diff options
| -rw-r--r-- | FORMAT | 14 | ||||
| -rw-r--r-- | README | 67 | ||||
| -rw-r--r-- | SPECIFICATION | 56 |
3 files changed, 81 insertions, 56 deletions
@@ -0,0 +1,14 @@ + + FARBFELD IMAGE FORMAT SPECIFICATION + + +--------+-----------------------------------------------------------+ + | Bytes | Description | + +--------+-----------------------------------------------------------+ + | 8 | "farbfeld" magic value | + +--------+-----------------------------------------------------------+ + | 4 | 32-Bit BE unsigned integer (width) | + +--------+-----------------------------------------------------------+ + | 4 | 32-Bit BE unsigned integer (height) | + +--------+-----------------------------------------------------------+ + | [2222] | 4*16-Bit BE unsigned integers [RGBA] / pixel, row-aligned | + +--------+-----------------------------------------------------------+ @@ -0,0 +1,67 @@ + + ███ ███ ██ ██ ███ ███ █ ██ + █ █ █ █ █ █ █ █ █ █ █ █ + ██ ███ ███ ███ ██ ██ █ █ █ + █ █ █ █ █ █ █ █ █ █ █ █ + █ █ █ █ █ ██ █ ███ ███ ██ + + +WHAT IS FARBFELD? + Farbfeld is a lossless image-format designed to be + parsed and piped easily. + It can be compressed easily and beats PNG's filesize + in many cases. + +WHY FARBFELD? + Current image-formats have integrated compression, + making it complicated to read the image data. + One is forced to use complex libraries like libpng, + libjpeg, libjpeg-turbo, giflib and others, read the + documentation and write a lot of boilerplate in order + to get started. + Farbfeld leaves this behind and is designed to be as + simple as possible, leaving the task of compression + to outside tools. + The simple design, which was the primary objective, + implicitly lead to the very good compression + characteristics, as it often happens when you go with + the UNIX philosophy. + Reading farbfeld images doesn't require any special + libraries. The tools in this folder are just a toolbox + to make it easy to convert between common image formats + and farbfeld. + +HOW DOES IT WORK? + In Farbfeld, pattern resolution is not done while + converting, but while compressing the image. + For example, farbfeld always stores the alpha-channel, + even if the image doesn't have alpha-variation. + This may sound like a big waste at first, but as + soon as you compress an image of this kind, the + compression-algorithm (e.g. bz2) recognizes the + pattern that every 48 bits the 16 bits store the + same information. + And the compression-algorithms get better and better + at this. + Same applies to the idea of having 16 bits per channel. + It sounds excessive, but if you for instance only have + a greyscale image, the R, G and B channels will store + the same value, which is recognized by the compression + algorithm easily. + This effectively leads to filesizes you'd normally only + reach with paletted images, and in some cases bz2 even + beats png's compression, for instance when you're dealing + with grayscale data, line drawings, decals and even + photographs. + +HOW DO I USE THE TOOLS? + encoding: + png2ff < example.png > example.ff + png2ff < example.png | bzip2 > example.ff.bz2 + + decoding: + ff2png < example.ff > example.png + bzcat example.ff.bz2 | ff2png > example.png + + bz2 is recommended for compression, but you can use + any algorithm you want. diff --git a/SPECIFICATION b/SPECIFICATION deleted file mode 100644 index 837fe0e..0000000 --- a/SPECIFICATION +++ /dev/null @@ -1,56 +0,0 @@ -The farbfeld-format is meant to be parsed easily -and used to pipe images losslessly. - -# WHY FARBFELD? - -Most current image-formats have their compression -incorporated in their format itself. -This has some advantages, but reaches its limits -with lossless formats (e.g. PNG). -The basic idea of the farbfeld-format is to separate -these to and having a completely transparent image-format. - -Pattern resolution is done while compressing, not while -converting the image. -For example, farbfeld always stores an alpha-channel, -even if the image doesn't have alpha-variation. -This may sound like a big waste, but as soon as you -compress an image of this kind, the bzip2-algorithm -takes care of the easy pattern, that each 4th character -has the same value. -This leads to almost no overhead while keeping parsing -really simple. -Same applies to the idea of having 64 bits per channel. -It sounds excessive, but if you only have 8 bits of -entropy, the compression will take care of it, while -you only need to do the endian-handling for 64 bit -chunks. - -# FORMAT: - -Bytes Description -8 farbfeld -4 32 bit BE Unsigned Integer (width) -4 32 bit BE Unsigned Integer (height) -[2222] RGBA-row-aligned-pixel-array - 16 bit BE Unsigned Integers (per channel) - -# EXAMPLES: - -encoding: -png2ff < example.png > example.ff -png2ff < example.png | bzip2 > example.ff.bz2 - -decoding: -ff2png < example.ff > example.png -bzcat example.ff.bz2 | ff2png > example.png - -# WHY BZ2? - -Using BZ2, you can get smaller filesizes than with PNG, -especially with line-drawings. -For normal pictures, the bz2-compression yields roughly -the same sizes as png does. -Always keep in mind that using PNG involves having to -rely on libpng to decode the image for you, whereas -farbfeld is a completely transparent format. |