Inspired by ponysay, I think wicked ascii/ansi artwork on the terminal is great.
I decided to survey all the tools I could find that aid in this conversion to see if there were any dramatic differences in results.
For these tests I used an image with a 160px width, twice that of
a standard terminal. Then I
cat'd the image in plain xterm and
took a screenshot of the results.
The original has been scaled up (6X) to be the same relative size as the resulting screenshots.
My entire methodology is on github
if you wish to see exactly how I made these images. In theory
it is 100% reproducible from
make. (assuming on a linux desktop)
img2xterm stands out to me as the most accurate and true to the original, with util-say as a close second. Both of these tools understand “half-block” characters with two colors, effectively doubling the horizontal resolution of the resulting characters. (two colors per “pixel”)