (FOURCC) that has the value â€˜RIFFâ€™, fileSize is a 4 byte number giving the size of the data in the file, and fileType is a FOURCC that identifies the specific file type. The value of fileSize includes the size of the fileType FOURCC and the size of the data that follows, but does not include the size of the â€˜RIFFâ€™ FOURCC or the size of fileSize. The file data includes data chunks and lists, in any order.

A chunk has the form:

- ckID ckSize ckData
  where ckID is a FOURCC that identifies the data contained in the chunk, ckData is a 4 byte number giving the size of the data in ckData, and ckData is zero or more bytes of data. The data is always padded to nearest WORD boundary. ckSize gives the size of the valid data in the chunk, but it does not include the padding, the size of ckID, or the size of ckSize.

A list is an ordered collection of other chunks, for example a collection of movie frames. In RIFF, a list has the form:

- â€˜LISTâ€™ listSize listType listData
  where â€˜LISTâ€™ is the literal FOURCC code â€˜LISTâ€™, listSize is a 4 byte number giving the size of the list, listType is a FOURCC code identifying the type of the list, and listData consists of chunks or lists, in any order. The value of listSize includes the size of listType plus the size of listData, but it does not include the â€˜LISTâ€™ FOURCC or the size of listSize.

It is customary and efficient to imply the chunk size and adopt a simplified notation to represent a RIFF chunk:

- ckID (ckData)
  Adopting a similar simplified notation, a list can be represented as:
- â€˜LISTâ€™ (listType (listData))
  The notation places optional elements in brackets, [optional element]

An AVI file is identified by a FOURCC â€˜AVIâ€™ in the RIFF header. All AVI files include two mandatory LIST chunks: the stream format and the stream data. An AVI file may also include an index chunk, which gives the location of the data chunks within the file. An AVI file with these components has the form:

RIFF(â€˜AVIâ€™[â€˜CSETâ€™ (Character Set)][LIST (â€˜INFOâ€™)]LIST (â€˜hdrlâ€™ ...)LIST (â€˜moviâ€™ ...)[â€˜idx1â€™ (<AVI Index>)])

The â€˜hdr1â€™ list defines the format of the data and is the first required LIST chunk. The â€˜moviâ€™ list contains the data for the AVI sequence and is the second required LIST chunk. The â€˜idx1â€™ list contains the index, which is optional. AVI files keep these three components in the proper sequence.

The â€˜hdr1â€™ and â€˜moviâ€™ lists use subchunks for their data. The following example shows the AVI RIFF form expanded with the chunks needed to complete these lists:

RIFF(â€˜AVI â€™[â€˜CSETâ€™ (Character Set)][LIST (â€˜INFOâ€™)]LIST â€˜hdrlâ€™â€˜avihâ€™(<Main AVI Header>)LIST (â€˜strlâ€™â€˜strhâ€™(<Stream header>)â€˜strfâ€™ (<Stream format>)[â€˜strdâ€™(<Additional header data>)][â€˜strnâ€™(<Stream name>)]...)...)LIST (â€˜moviâ€™{SubChunk |â€‚LIST (â€˜rec â€™ SubChunk1SubChunk2...)...} ...)[â€˜idx1â€™ (<AVI Index>)])

The character set (CSET) chunk is typically used to define a character set and language information for a RIFF file, a LIST, or a stream. The CSET chunk is defined as follows:

CSET(WORD wCodePageWORD wCountlyCodeWORD wLanguageCodeWORD wDialect)

In accordance with a preferred embodiment of the present invention, the functions and formats of the fields in the CSET chunk are defined as follows:

wCodePage specifies the code page used for file elements.

If the CSET chunk is not present or if this field has a value of zero, a standard ISO 8859/1 code page (identical to code page 1004 without code points defined in hex columns 0, 1, 8, and 9) is assumed in accordance with an embodiment of the present invention.

wCountryCode specifies the country code used for file elements. If the CSET chunk is not present or if this field has value zero, USA (country code 001) is assumed in accordance with an embodiment of the present invention. By way of example, the country codes used in the wCountryCode field of CSET chunk are listed in Table 1.

wLanguage and wDialect specify the language and dialect used for file elements. If the CSET chunk is not present or if these fields have value zero, US English (language code 9, dialect code 1) is assumed in accordance with an embodiment of the present invention. By way of example, the language and dialect codes used in the wLanguage and wDialect fields of CSET chunk are listed in Table 2.

The information â€˜INFOâ€™ list is a registered global form type that can store information, e.g., copyright information and comments, that helps identify the contents of the chunk. This information, although useful, does not affect the way a program interprets the file. An â€˜INFOâ€™ list is a â€˜LISTâ€™ chunk with list type â€˜INFOâ€™.

In accordance with a preferred embodiment, an â€˜INFOâ€™ list may contains the chunks listed in Table 3. Additional chunks may be defined. Preferably, an application ignores any chunk it doesn't understand. Each chunk contains a null-terminated Unicode text string. The character set used in the string is specified by the global CSET chunk.

The AVI file header (â€˜hdr1â€™) list includes a main AVI header in an â€˜avihâ€™ chunk. One or more stream descriptor lists follow the main AVI header. Each stream descriptor is contained in an â€˜str1â€™ list.

