jp2a(1) convert JPEG images to ASCII


jp2a [ options ] [ filename(s) | URL(s) ]


jp2a will convert JPEG images to ASCII characters. You can specify a mixture of files and URLs.


Read JPEG image from standard input
--background=light --background=dark
If you don't want to mess with --invert all the time, just use these instead. If you are using white characters on a black display, then use --background=dark, and vice versa.
-b --border
Frame output image in a border
Use the given characters when producing the output ASCII image. Default is " ...',;:clodxkO0KXNWM".
Use ANSI color for text output and CSS color for HTML output.
-d --debug
Print debugging information when using libcurl to download images from the net.
-f --term-fit
Use the largest dimension that makes the image fit in your terminal display.
Use terminal display height and calculate width based on image aspect ratio.
Use terminal display width and calculate height based on image aspect ratio.
-z --term-zoom
Use terminal display width and height.
When used with --html and --color, then color each output character's background color. For instance, if you want to use fill-output on a light background, do

jp2a --color --html --html-fill --background=light somefile.jpg --output=dark.html

To do the same on a light background:

jp2a --color --html --html-fill --background=dark somefile.jpg --output=light.html

The default is to have fill disabled.

-x --flipx
Flip output image horizontally
-y --flipy
Flip output image vertically
Set output height. If only --height is specified, then output width will be calculated according to the source images aspect ratio.
-h --help
Display a short help text
Converts image to grayscale when using --html or --colors.
Make ASCII output in strict XHTML 1.0, suitable for viewing with web browsers. This is useful with big output dimensions, and you want to check the result with a browser with small font.
Same as --fill. You should use that option instead.
Do not use bold text for HTML output.
Output only the image in HTML codes, leaving out the rest of the webpage, so you can construct your own.
Set fontsize when using --html output. Default is 4.
Set HTML output title.
Write ASCII output to given filename. To explicitly specify standard output, use --output=-.
-i --invert
Invert output image. If you view a picture with white background, but you are using a display with light characters on a dark background, you shoudl invert the image.
When converting from RGB to grayscale, use the given weights to calculate luminance. These three floating point values must add up to exactly 1.0. The default is red=0.2989, green=0.5866 and blue=0.1145.
Set output dimension.
-v --verbose
Print some verbose information to standard error when reading each JPEG image.
Set output width. If you only specify the width, the height will be calculated automatically.
-V --version
Print program version.
Sets output dimensions to your entire terminal window, disregarding source image aspect ratio.


jp2a returns 1 when errors are encountered, zero for no errors.


Convert and print imagefile.jpg using ASCII characters in 40 columns and 20 rows:

jp2a --size=40x20 imagefile.jpg

Download an image off the net, convert and print:


Output picture.jpg and picture2.jpg, each 80x25 characters, using the characters " ...ooxx@@" for output:

jp2a --size=80x25 --chars=" ...ooxx@@" picture.jpg picture2.jpg

Output image.jpg using 76 columns, height is automatically calculated from aspect ratio of image.jpg

cat image.jpg | jp2a --width=76 -

If you use jp2a together with ImageMagick's convert(1) then you can make good use of pipes, and have ImageMagick do all sorts of image conversions and effects on the source image. For example:

convert somefile.png jpg:- | jp2a - --width=80

Check out convert(1) options to see what you can do. Convert can handle almost any image format, so with this combination you can convert images in e.g. PDF or AVI files to ASCII.

Although the default build of jp2a includes automatic downloading of files specified by URLs, you can explicitly download them by using curl(1) or wget(1), for example:

curl -s | convert - jpg:- | jp2a -


If you have compiled jp2a with libcurl(3), you can download images by specifying URLs:

jp2a https://user:[email protected]/bar.jpg

The protocols recognized are ftp, ftps, file, http, https and tftp.

If you need more control of the downloading, you should use curl(1) or wget(1) and jp2a read the image from standard input.

jp2a uses pipe and fork to download images using libcurl (i.e., no exec or system calls) and therefore does not worry about malevolently formatted URLs.


You can extract the red channel by doing this:

jp2a somefile.jpg --red=1.0 --green=0.0 --blue=0.0

This will calculate luminance based on Y = R*1.0 + G*0.0 + B*0.0. The default values is to use Y = R*0.2989 + G*0.5866 + B*0.1145.


The latest version of jp2a and news is always available from


jp2a does not interpolate when resizing. If you want better quality, try using convert(1) and convert the source image to the exact output dimensions before using jp2a.

Another issue is that jp2a skips some X-pixels along each scanline. This gives a less precise output image, and will probably be corrected in future versions.


Christian Stigen Larsen <[email protected]> --

jp2a uses jpeglib to read JPEG files. jpeglib is made by The Independent JPEG Group (IJG), who have a page at


jp2a is distributed under the GNU General Public License v2.