This program is used to convert images to particular color formats and source includes. It primarily is used for the TI-84+CE and related calculator series, however can be used as a standalone program.
Usage:
convimg [options]
Optional options:
-i, --input <yaml file> Input file, format is described below.
-n, --new Create a new template YAML file.
-h, --help Show this screen.
-v, --version Show program version.
-c, --clean Deletes files listed in 'convimg.out' and exits.
-l, --log-level <level> Set program logging level:
0=none, 1=error, 2=warning, 3=normal
Optional icon options:
--icon <file> Create an icon for use by shell.
--icon-description <txt> Specify icon/program description.
--icon-format <fmt> Specify icon format, 'ice' or 'asm'.
--icon-output <output> Specify icon output filename.
YAML File Format:
A YAML file is a human-readable text file that is used by this program
to create palettes, convert images, and output the results to a supported
format. A palette is a list of colors that an image or group of images uses.
This saves memory by having a fixed number of colors rather than storing
a color for every pixel in the image.
The YAML file consists of three key components: palette generation, image
conversion, and output formatting. These components are broken down into
different sections of a YAML file, shown below:
palettes:
- name: mypalette : Palette and name of palette
images: automatic : Automatically get list of images
converts:
- name: myimages : Convert images section
palette: mypalette : Use the above generated palette
images: : List of input images
- image1.png
- image2.png
- image3.png
outputs: : List of outputs
- type: c : Sets the output format to C
include-file: gfx.h : File name containing all images/palettes
palettes: : Start of palettes list in the output
- mypalette : Name of each palette, separated by line
converts: : Converted sections, based on name
- myimages : Name of each convert section
To create the above template, use the '--new' command line option.
By default 'convimg.yaml' in the same directory the program is executed
will be used for conversion - this can be overriden with the -i flag.
The following describes each section of the YAML file in more detail, and
the available options for conversion.
--------------------------------------------------------------------------------
palettes:
The YAML palettes section constists of a list of different palettes to be
generated. List entires start with the name of the palette to create.
- name: <palette name> : Name of palette. This is used by the
: convert blocks for detecting which
: palette to use, as well as the name
: used in the output code files.
: Required parameter.
max-entries: <num> : Number of maximum colors available
: in the palette. Range is 2-256.
: Default is 256.
fixed-entries: : Adds fixed color entries to the
: palette. The format is:
:
: fixed-entries:
: - color: {index: 0, r: 9, g: 10, b: 0}
: - color: {index: 1, r: 2, g: 83, b: 5}
: - color: {index: 2, hex: "#945D3A"}
: - color: {index: 3, hex: "#C54DCE"}
:
: Note the spaces between key/value pairs.
: 'index' represents the palette entry
: used for the color, r is the red amount,
: g is the green amound, and b is the blue
: amount of the color. This is often
: called the "rgb" value of a color.
: A quoted "hex" string beginning with '#'
: can also be used instead.
: 'exact' is used to match colors exactly
: to the input color before quantization.
: It is an optional parameter, defaults to
: 'false' but can be 'true'.
: Fixed entries can also be supplied via
: an image, 1 pixel in height (and a
: width <= 256) where each pixel in the
: image represents an entry:
:
: fixed-entries:
: - image: palette.png
:
: There can be up to 256 fixed entries.
: With the image method, the starting
: index is 0 and increments by 1 moving
: from left to right.
quality: <value> : Sets the quality of the generated palette.
: Value range is 1-10, where 1 is the worst
: and 10 is the best. Lower quality may use
: less palette entries.
: Default is 8.
images: <option> : A list of images separated by a newline
: and indented with a leading '-'
: character. These images are quantized
: in order to produce a palette that
: best matches with these images.
: <option> can be 'automatic' to
: automatically get the images from the
: convert blocks that are using them,
: without having to specify the images
: twice. Note that common "globbing"
: patterns can be used, such as '*'.
--------------------------------------------------------------------------------
converts:
The YAML converts section constists of a list of different "groups" of
images to be converted that share similar properties, such as the palette
used, or the compression mode. The available options are:
- name: <convert name> : Name of convert. This is used by the
: outputs for detecting which converted
: images to output.
: Required parameter.
palette: <palette name> : Name of palette used to quantize images
: against. This is the name of a palette
: block. Built-in palettes also exist,
: namely 'xlibc' and 'rgb332'.
palette-offset: <index> : Converts assuming a palette has
: a different base index.
: Defaults to '0'.
images: : A list of images separated by a newline
: and indented with a leading '-'
: character. These images are quantized
: against the palette.
: Note that common "globbing"
: patterns can be used, such as '*'.
tilesets: : A group of tilesets. There can be only one
: tileset group per conversion block.
: Tilesets are special images consisting of
: multiple "tiles" of fixed width and
: height. The format for this option is:
:
: tilesets:
: tile-width: 16
: tile-height: 16
: images:
: - tileset1.png
: - tileset2.png
:
: Above, each tileset in the list is
: composed of tiles of a width and height
: of 16 pixels. Tile numbers are determined
: starting from the top left, moving right
: to the bottom-right.
: Optional fields are:
: 'tile-rotate': rotate tiles 90, 180, 270.
