BTIC5A

BT Palette Image (BTIC 5A)

• FOURCC: 'bt5a' ?

Focus will be on static images or short animations (animated textures or sprites), with a relatively low-cost decoder. Image quality isn't a particularly high priority.

Will internally use a color palette, but decoded frames will be assumed to use HiColor or similar. The palette may be either encoded with each I-Frame, or a static global palette may be used.

It may also be used as a video codec if low complexity decoding is needed.

Palettes will be stored as RGB555:

• 0rrr-rrgg-gggb-bbbb
• Nominally, color 255 will be reserved for a transparent color.
• A default palette may also be used.
  • The MSB bit will conventionally be assumed to be an alpha bit.
  • Setting this bit for Color 255 may indicate that transparency is used.

Default Palette:

• 0..15: Traditional 16-color palette (RGBI)
• 16..31: Grayscale, 16 shades.
• 32..247: 216 color palette
• 248..254: Special Colors
• 255: Transparent

This palette may be used if no dynamic palette is used.

Frame TLV:

• 0x00..0x1F: 8K, TWOCC
• 0x20..0x3F: 8K, FOURCC
• 0x40..0x5F: 512M, TWOCC
• 0x60..0x7F: 512M, FOURCC
• 0x80..0x9F: 2M, TWOCC
• 0xA0..0xBF: 2M, FOURCC

TwoCC:

• 'IX': Image Data (I-Frame), No LZ
• 'PX': Image Data (P-Frame), No LZ
• 'HX': Optional Header
• 'PT': Color Palette
• 'Z3': Frame Encoded with RP2
• 'Z4': Frame Encoded with LZ4
  • LZ Compressed frames will be preceded by a small 32-bit header:
    • (23: 0): Uncompressed Size
    • (31:24): Reserved / Zero
  • LZ may not be applied recursively.

Header:

• u16 width; //image frame width
• u16 height; //image frame height
• u16 flags; //some control flags
  • 1: Uses simple skip rather than motion compensation.
    • This mode does not need a second block buffer or copying in the decoder.

The frame size will be internally padded up to a multiple of 4 pixels, though the width and height will represent the actual / non-padded resolution.

Image Data

Frame Data:

• ppppppp0 pppppppp aaaaaaaa bbbbbbbb //Raw Simple Block (4x4x1, 2-color)
  • Emit block, selecting A or B based on each pixel bit.
  • Updates both colors A and B.
• yyyxxx01, Small Delta Vector (0..6 => -3..+3)
  • Copy block with small delta relative to prior vector.
• nnn11101, Short Repeat Vector
  • Copy blocks repeating the last vector.
• 111nnn01, Short Skip (Vector set to 0,0)
  • Copy blocks while setting motion vector to zero.
• 11111101, Reserved
• nn000011 nnnnnnnn, Larger Skip / Repeat
  • Copy larger run of blocks reusing prior vector.
  • Functions like a Skip if the vector is zero.
• nn001011 yyyyxxxx, Copy 1..4 blocks with Absolute Vector
  • Vector range is (0..14 => -7..+7), 15 is reserved.
• nn010011 aaaaaaaa, Flat Color (1..4 blocks)
  • Updates color A.
• nn011011 bbbbbbbb, Flat Color (1..4 blocks)
  • Updates color B.
• nn100011 (n * 1B), Emit 2..8 blocks, 2x2x1, reusing prior colors
• nn101011 (n * 2B), Emit 1..4 blocks, reusing prior colors
• nn110011 (n * 8B), Emit 1..4 blocks, 4-color + 4x4x2bpp
• nn111011 (n * 16B), Emit 1..4 blocks, each with 16 raw pixels.
• pppp0111, 2x2x1 block, reusing prior colors.
• nnp01111, Emit 1..4 flat blocks, reuse colors, P selects A or B.

I-Frame:

• Always updates the whole image.
• I-Frame images may use secondary LZ compression (RP2).

P-Frame:

• Will update the image based on the prior image;
• P-Frame images may use secondary LZ compression (RP2).
• Motion compensation will be based on a relative number of blocks.

Pixel Bits will be 1bpp, where 0=ColorA, 1=ColorB.

Pixel ordering will be LSB first, with the LSB corresponding to the lower-left corner of a block in raster order:

• 12 13 14 15
• 8 9 10 11
• 4 5 6 7
• 0 1 2 3

With the image nominally being in raster order at the lower-left corner.

RP2

BtRP2 is a yet another byte-based LZ77 variant.

BtRP2 (Transposed, LE):

• dddddddd-dlllrrr0 (l=3..10, d=0..511, r=0..7)
• dddddddd-dddddlll-lllrrr01 (l=4..67, d=0..8191)
• dddddddd-dddddddd-dlllllll-llrrr011 (l=4..515, d=0..131071)
• rrrr0111 (Raw Bytes, r=(r+1)*8, 8..128)
• * rrr01111 (Long Match)
• rr011111 (r=1..3 bytes, 0=EOB)
• rrrrrrrr-r0111111 (Long Raw, r=(r+1)*8, 8..4096)
  • d: Distance
  • l: Match Length
  • r: Literal Length

Values are encoded in little-endian order, with tag bits located in the LSB. Bits will be contiguous within the value, with shift-and-mask being used to extract individual elements.

Long Match will encode length and distance using variable-length encodings directly following the initial tag byte.

Length VLN:

• lllllll0, 4.. 131
• llllllll-llllll01, 132..16383

Distance VLN:

• dddddddd-ddddddd0, 32K (0..32767)
• dddddddd-dddddddd-dddddd01, 4M