The main AVI header contains global information for the entire AVI file, such as the number of streams within the file and the width and height of the AVI sequence. The main header chunk includes an AVIMAINHEADER structure, whose syntax is defined as:

typedef struct_avimainheader {FOURCCfcc;DWORDcb;DWORDdwMicroSecPerFrame;DWORDdwMaxBytesPerSec;DWORDdwPaddingGranularity;DWORDdwFlags;DWORDdwTotalFrames;DWORDdwInitialFrames;DWORDdwStreams;DWORDdwSuggestedBufferSize;DWORDdwWidth;DWORDdwHeight;DWORDdwReserved[4];} AVIMAINHEADER

In accordance with a preferred embodiment, the members in the AVIMAINHEADER structure have the following variables:

- fcc specifies a FOURCC code with the value being â€˜avihâ€™.
- cb specifies the size of the structure, not including the initial 8 bytes of fcc and cb.
- dwMicroSecPerFrame
  - specifies the number of microseconds between frames and indicates the overall timing for the file.
- dwMaxBytesPerSec
  - specifies the approximate maximum data rate of the file. This value indicates the number of bytes per second the system must handle to present an AVI sequence as specified by other parameters contained in the main header and stream header chunks.
- dwPaddingGranularity
  - specifies the alignment for data, in bytes. Data are padded to multiples of this value.
- dwFlags
  - includes a bitwise combination of zero or more of the following flags:
  - AVIF_HASINDEXâ€”Indicates the AVI file has an index.
  - AVIF_MUSTUSEINDEXâ€”Indicates that the application should use the index, instead of the physical ordering of the chunks in the file, to determine the order of presentation of the data. For example, this flag could be used to create a list of frames for editing.
  - AVIF_ISINTERLEAVEDâ€”Indicates the AVI file being interleaved.
  - AVIF_WASCAPTUREFILEâ€”Indicates the AVI file as a specially allocated file used for capturing real-time video. Applications should warn the user before writing over a file with this flag set because the user may defragment this file.
  - AVIF_COPYRIGHTEDâ€”Indicates the AVI file contains copyrighted data and software. When this flag is used, software should not permit the data to be duplicated.
- dwTotalFrames
  - specifies the total number of frames of data in the file.
  - dwInitialFrames:
- specifies the initial frame for interleaved files. For non-interleaved files the value should specify zero. When creating interleaved files, the number of frames in the file should be specified prior to the initial frame of the AVI sequence in this member.
- dwStreams
  - specifies the number of streams in the file. For example, a file with audio and video has at least two streams.
- dwSuggestedBufferSize
  - specifies the suggested buffer size for reading the file. Preferably, this size should be large enough to contain the largest chunk in the file. If set to zero, or too small, the playback software will have to reallocate memory during playback, which will reduce performance. For an interleaved file, the buffer size should be large enough to read an entire record, and not just a chunk.
- dwwidth
  - specifies the width of the AVI file in pixels.
- dwHeight:
  - specifies the height of the AVI file in pixels.
- dwReserved
  - reserved, set to zero.

By way of example, the flags in the member dwFlags include the following bitwise combinations:

/* flags for use in <dwFlags> in AVIFileHdr */#define AVIF_HASINDEX0x00000010#define AVIF_MUSTUSEINDEX0x00000020#define AVIF_ISINTERLEAVED0x00000100#define AVIF_TRUSTCKTYPE0x00000800#define AVIF_WASCAPTUREFILE0x00010000#define AVIF_COPYRIGHTED0x00020000

TABLE 1Country codes (wCountryCode)Country Code(wCountryCode)Country or Region000None (ignore this field)001USA002Canada003Latin America030Greece031Netherlands032Belgium033France034Spain039Italy041Switzerland043Austria044United Kingdom045Denmark046Sweden047Norway049Germany052Mexico055Brazil061Australia064New Zealand081Japan082Korea086People's Republic of China088Taiwan090Turkey351Portugal352Luxembourg354Iceland358Finland

TABLE 2Language codes and dialect codesLanguage CodeDialect Code(wLanguage)(wDialect)Language00None (ignore these fields)11Arabic21Bulgarian31Catalan41Traditional Chinese42Simplified Chinese51Czech61Danish71German72Swiss German81Greek91US English92UK English101Spanish102Spanish Mexican111Finnish121French122Belgian French123Canadian French124Swiss French131Hebrew141Hungarian151Icelandic161Italian162Swiss Italian171Japanese181Korean191Dutch192Belgian Dutch201Norwegian - Bokmal202Norwegian - Nynorsk211Polish221Brazilian Portuguese222Portuguese231Rhaeto-Romanic241Romanian251Russian261Serbo-Croatian (Latin)262Serbo-Croatian (Cyrillic)271Slovak281Albanian291Swedish301Thai311Turkish321Urdu331Bahasa

TABLE 3Information List (INFO) chunksChunk IDDescriptionIARLArchival Location, indicating where the subject of the fileis archived.IARTArtist, listing the artist of the original subject of the file.ICMSCommissioned, listing the name of the person ororganization that commissioned the subject of the file.ICMTComments, providing general comments about the file orthe subject of the file, if multiple sentences in length, eachsentence ending with a period, no new line characters.ICOPCopyright, recording the copyright information for thefile, multiple copyrights separated by a semicolonfollowed by a space.ICRDCreation date, specifying the date the subject of the file wascreated, listing dates in year-month-day format, paddingone-digit months and days with a zero on the left.ICRPCropped, indicating whether an image has been croppedand, if so, how it was cropped, e.g., â€œlower right cornerâ€ .IDIMDimensions, specifying the size of the original subject ofthe file, e.g., 8.5 inches in height, 11 inches in width.IDPIDots Per Inch, specifying dots per inch setting of thedigitizer used to produce the file, such as 300.IENGEngineer, specifying the name of the engineer who workedon the file. If there are multiple engineers, the names areseparated by a semicolon and a blank, e.g., Smith, John;Adams, Joe.IGNRGenre, describing the original work, such as, landscape,portrait, still life, etc.IKEYKeywords, providing a list of keywords that refer to thefile or subject of the file, with multiple keywords separatedwith a semicolon and a blank, e.g., â€œSeattle; aerial view; sceneryâ€ .ILGTLightness, describing the changes in lightness settings onthe digitizer required to produce the file, its format depending on hardware used.IMEDMedium, describing the original subject of the file, e.g.,computer image, drawing, lithograph.INAMName, storing the title of the subject of the file, such as,â€œSeattle From Aboveâ€ .IPLTPalette Setting specifying the number of colors requestedwhen digitizing an image.IPRDProduct, specifying the name of the product, for which thefile was originally intended, e.g., â€œEncyclopedia ofPacific Northwest Geographyâ€ .ISBJSubject, describing the contents of the file, e.g., â€œAerialview of Seattleâ€ .ISFTSoftware, identifying the name of the software packageused to create the file, e.g., â€œMicrosoft WaveEditâ€ .ISHPSharpness, identifying the changes in sharpness for thedigitizer required to produce the file, its formatdepending on the hardware used.ISRCSource, identifying the person or organization thatsupplied the original subject of the file.ISRFSource Form, identifying the original form of the materialthat was digitized, e.g., slide, paper, map, which may be different from IMED.ITCHTechnician, identifying the technician who digitized thesubject file.

One or more steam descriptor (â€˜str1â€™) lists follow the main header â€˜hdr1â€™. Each â€˜str1â€™ list corresponds to a data stream and includes information about the data stream in the file. A â€˜str1â€™ list contains a stream header chunk (â€˜strhâ€™) and a stream format chunk (â€˜strfâ€™). In addition, a â€˜str1â€™ list may contain a stream header data chunk (â€˜strdâ€™) and a stream name chunk (â€˜strnâ€™). The stream descriptors in the â€˜hdr1â€™ list are associated with the stream data in the â€˜moviâ€™ list according to the order of the â€˜str1â€™ lists. The first â€˜str1â€™ list applies to stream 0, the second applies to stream 1, and so forth.

The stream header chunk (â€˜strhâ€™) in the â€˜str1â€™ list includes an AVISTREAMHEADER structure containing information about a stream in the AVI file. The AVISTREAMHEADER structure has the syntax:

typedef struct_avistreamheader {FOURCCfcc;DWORDcb;FOURCCfccType;FOURCCfccHandler;DWORDdwFlags;WORDwPriority;WORDwLanguage;DWORDdwInitialFrames;DWORDdwScale;DWORDdwRate;DWORDdwStart;DWORDdwLength;DWORDdwSuggestedBufferSize;DWORDdwQuality;DWORDdwSampleSize;struct{WORDleft;WORDtop;WORDright;WORDbottom;} rcFrame;} AVISTREAMHEADER

In accordance with a preferred embodiment, the members in the AVISTREAMHEADER structure have following variables:

- fcc specifies a FOURCC, with the value being â€˜strhâ€™.
- cb specifies the size of the structure, not including the initial 8 bytes.
- fccType
  - contains a FOURCC that specifies the type of the data in the stream, with the following standard AVI values for video and audio:
  - â€˜audsâ€™ Audio stream
  - â€˜midsâ€™ MIDI stream
  - â€˜txtsâ€™ Text stream
  - â€˜vidsâ€™ Video stream
- fccHandler
  - optional, may contain a FOURCC that identifies a specific data handler preferred handler for the stream. For audio and video streams, this specifies the codec for decoding the stream.
- dwFlags
  - contains flags for the data stream. The bits in the high-order word of these flags are specific to the type of data contained in the stream. The standard flags are:
  - AVISF_DISABLEDâ€”Indicates the stream should not be enabled by default.
  - AVISF_VIDEO_PALCHANGESâ€”Indicates the video stream contains palette changes, thereby warning the playback software that it will need to animate the palette.
- dwPriority
  - specifies the priority of a stream type. For example, in a file with multiple audio streams, the one with the highest priority might be the default stream.
- dwInitialFrames
  - specifies how far audio data is skewed ahead of the video frames in interleaved files, e.g., 0.75 seconds. For an interleaved file, dwInitialFrames specifies the number of frames in the file prior to the initial frame of the AVI sequence.
- dwScale
  - specifies, in combination with dwRate, the time scale that the stream will use.
- dwRate
  - specifies, in combination with dwScale, the time scale that the stream will use. Dividing dwRate by dwScale gives the number of samples per second. For video streams, this is the frame rate. For audio streams, this rate corresponds to the time needed to play nBlockAlign bytes of audio. For pulse code modulation (PCM) audio this rate corresponds to sample rate.
- dwStart
  - specifies the starting time for this stream, with units defined by the dwRate and dwScale members in the main file header. Usually, its value is zero. A nonzero value specifies a delay time for a stream that does not start concurrently with the file.
- dwLength
  - specifies the length of the stream. The units are defined by dwRate and dwScale.
- dwSuggestedBufferSize
  - specifies how large a buffer should be used to read this stream. Preferably, it has a value corresponding to the largest chunk present in the stream. Using the correct buffer size makes playback more efficient. The value can be set to zero if the correct buffer size is unknown.
- dwQuality
  - specifies the quality of the data in the stream, represented as a number between 0 and 10,000. For compressed data, this typically represents the value of the quality parameter passed to the compression software. If set to âˆ’1, drivers use the default quality value.
- dwSampleSize
  - specifies the size of a single sample of data. This is set to zero if the samples can vary in size. For nonzero values, multiple samples of data can be grouped into a single chunk within the file. For a value of zero, each sample of data, e.g., a video frame, must be in a separate chunk. For video streams, the value is typically zero, although it can be nonzero if all video frames are the same size. For audio streams, the value should be the same as the nBlockAlign member of the WAVEFORMATEX structure describing the audio.
- rcFrame
  - specifies, in pixels, the destination rectangle for a text or video stream within the movie rectangle specified by the dwwidth and dwHeight members of the AVI main header structure. The rcFrame member is typically used in support of multiple video streams. The rectangle is preferably set to the coordinates corresponding to the movie rectangle to update the whole movie rectangle. The upper left corner of the destination rectangle is relative to the upper left corner of the movie rectangle. In accordance with the present invention the members in RcFrame may be defined as DWORD as well as WORD.

A stream format (â€˜strfâ€™) chunk follows the stream header (â€˜str1â€™) chunk. The stream format chunk describes the format of the data in the stream. The data contained in this chunk depends on the stream type.