: 'tile-flip-x': flip tiles across x axis.
: 'tile-flip-y': flip tiles across y axis.
: 'pointer-table': output tile pointers
transparent-index: <index> : Transparent color index in the palette
: that represents a transparent color.
: This only is used in two cases!
: When converting 'rlet' images,
: or for assigning pixels in the source
: image that have an alpha channel.
: If the alpha chanel is 0 in an pixel,
: the converted output will be assigned
: this value.
: Generally, you want transparency to be.
: defined as a 'fixed-entry' color in the
: palette instead of using this option.
: Default is '0'.
style: <mode> : Style controls the converted format of
: images. In 'palette' mode, converted
: images map 1-1 where each pixel byte
: represents a palette index.
: In 'rlet' mode, the transparent color
: index is used to compress runs of
: transparent colors, essentially reducing
: the output size if there are many
: transparent pixels.
: In 'direct' mode, the color format
: defined by the 'color-format' option
: converts the input image colors to the
: desired format. This may prevent some
: palette-specific options from being used.
: Default is 'palette'.
color-format: <format> : In direct style mode, sets the colorspace
: for the converted pixels. The available
: options are 'rgb565', 'bgr565', 'rgb888',
: 'bgr888', and 'rgb1555'.
: Default is 'rgb565'.
compress: <mode> : Images can be optionally compressed after
: conversion using modes 'zx0' or 'zx7'.
: The 'zx7' compression time is much faster,
: however 'zx0' usually has better results.
width-and-height: <bool> : Optionally control if the width and
: height should be placed in the converted
: image; the first two bytes respectively.
: Default is 'true'
flip-x: <bool> : Flip input images vertically across the
: x-axis
flip-y: <bool> : Flip input images horizontally across the
: y-axis
rotate: <degrees> : Rotate input images 0, 90, 180, 270 degrees
bpp: <bits-per-pixel> : Map input pixels to number of output bits.
: Available options are 1, 2, 4, 8.
: For example, a value of 2 means that 1 pixel
: in the input image maps to 2 bits of output.
: The palette option 'max-entries' should be
: used to limit the palette size for this
: option to ensure correct quantization.
: Default is '8'.
omit-indices: [<list>] : Omits the specified palette indices
: from the converted image. May be useful
: by a custom drawing routine. A comma
: separated list in [] should be provided.
--------------------------------------------------------------------------------
outputs:
The YAML file outputs section is a list of different "groups" of outputs,
differing in types or images. The currently available formats are:
- type: c : C source and header files
- type: asm : Assembly and include files
- type: bin : Binary and listing files
- type: basic : Single string formatted file
- type: appvar : TI archived AppVar format
The different options available for outputs are:
include-file: <file> : Output file containing all the graphics
: Defaults to "gfx.h", "gfx.inc",
: "bin.txt", and "basic.txt" for C, Asm,
: Binary, and BASIC formats respectively.
palettes: : A list of generated palettes, separated
: by a newline and indented with the '-'
: character.
converts: : A list of convert sections, separated
: by a newline and indented with the '-'
: character.
directory: <directory> : Supply the name of the output directory.
: Defaults to the current directory.
prepend-palette-sizes: : Prepend the palette with a 2-byte
<bool> : entry containing the total size in
: bytes of the palette.
: Default is 'false'.
const: <bool> : Only applicable to C outputs.
: Adds the 'const' parameter to output types.
: Default is 'false'.
AppVars are a special type of output and require a few more options.
The below options are only available for AppVars, however the above
options can also be used.
name: <appvar name> : Name of AppVar, maximum 8 characters.
: Required parameter.
source-format: <format> : Source files to create to access
: image and palette data. format can be
: 'c', 'ice', 'asm', or 'none'.
: Default is 'none'.
source-init: <bool> : Whether to output AppVar initialization
: code for loading and setting up image
: and palette pointers. Default is 'true'
: Only compatible with source-format: 'c'.
lut-entries: <bool> : Embeddeds a lookup table (LUT) for
: determining the image offsets in the
: appvar. Format consists of an entry
: containing the table size, followed by
: the LUT entries. Entry sizes are
: controlled via the parameter
: 'lut-entry-size'. Default is 'false'.
lut-entry-size: <int> : Controls the size of LUT entries. Can be
: '2' for 2-byte or '3' for 3-byte entries.
: Default is '3'.
compress: <mode> : Compress AppVar data.
: Available 'mode's: 'zx0', 'zx7'.
: The AppVar then needs to be decompressed
: to access image and palette data.
: Optional parameter.
header-string: <string> : Prepends <string> to the start of the
: AppVar's data.
: Use double quotes to properly interpret
: escape sequences.
archived: <bool> : 'true' makes the AppVar archived, while
: 'false' leaves it unarchived.
: Default is 'false'.
--------------------------------------------------------------------------------
Credits:
(c) 2017-2024 by Matt "MateoConLechuga" Waltz.
This program utilizes the following libraries:
libimagequant: (c) 2009-2022 by Kornel Lesiński.
libyaml: (c) 2006-2022 by Ingy döt Net & Kirill Simonov.
stb: (c) 2017 by Sean Barrett.
zx0,zx7: (c) 2012-2022 by Einar Saukas.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent
