Provided by: libsixel-bin_1.10.3-3_amd64 
NAME
img2sixel - image converter to DEC SIXEL graphics
SYNOPSIS
img2sixel [ -options ] imagefiles
img2sixel [ -options ] < imagefiles
DESCRIPTION
img2sixel converts various images into high quality DEC SIXEL image format.
COMMAND-LINE OPTIONS
img2sixel has the following command-line options:
-o, --outfile
specify output file name (default:stdout).
-7, --7bit-mode
generate a sixel image for 7bit terminals or printers (default).
-8, --8bit-mode
generate a sixel image for 8bit terminals or printers.
-R, --gri-limit
limit arguments of DECGRI('!') to 255.
-p COLORS, --colors=COLORS
specify number of colors to reduce the image to (default=256).
-m FILE, --mapfile=FILE
transform image colors to match this set of colorsspecify map.
-e, --monochrome
output monochrome sixel image. this option assumes the terminal background color is
black.
-k, --insecure
allow to connect to SSL sites without certs (enabled only when configured with
--with-libcurl)
-i, --invert
assume the terminal background color is white. make sense only when -e option is
given.
-I, --high-color
output 15bpp sixel image
-u, --use-macro
use DECDMAC and DEVINVM sequences to optimize GIF animation rendering.
-n MACRONO, --macro-number=MACRONO
specify an number argument for DECDMAC and make terminal memorize SIXEL image. No
image is shown if this option is specified
-C COMPLEXIONSCORE, --complexion-score=COMPLEXIONSCORE
specify an number argument for the score of complexion correction. COMPLEXIONSCORE
must be 1 or more.
-g, --ignore-delay
render GIF animation without delay.
-S, --static
render animated GIF as a static image.
-d DIFFUSIONTYPE, --diffusion=DIFFUSIONTYPE
choose diffusion method which used with color reduction.
DIFFUSIONTYPE is one of them:
auto -> choose diffusion type automatically (default)
none -> do not diffuse
fs -> Floyd-Steinberg method
atkinson -> Bill Atkinson's method
jajuni -> Jarvis, Judice & Ninke
stucki -> Stucki's method
burkes -> Burkes' method
a_dither -> positionally stable arithmetic dither
x_dither -> positionally stable arithmetic xor based dither
-f FINDTYPE, --find-largest=FINDTYPE
choose method for finding the largest dimension of median cut boxes for splitting,
make sense only when -p option (color reduction) is specified.
FINDTYPE is one of them:
auto -> choose finding method automatically (default)
norm -> simply comparing the range in RGB space
lum -> transforming into luminosities before the comparison
-s SELECTTYPE, --select-color=SELECTTYPE
choose the method for selecting representative color from each median-cut box, make
sense only when -p option (color reduction) is specified.
SELECTTYPE is one of them:
auto -> choose selecting method automatically (default)
center -> choose the center of the box
average -> calculate the color average into the box
histogram -> similar with average but considers color histogram
-c REGION, --crop=REGION
crop source image to fit the specified geometry.
REGION should be formatted as '%dx%d+%d+%d'.
-w WIDTH, --width=WIDTH
resize image to specified width.
WIDTH is represented by the following syntax:
auto -> preserving aspect ratio (default)
<number>% -> scale width with given percentage
<number> -> scale width with pixel counts
<number>px -> scale width with pixel counts
-h HEIGHT, --height=HEIGHT
resize image to specified height.
HEIGHT is represented by the following syntax
auto -> preserving aspect ratio (default)
<number>% -> scale height with given percentage
<number> -> scale height with pixel counts
<number>px -> scale height with pixel counts
-r RESAMPLINGTYPE, --resampling=RESAMPLINGTYPE
choose resampling method used with -w or -h option (scaling).