For video streams, the information is a BITMAPINFOHEADER structure, including palette information if appropriate. The structure of BITMAPINFOHEADER is defined as:

typedef struct BITMAPINFOHEADER {DWORDbiSize;LONGbiWidth;LONGbiHeight;WORDbiPlanes;WORDbiBitCount;DWORDbiCompression;DWORDbiSizeImage;LONGbiXPelsPerMeter;LONGbiYPelsPerMeter;DWORDbiClrUsed;DWORDbiClrImportant;} BITMAPINFOHEADER

In accordance with a preferred embodiment, the members in the BITMAPINFOHEADER structure have following variables:

- BiSize
  - specifies the number of bytes required by the structure.
- BiWidth
  - specifies the width of the bitmap in pixels, or specifies the width of the decompressed JPEG image file for Microsoft Windows 98, Windows NT 5.0 and later versions if bicompression is BI_JPEG.
- BiHeight
  - specifies the height of the bitmap in pixels, or specifies the height of the decompressed JPEG image file for Microsoft Windows 98, Windows NT 5.0 and later versions if biCompression is BI_JPEG. If biHeight is positive, the bitmap is a bottom-up device independent bitmap (DIB) and its origin is the lower-left corner. If biHeight is negative, the bitmap is a top-down DIB and its origin is the upper-left corner.
- biplanes
  - specifies the number of planes for the target device. This value is set to 1.
- biBitCount
  - specifies the number of bits per pixel and determining the number of bits that define each pixel and the maximum number of colors in the bitmap. Its values and their meanings are:
  - 0 for Windows 98, Windows NT 5.0, and later, the number of bits per pixel is specified or is implied by the JPEG format.
  - 1 specifies that the bitmap is monochrome, and bmiColors contains two entries. Each bit in the bitmap array represents a pixel. If the bit is clear, the pixel is displayed with the color of the first entry in the bmiColors table. If the bit is set, the pixel has the color of the second entry in the table.
  - 4 specifies that the bitmap has a maximum of 16 colors, and bmiColors contains up to 16 entries. Each pixel in the bitmap is represented by a 4-bit index into the color table. For example, if the first byte in the bitmap is 0Ã—1F, the byte represents two pixels. The first pixel contains the color in the second table entry, and the second pixel contains the color in the sixteenth table entry.
  - 8 specifies that the bitmap has a maximum of 256 colors, and bmiColors contains up to 256 entries. Each byte in the array represents a single pixel.
  - 16 specifies that the bitmap has a maximum of 2¹⁶colors. If biCompression is BI_RGB, bmiColors is NULL. Each WORD in the bitmap array represents a single pixel. The relative intensities of red, green, and blue are represented with 5 bits for each color component. The value for blue is in the least significant 5 bits, followed by 5 bits each for green and red. The most significant bit is not used. The bmiColors color table is used for optimizing colors used on palette-based devices, and contains the number of entries specified by biClrUsed. If bicompression is BI_BITFIELDS, bmiColors member contains three DWORD color masks that specify the red, green, and blue components, respectively, of each pixel. Each WORD in the bitmap array represents a single pixel. For Windows NT: When bicompression is BI_BITFIELDS, bits set in each DWORD mask are contiguous and should not overlap the bits of another mask. All the bits in the pixel do not have to be used.
    - For Windows 95 and Windows 98: When biCompression is BI_BITFIELDS, the system supports only the following 16 bits per pixel (bpp) color masks: A 5-5-5 16-bit image, where the blue mask is 0Ã—001 F, the green mask is 0Ã—03E0, and the red mask is 0Ã—7C00; and a 5-6-5 16-bit image, where the blue mask is 0Ã—001 F, the green mask is 0Ã—07E0, and the red mask is 0Ã—F800.
  - 24 specifies that the bitmap has a maximum of 2²⁴colors, and the bmiColors member is NULL. Each 3-byte triplet in the bitmap array represents the relative intensities of blue, green, and red, respectively, for a pixel. The bmiColors color table is used for optimizing colors used on palette-based devices, and contains the number of entries specified by biClrUsed.
  - 32 specifies that the bitmap has a maximum of 2³²colors. If the biCompression member of the BITMAPINFOHEADER is BI_RGB, the bmiColors member is NULL. Each DWORD in the bitmap array represents the relative intensities of blue, green, and red, respectively, for a pixel. The high byte in each DWORD is not used. The bmiColors color table is used for optimizing colors used on palette-based devices, and must contain the number of entries specified by biClrUsed.
    - If biCompression is BI_BITFIELDS, bmiColors contains three DWORD color masks that specify the red, green, and blue components, respectively, of each pixel. Each DWORD in the bitmap array represents a single pixel. For Windows NT: When biCompression is BI_BITFIELDS, bits set in each DWORD mask must be contiguous and should not overlap the bits of another mask. All the bits in the pixel do not need to be used.
    - For Windows 95 and Windows 98: When biCompression is BI_BITFIELDS, the system supports only the following 32 bpp color mask. The blue mask is 0Ã—000000FF, the green mask is 0Ã—0000FF00, and the red mask is 0Ã—00FF0000.
- BiCompression
  - specifies the type of compression for a compressed bottom-up bitmap. If biHeight is negative, indicating a top-down DIB, biCompression must be either BI_RGB or BI_BITFIELDS. Top-down DIBs cannot be compressed. The member can be one of the following values:
  - BI_RGB
    - specifies an uncompressed format.
  - BI_RLE8
    - specifies a run-length encoded (RLE) format for bitmaps with 8 bits per pixel. The compression format is a 2-byte format consisting of a count byte followed by a byte containing a color index.
  - BI_RLE4
    - specifies an RLE format for bitmaps with 4 bits per pixel. The compression format is a 2-byte format consisting of a count byte followed by two wordlength color indexes.
  - BI_BITFIELDS
    - specifies that the bitmap is not compressed and that the color table consists of three DWORD color masks that specify the red, green, and blue components, respectively, of each pixel. This is valid when used with 16 bpp and 32 bpp bitmaps.
  - BI_JPEG
    - Indicates that the image is a JPEG image for Windows 98, Windows NT 5.0, and later versions.
- BiSizeImage
  - specifies the size, in bytes, of the image. Its value may be set to zero for BI_RGB bitmaps. For Windows 98, Windows NT 5.0, and later versions: If bicompression is JBI_JPEG, biSizeImage indicates the size of the JPEG image buffer.
- BiXPelsPerMeter
  - specifies the horizontal resolution, in pixels per meter, of the target device for the bitmap. An application can use this value to select a bitmap from a resource group that best matches the characteristics of the current device.
- BiYPelsPerMeter
  - specifies the vertical resolution, in pixels per meter, of the target device for the bitmap.
- BiClrUsed
  - specifies the number of color indexes in the color table that are actually used by the bitmap.
  - If BiClrUsed is zero, the bitmap uses the maximum number of colors corresponding to the value of the biBitCount for the compression mode specified by biCompression.
  - If biClrUsed is nonzero and biBitCount is less than 16, biClrUsed specifies the actual number of colors the graphics engine or device driver accesses. If biBitCount is 16 or greater, biClrUsed specifies the size of the color table used to optimize performance of the system color palettes. If biBitCount equals 16 or 32, the optimal color palette starts immediately following the three DWORD masks.
  - If the bitmap is a packed bitmap (a bitmap in which the bitmap array immediately follows the BITMAPINFOHEADER and is referenced by a single pointer), biClrUsed should be either zero or the actual size of the color table.
- BiClrImportant
  - specifies the number of color indexes that are required for displaying the bitmap. If its value is zero, all colors are required.

For audio streams, the information is a WAVEFORMATEX structure. For text streams, the information has a TEXTINFO structure:

typedef_textinfo {WORDwCodePage;WORDwCountryCode;WORDwLanguageCode;WORDwDialect} TEXTINFO

where the meaning of all the fields (wCodePage, wCountryCode, wLanguageCode, and wDialect) is the same as those defined above with reference to CSET chunk. Different languages can be set for each of the text streams in a file having multiple text streams.

If the optional stream header data (â€˜strdâ€™) chunk is present in an AVI file, it follows the stream format chunk. The format and content of the â€˜strdâ€™ chunk are defined by the codec driver. Typically, drivers use this information for configuration. Applications that read and write AVI files do not need to interpret this information, they simple transfer it to and from the driver as a memory block.

The information block for achieving the digital rights management (DRM) protection in the AVI file is presented in the â€˜strdâ€™ chunk associated with the main video stream. The format of the DRM information data for the video stream in the â€˜strdâ€™ should be as following:

typedef_DRMinfo{WORD wVersion;STR sDRMInfo;} DRMINFO

where the two members in the structure DRMINFO are defined as:

- wVersion
  - specifies the version of the DRM.
- sDRMInfo
  - specifies the information for the DRM protection, e.g., in an encrypted binary string.

The optional stream name â€˜strnâ€™ chunk includes a null terminated text string describing the stream. In accordance with an embodiment of the present invention, the string is â€œVideo [â€”Description]â€ for a video stream, where optional [â€”Description] part is any string that describes the video stream, e.g., it can be â€œVideoâ€”Mainâ€ . For an audio stream, the string can be â€œAudio [â€”Description]â€ , where the optional [â€”Description] part is any string that describes the audio stream, e.g., it can be â€œAudioâ€”Englishâ€ , â€œAudioâ€”Frenchâ€ , â€œAudioâ€”Mainâ€ , or â€œAudioâ€”Auxiliaryâ€ , etc. For a chapter stream, which is a text stream, the string can be â€œChapter [â€”Description]â€ , where the optional [â€”Description] part is any string that describes the chapter stream. For a subtitle stream, which can be either a text stream or a video stream, the string can be â€œSubtitle [â€”Description]â€ , where the optional [â€”Description] part is any string that describes the subtitle stream, e.g., it can be â€œSubtitleâ€”Englishâ€ , or â€œSubtitleâ€”Chineseâ€ .

AVI stream data â€˜moviâ€™ list follows the header information in the AVI RIFF file format. The â€˜moviâ€™ list contains the actual data in the streams, e.g., the video frames and audio samples. The data chunks can reside directly in the â€˜moviâ€™ list, or be grouped together as subchunks within â€˜recâ€™ lists. The â€˜recâ€™ grouping implies that the grouped subchunks should be read from disk all at once, and is intended for files that are interleaved to play from CD-ROM.

Each data chunk in the â€˜moviâ€™ list is identified by a FOURCC that includes a two-digit stream number followed by a two-character code that defines the type of information in the chunk. In accordance with an embodiment of the present invention, the two-character codes for defining the data type are:

dbuncompressed video framedecompressed video frameddDRM key info for the video framepcpalette changewbaudio datastsubtitle (text mode)sbsubtitle (bitmap mode)chchapter

It should be noted that, in accordance with the present invention, additional two-character codes may be used to identify data streams not specified herein above.

By way of example, if stream 0 contains audio, the FOURCC for the stream would be â€˜00wbâ€™. If stream 1 contains video, the FOURCC for the stream would be â€˜01dbâ€™ for uncompressed video or â€˜01dcâ€™ for compressed video. Video data chunks can also define new palette entries to update the palette during an AVI sequence. Each palette-change chunk (â€˜xxpcâ€™) contains an AVIPALCHANGE structure. If a stream contains palette changes, the AVISF_VIDEO_PALCHANGES flag in the dwFlags member of the AVISTREAMHEADER structure for that stream is set accordingly.

The optional index list follows the â€˜moviâ€™ list in the AVI RIFF file format. The index contains a list of the data chunks and their location in the file. If the AVI file contains an index, the dwFlags member of the AVIMAINHEADER structure is set to AVIF_HASINDEX.

In version AVI 1.0, the index (â€˜idx1â€™) list includes an AVIOLDINDEX structure with entries for each data chunk, including â€˜recâ€™ chunks. The AVIOLDINDEX structure has the syntax:

typedef struct_avioldindex {FOURCCfcc;DWORDcb;struct_avioldindex_entry {DWORDdwChunkId;DWORDdwFlags;DWORDdwOffset;DWORDdwSize;} aIndex [ ];} AVIOLDINDEX

The members in the structure have following characters:

fccspecifies a FOURCC code, with the value â€˜idx1â€™.cbspecifies the size of the structure, not includingthe initial 8 bytes.DwChunkIdspecifies a FOURCC that identifies a stream in theAVI file, having the form â€˜nnyyâ€™ where nn is thestream number and yy is a two-character code thatidentifies the contents of the stream:dbuncompressed video framedccompressed video framepcpalette changewbaudio datadwFlagsspecifies a bitwise combination of zero or more ofthe following flags:AVIIF_LIST0x00000001L// The data chunkis a â€˜recâ€™ list.AVIIF_KEYFRAME0x00000010L// The data chunkis a key frame.AVIIF_NO_TIME0x00000100L// The data chunkdoes not affect the timing of the stream,e.g., for palette changes.AVIIF_NO_COMPUS0x0FFF0000L// The data arefor compressor use.DwOffsetspecifies the location of the data chunk in thefile. In one embodiment, the value is specified asan offset, in bytes, from the start of the â€˜moviâ€™list. In another embodiment, the value is theoffset from the start of the file.DwSizespecifies the size of the data chunk, in bytes.

In accordance with a preferred embodiment, the AVIOLDINDEX structure includes the initial RIFF chunk (the fcc and cb members) followed by one index entry for each data chunk in the â€˜moviâ€™ list. The AVIOLDINDEX structure describes an AVI 1.0 index (â€˜idx1â€™ format). New AVI files should use an AVI 2.0 index (â€˜indxâ€™ format).

Additional data can be aligned in an AVI file by inserting â€˜JUNKâ€™ chunks as needed. Applications will ignore the contents of a â€˜JUNKâ€™ chunk.

In accordance with the present invention, the video tracks of one or more movies are stored in an AVI file as AVI video streams or tracks. A single AVI file may include multiple video tracks. Preferably, the first of the multiple video tracks is the main video track.

The stream descriptor (â€˜str1â€™) list for a video stream should include a stream header (â€˜strhâ€™) chunk, a stream format (â€˜strfâ€™), a stream header data (â€˜strdâ€™) chunk if the stream is DRM protected, and a stream name (â€˜strnâ€™) chunk. In accordance with an embodiment of the present invention, the member fccType in the structure AVISTREAMHEADER the stream header (â€˜strhâ€™) chunk for a video stream has the value â€˜vidsâ€™. The stream header data (â€˜strdâ€™) chunk of a video stream should exist only for DRM protected video. If the â€˜strdâ€™ chunks exists, the video stream is protected, and there will be â€˜xxddâ€™ DRM chunks in the video stream. The stream name data (â€˜strnâ€™) chunk for a video stream includes a string of the form â€œVideo [â€”Descriptions]â€ .

The stream data (â€˜moviâ€™) list of a video stream in includes an â€˜nndbâ€™ chunk for an uncompressed video data chunk or an â€˜nndcâ€™ for a compressed video data chunk, where â€˜nnâ€™ is a two digit data chunk index. If a video data chunk is DRM protected, the â€˜moviâ€™ list also includes a â€˜nnddâ€™ chunk preceding the corresponding â€˜nndbâ€™ or â€˜nndcâ€™ chunk of the protected video data chunk. In accordance with a specific embodiment of the present invention, the member dwFlags in the structure AVIOLDINDEX of the index entry for the â€˜nnddâ€™ chunk is set to AVIF_NO_TIME.

In one embodiment of the present invention, each video data chunk includes one video frame in variable bit rate coding. For video frames encoded in predicted frames (P frames) and bidirectional frames (B frames), a B frame is preferably placed in a chunk with the following P frame. In such cases, an uncoded dummy P frame (N in the following illustration) is preferably inserted by the codec to keep the timing. For example, a sequence of image frames (I frames), B frames, and P frames I_mB_m+1P_m+2P_m+3P_m+4. . . is preferably arranged into the following video stream chunk sequence:

- [I_m] [P_m+2, B_m+1] [N] [P_m+4, B_m+3] [N] . . .
  In the expression, the square brackets indicate the data chunks in the AVI stream.

In accordance with the present invention, the audio tracks of one or more movies are stored in an AVI file as AVI audio streams or tracks. A single AVI file may include multiple audio tracks. Preferably, the first of the multiple audio tracks is the main audio track.

The stream descriptor (â€˜str1â€™) list for an audio stream should include a stream header (â€˜strhâ€™) chunk, a stream format (â€˜strfâ€™), and a stream name (â€˜strnâ€™) chunk. In accordance with a specific embodiment the â€˜str1â€™ list for an audio stream does not include the stream header data (â€˜strdâ€™) chunk. In this embodiment, the application should ignore any data chunk with the â€˜strdâ€™ code in the steam descriptor (â€˜str1â€™) list of the AVI file.

In accordance with an embodiment of the present invention, the member fccType in the structure AVISTREAMHEADER the stream header (â€˜strhâ€™) chunk for a video stream has the value â€˜audsâ€™. The stream name data (â€˜strnâ€™) chunk for a video stream includes a string of the form â€œAudio [â€”Descriptions]â€ .

The stream data (â€˜moviâ€™) list of an audio stream in includes an â€˜nnwbâ€™ chunk for identifying an audio data chunk, where â€˜nnâ€™ is a two digit data chunk index. In one embodiment of the present invention, each audio data chunk includes one audio frame in variable bit rate coding. In another embodiment of the present invention, each audio data chunk includes one or more audio frames in constant bit rate coding.

