IMG2SIXEL(1) User Commands IMG2SIXEL(1)

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>
/Oyvind Koloas <pippin@gimp.org>
Henri Salo (@fgeek)
hongxu (@HongxuChen)
pwd (@YourButterfly)
Nicholas Luedtke (@nluedtke)
cool-tomato (@cool-tomato)

BUGS

+o Send bug-reports, fixes, enhancements to saitoha@me.com


4th Berkeley Distribution Aug 2016 IMG2SIXEL(1)