RESAMPLINGTYPE is one of them:
nearest -> Nearest-Neighbor method
gaussian -> Gaussian filter
hanning -> Hanning filter
hamming -> Hamming filter
bilinear -> Bilinear filter (default)
welsh -> Welsh filter
bicubic -> Bicubic filter
lanczos2 -> Lanczos-2 filter
lanczos3 -> Lanczos-3 filter
lanczos4 -> Lanczos-4 filter
-q QUALITYMODE, --quality=QUALITYMODE
select quality of color quanlization.
QUALITYMODE is one of them:
auto -> decide quality mode automatically (default)
high -> high quality and low speed mode
low -> low quality and high speed mode
full -> quality and careful speed mode
-l LOOPMODE, --loop-control=LOOPMODE
select loop control mode for GIF animation.
auto -> honer the setting of GIF header (default)
force -> always enable loop
disable -> always disable loop
-t PALETTETYPE, --palette-type=PALETTETYPE
select palette color space type.
auto -> choose palette type automatically (default)
hls -> use HLS color space
rgb -> use RGB color space
-b BUILTINPALETTE, --builtin-palette=BUILTINPALETTE
select built-in palette type
xterm16 -> X default 16 color map
xterm256 -> X default 256 color map
vt340mono -> VT340 monochrome map
vt340color -> VT340 color map
-E ENCODEPOLICY, --encode-policy=ENCODEPOLICY
select encoding policy
auto -> choose encoding policy automatically (default)
fast -> encode as fast as possible
size -> encode to as small sixel sequence as possible
-B BGCOLOR, --bgcolor=BGCOLOR
specify background color
BGCOLOR is represented by the following syntax
#rgb
#rrggbb
#rrrgggbbb
#rrrrggggbbbb
rgb:r/g/b
rgb:rr/gg/bb
rgb:rrr/ggg/bbb
rgb:rrrr/gggg/bbbb
-P, --penetrate
penetrate GNU Screen using DCS pass-through sequence.
-D, --pipe-mode
[[deprecated]] read source images from stdin continuously.
-v, --verbose
show debugging info.
-V, --version
show version and license info.
-H, --help
print help.
ENVIRONMENT VARIABLES
img2sixel has the following command-line options:
SIXEL_BGCOLOR
specify background color.
overrided by -B(--bgcolor) option.
represented by the following syntax:
#rgb
#rrggbb
#rrrgggbbb
#rrrrggggbbbb
rgb:r/g/b
rgb:rr/gg/bb
rgb:rrr/ggg/bbb
rgb:rrrr/gggg/bbbb
SIXEL_NCOLORS
specify number of colors to reduce the image to (default=256).
overrided by -p(--colors) option.
Image loaders
img2sixel includes two or more image decoder components.
stb_image
libsixel includes stb_image, a public domain image loader.
img2sixel uses it as default built-in image decoder. It can decode almost all
images. but a few images can not be decoded by its limitations.
Supported source formats:
JPEG baseline & progressive (12 bpc/arithmetic not supported, same as stock IJG
lib)
PNG 1/2/4/8/16-bit-per-channel
TGA (not sure what subset, if a subset)
BMP non-1bpp, non-RLE
PSD (composited view only, no extra channels)
PIC (Softimage PIC)
PNM (PPM and PGM binary only)
Limitations:
no 12-bit-per-channel JPEG
no JPEGs with arithmetic coding / JPEG 2000
no 1-bit BMP
libpng
If libpng library is linked at compile time, img2sixel uses it for decoding PNG
image.
libjpeg
If libjpeg library is linked at compile time, img2sixel uses it for decoding JPEG
image.
gdk-pixbuf2
If gdk-pixbuf2 library is linked at compile time, img2sixel uses it automatically
in some cases.
GD
If GD library is linked at compile time, img2sixel uses it automatically in some
cases.
libsixel
img2sixel can load SIXEL as source image format, because it uses libsixel as a
SIXEL image decoder.
HISTORY
Former SIXEL encoders(such as ppmtosixel) are mainly designed for dot-matrix printers.
They minimize the amount of printer-head movement distance. But nowadays this method did
not represent the best performance for displaying sixel data on terminal emulators.
Encoded SIXEL data for VT-2xx/VT-3xx terminals were found in 80's Usenet, But the
technology of how to create them seems to be lost.
kmiya's sixel(kmiya,2014) introduces an efficient encoding method which is re-designed for
terminal emulators to optimize the overhead of transporting SIXEL with keeping
compatibility with former SIXEL terminal. Now libsixel and ImageMagick's sixel coder
follow it.
Araki Ken, known as the maintainer of mlterm, proposed the method for more compressed
SIXEL encoding. Now libsixel adopted that method. Araki Ken describes about the way to
generate high quality SIXEL.
See http://mlterm.sourceforge.net/libsixel.pdf(in Japanese).
SEE ALSO
sixel(5) sixel2png(1)
AUTHORS
img2sixel is maintained by Hayaki Saito. We imported whole code of stb_image v2.12,
written by Sean Barrett and its contributers, for loading various images, and imported
some code from pnmquant.c (netpbm library) for image quantization.
COPYRIGHT
Copyright (c) 2014-2016 Hayaki Saito
Permission is hereby granted, free of charge, to any person obtaining a copy of this
software and associated documentation files (the "Software"), to deal in the Software
without restriction, including without limitation the rights to use, copy, modify, merge,
publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons
to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or
substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE
FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN THE SOFTWARE.
THANKS
This software derives from the following implementations.
stb_image-v2.12
This software includes stb_image-v2.12 (stb_image.h), a public domain JPEG/PNG
reader.
https://github.com/nothings/stb
pnmquant.c (netpbm library)
The implementation of median cut algorithm for color quantization in quant.c is
imported from pnmcolormap included in netpbm library.
http://netpbm.sourceforge.net/
pnmcolormap was derived from ppmquant, originally by Jef Poskanzer.
Copyright (C) 1989, 1991 by Jef Poskanzer.
Copyright (C) 2001 by Bryan Henderson.
Permission to use, copy, modify, and distribute this software and its documentation
for any purpose and without fee is hereby granted, provided that the above copyright
notice appear in all copies and that both that copyright notice and this permission
notice appear in supporting documentation. This software is provided "as is" without
express or implied warranty.
sixel 2014-3-2
some part of converters/loader.c are derived from kmiya's "sixel" original version
(2014-3-2).
http://nanno.dip.jp/softlib/man/rlogin/sixel.tar.gz
It is written by kmiya@culti.
He distributes it under very permissive license which permits useing, copying,
modification, redistribution, and all other public activities without any
restrictions.
He declares this is compatible with MIT/BSD/GPL.
REFERENCES
resize.c (ImageMagick)
We added some resampling filters in reference to the line-up of filters of
MagickCore's resize.c.
http://www.imagemagick.org/api/MagickCore/resize_8c_source.html
CONTRIBUTORS
Araki Ken (@arakiken)
Markus Elfring (@elfring)
Akinori Hattori (@hattya)
isaki (@isaki68k)
NOKUBI Takatsugu (@knok)
Yasuhiro MATSUMOTO (@mattn)
Masami HIRATA(@msmhrt)
OBATA Akio (@obache)
Izumi Tsutsui (@tsutsui)
Iwamoto Kouichi (@ttdoda)
haru (@uobikiemukot)
Vertis Sidus (@vrtsds)
Bruce Mitchener (@waywardmonkeys)
Kazuhiro YOSHIKAWA (@yoshikaw)
Turenar <sora@turenar.xyz>
Yusuke Endoh <mame@ruby-lang.org>
mattn <mattn.jp@gmail.com>
Akinori Hattori <hattya@gentoo.org>
Øyvind Kolås <pippin@gimp.org>
Henri Salo (@fgeek)
hongxu (@HongxuChen)
pwd (@YourButterfly)
Nicholas Luedtke (@nluedtke)
cool-tomato (@cool-tomato)
BUGS
• Send bug-reports, fixes, enhancements to saitoha@me.com