In accordance with the present invention, the chapter tracks are stored in an AVI file as AVI text streams or tracks. A single AVI file may include multiple chapter tracks. The stream descriptor (â€˜str1â€™) list for a chapter stream should include a stream header (â€˜strhâ€™) chunk, a stream format (â€˜strfâ€™), and a stream name (â€˜strnâ€™) chunk. In accordance with a specific embodiment the â€˜str1â€™ list for a chapter stream does not include the stream header data (â€˜strdâ€™) chunk. In this embodiment, the application should ignore any data chunk with the â€˜strdâ€™ code in the steam descriptor (â€˜str1â€™) list of the AVI file.

In accordance with an embodiment of the present invention, the member fccType in the structure AVISTREANHEADER the stream header (â€˜strhâ€™) chunk for a video stream has the value â€˜txtsâ€™. The stream format (â€˜strfâ€™) chunk for a chapter stream has the TEXTINFO structure. The stream name data (â€˜strnâ€™) chunk for a video stream includes a string of the form â€œChapter [â€”Descriptions]â€ .

The stream data (â€˜moviâ€™) list of a chapter stream in includes an â€˜nnchâ€™ chunk for identifying a chapter data chunk, where â€˜nnâ€™ is a two digit data chunk index. In one embodiment of the present invention, each chapter data chunk has a CHAPTERCHUNK structure:

typedef struct_chapterchunk {FOURCCfcc;DWORDcb;STRtime;STRdescription} CHAPTERCHUNK

The members in the structure CHAPTERCHUNK are

- fcc specifies a FOURCC code having the value â€˜nnxxâ€™.
- cb specifies the size of the structure, not including the initial 8 bytes.
- time specifies the time at the starting of the chapter, having the form [hh:mm:ss.xxx], where hh is a two digit number representing the hours, mm a two digit number representing the minutes, ss a two digit number representing the seconds, and xxx a three digit number representing the milliseconds, of the starting point of the chapter.
- Description
  - specifies the name of the chapter or other description information.

The chapter stream is not a regular interval stream. In accordance with a specific embodiment of the present invention, the member dwFlags in the structure AVIOLDINDEX of the index entry for the â€˜nnchâ€™ chunk is set to AVIF_NO_TIME.

In accordance with one embodiment of the present invention, the subtitle tracks are stored in an AVI file as AVI text streams or tracks. In accordance with another embodiment of the present invention, the subtitle tracks are stored in an AVI file as AVI bitmap streams or tracks. A single AVI file may include multiple subtitle tracks. The stream descriptor (â€˜str1â€™) list for a subtitle stream should include a stream header (â€˜strhâ€™) chunk, a stream format (â€˜strfâ€™), and a stream name (â€˜strnâ€™) chunk. In accordance with a specific embodiment the â€˜str1â€™ list for a subtitle stream does not include the stream header data (â€˜strdâ€™) chunk. In this embodiment, the application should ignore any data chunk with the â€˜strdâ€™ code in the steam descriptor (â€˜str1â€™) list of the AVI file.

In accordance with an embodiment of the present invention, the member fccType in the structure AVISTREAMHEADER the stream header (â€˜strhâ€™) chunk for a video stream has the value â€˜txtsâ€™ for text form subtitles or â€˜vidsâ€™ for bitmap form subtitles. The stream format (â€˜strfâ€™) chunk for a subtitle stream has the TEXTINFO structure for text form subtitles and the BITMAPINFOHEADER structure for bitmap form subtitles. The stream name data (â€˜strnâ€™) chunk for a video stream includes a string of the form â€œSubtitle [â€”Descriptions]â€ .

The stream data (â€˜moviâ€™) list of a subtitle stream includes an â€˜nnstâ€™ chunk for identifying a text form subtitle data chunk and/or an â€˜nnsbâ€™ chunk for identifying a bitmap form subtitle data chunk, where â€˜nnâ€™ is a two digit data chunk index. In one embodiment of the present invention, each subtitle data chunk has a SUBTITLECHUNK structure:

typedef struct_subtitlechunk {FOURCCfcc;DWORDcb;STRduration;STRsubtitle} SUBTITLECHUNK

The members in the structure SUBTITLECHUNK are

- fcc specifies a FOURCC code having the value â€˜nnxxâ€™.
- cb specifies the size of the structure, not including the initial 8 bytes.
- duration
  - specifies the time at the starting of the chapter, having the form [hh:mm:ss.xxxâ€”HH:MM:SS:XXX], where hh and HH are a two digit numbers representing the hours, mm and MM two digit numbers representing the minutes, ss and SS two digit numbers representing the seconds, and xxx and XXX three digit numbers representing the milliseconds, of the starting point and ending point, respectively, for displaying the subtitles.
- subtitle
  - contains either the Unicode text of the subtitles for text mode, or a compressed bitmap image of the subtitles for bitmap mode.

The subtitle stream is not a regular interval stream. In accordance with a specific embodiment of the present invention, the member dwFlags in the structure AVIOLDINDEX of the index entry for the subtitle chunk is set to AVIF_NO_TIME.

For bitmap format subtitles, it is preferred to have compressed subtitle bitmaps in the subtitle field in the subtitle chunks. A compressed subtitle bitmap will have the following fields:

WORDwidth;WORDheight;WORDleft;WORDtop;WORDright;WORDbottom;struct {BYTEred;BYTEgreen;BYTEblue;} color_background, color_pattern, color_emphasis1,color_emphasis2;BITMAP bitmap;

The â€œwidthâ€ and â€œheightâ€ fields specify the dimension of the subtitle bitmap. The members â€œleftâ€ , â€œtopâ€ , â€œrightâ€ , and â€œbottomâ€ fields specify the display rectangle of the subtitle bitmap relative to the main video rectangle. The BITMAP includes compressed bitmap data.

In accordance with a preferred embodiment, the subtitle bitmaps are four-level bitmaps with the following definition.

- 00 Background pixel
- 01 Pattern pixel
- 10 Emphasis pixelâ€”1
- 11 Emphasis pixelâ€”2

Compression of the subtitle bitmap uses a simple run-length coding according the rules in Table 4. In accordance with an embodiment of the present invention, the size of the run-length coded data within one line is 1440 bits or less.

In accordance with a preferred embodiment of the present invention, the streams in AVI files are interleaved. Audio stream chunks are interleaved ahead of corresponding video stream chunks in time. The amount of the audio stream that is interleaved ahead of corresponding points in the video stream should not exceed an predetermined upper limit, e.g., 2 seconds, 5 seconds, 10 seconds, 15 seconds, etc. The subtitle chunks are interleaved in the file ahead of the corresponding video chunk in time, with the amount of subtitle interleaved ahead of corresponding points in the video stream not exceeding a upper limit, e.g., 5 seconds, 10 seconds, 15 seconds, 20 seconds, etc. The interleaving of the chapter stream is not restricted. It could be all written in the beginning of the â€œmoviâ€ list, or interleaved with the other streams.

TABLE 4Subtitle bitmap coding rulesBitmap PixelsCoding1 to 3 pixels with theEnter the number of pixel(s) followed in thesame valuefirst 2 bits and the pixel data in the next 2 bits.follow(s).The 4 bits are considered to be one unit.4 to 15 pixels withSpecify â€˜0â€™ in the first 2 bits, and enter thethe same valuenumber of pixels in the following 4 bits andfollows.the pixel data in the next 2 bits. The 8 bits areconsidered to be one unit.16 to 63 pixels withSpecify â€˜0â€™ in the first 4 bits, and enter thethe same valuenumber of pixels in the following 6 bits andfollows.the pixel data in the next 2 bits. The 12 bitsare considered to be one unit.64 to 255 pixels withSpecify â€˜0â€™ in the first 6 bits, and enter thethe same valuenumber of pixels in the following 8 bits andfollows.the pixel data in the next 2 bits. The 16 bitsare considered to be one unit.The same pixelsSpecify â€˜0â€™ in the first 14 bits, and describefollow to the end ofthe pixel data in the following 2 bits. The 16a line.bits are considered to be one unit.The byte alignment isInsert dummy data of 4 bits â€˜0000bâ€™not accomplished whenfor adjustment.the description forpixels on one line iscompleted.

An AVI file typically does not contain a time stamp of the streams. Each stream has its frame rate specified in the stream descriptor (â€˜str1â€™) list in the AVI header. For variable bit rate streams such as video streams or variable bit rate audio streams, each chunk contains one and only one frame. Accessing the data of the variable bit rate stream at any given point is feasible with the known frame rate and the data chunk index. For constant bit rate streams, e.g., constant bit rate audio streams, each chunk may contain one or more frames. Because each frame has a known fixed size, locating data at any given point can be achieved by calculating the size of the stream data. Therefore, seeking an arbitrary location in an AVI file in accordance with the present invention can be achieved for either constant bit rate or variable bit rate streams by parsing and recording the index table for each frame.

Many playback devices, particularly consumer electronics devices such as DVD players, are not able to input pointers to arbitrary points as can a slider bar used in computer software. For such devices, it is beneficial to only record the chapter location, i.e., the starting point of audio, video, and subtitles, while parsing the index. For a memory restricted player, it may be preferred for the player to remember index records at the minute points to reduce memory usage, thereby saving limited memory space. The full index is not required during normal forward play because the chunk is self-contained.

In accordance with the present invention, the version of the video codec used in AVI files is signaled by the FourCC code in the fccHandler field or member of the AVISTREAMHEADER of the corresponding stream header â€˜strhâ€™ chunks, and the FourCC code bicompression field or member in the BITMAPINFOHEADER of the corresponding â€˜strfâ€™ chunks.

By way of example, for videos encoded according to a codec developed by DivX Networks, Inc., 10350 Science Center Drive, Building 14, Suite 140, San Diego, Calif. 92121, the FourCC codes fccHandler in the stream header (â€˜strhâ€™) of the AVISTREAMHEADER is set to â€œdivxâ€ or â€œDIVXâ€ . Furthermore, the FourCC (DWORD) code biCompression in the BITMAPINFOHEADER of the corresponding â€˜strfâ€™ chunks is set to signify the detailed codec version.

Specifically by way of example, for version DivX 3.11, â€˜div3â€™ or â€˜div4â€™ is used in AVISTREAMHEADER, and â€˜div3â€™ or â€˜div4â€™ is used in BITMAPINFOHEADER; for version DivX 4.x, â€˜divxâ€™ is used in AVISTREAMHEADER, and â€˜divxâ€™ is used in BITMAPINFOHEADER; and for version DivX 5.x, â€˜divxâ€™ is used in AVISTREAMHEADER, and â€˜dx50â€™ is used in BITMAPINFOHEADER.

By now it should be appreciated that a file format for storing digital data with a high compression rate has been described. A file format in accordance with the present invention is compatible with high level data compressing algorithms, such as MPEG-4. Its data compression ratio is about six to ten times higher than a standard DVD format. In accordance with the present invention, the file format is capable of storing data in multiple streams or tracks. The file format is also able to encode and archive video, audio, and text data on easily accessible streams or tracks. Furthermore, the file format is able to provide protection of the copyright of the digitized content.

While the invention is susceptible to various modifications and alternative constructions, certain illustrated embodiments thereof have been described above in detail. It should be understood, however, that there is no intention to limit the invention to the specific form or forms disclosed, but on the contrary, the intention is to cover all modifications, alternative constructions, and equivalents falling within the spirit and scope of the invention. The present invention is limited only by the claims that follow.