I wanted to experiment with vectoring and decided to try what color reduction on a simple image would turn out like.
It seems I have a weird habit of using Shana images… oh well.
Anyways, I vectored all the major shapes, ignoring the white flower prints on her yukata/kimono (what’s the difference, again?). This way I could progressively reduce the number of colors used and made it easier for experimentation.
Note: All the white background images lead to transparent background versions. The black background ones just have solid black matte bg.
I started with 4 colors (not including black and white, obviously).
The shading is a bit funny.
The black background provides a little more impact.
Now I wanted to see what happened if I just flattened the colors to red.
I think it came out pretty good, and I don’t have to deal with shading oddities with so few colors (unless I wanted to do them in black).
I cheated a bit as I had to color Shana’s face – keeping it white just looked too weird.
Now I wanted to see what a red + white version would look like (not including the black in the background).
As expected, looks like a stencil. A bit too bright.
Still jarring on the eyes. Serviceable enough I guess.
Finally we come to the black and white version.
A bit easier on the eyes than with the red only. But boring.
I was too lazy to add extra white regions for the ribbons; didn’t think they’d really add much anyways.
I think the experiment was largely a failure. I liked the 4 color and red/black/white versions best (the original being superior regardless). Also, the image is rather boring since the kimono got simplified so much – maybe if I added more detail, or some kind of background…
Oh, well, I’ll end with another experiment I did (just for fun), where I colored the trace-lines randomly as I was vectoring.
Avant-garde art! lol.
P.S. will add 1280×1024 resolutions when I’m less inebriated.
Update: Screw 1280×1024. Crop it in paint or something. Also, seems the links to the 4×3 resolution ones were all broken (pointing to the widescreen instead). Fixed now.
first
http://img188.imageshack.us/img188/4134/shanaabstr.jpg
http://img188.imageshack.us/img188/4889/shanaabstr2.jpg
http://img188.imageshack.us/img188/939/shanaabstr2b.jpg
http://img190.imageshack.us/img190/2626/shanaabstr3.jpg
http://img190.imageshack.us/img190/357/shanaabstr3b.jpg
http://img190.imageshack.us/img190/2875/shanaabstr3c.jpg
second
Nice, looks like fun, and yea weird shadow (I remember this img ^^). Looks like she just jumped out of something and it exploded… Shana Explosion!
You’re right, too much loss of details. Not just the pattern on the kimono though, you can’t even differentiate between the obi and the kimono itself in your vectors because you’re limiting yourself to so few colours. Without shading, there’s no sense of depth at all, which takes a lot away from an image.
I like the final B&W one. Apart from losing some of the ribbon just above her head, I think it’s pretty good. Would probably look nice as a print on a shirt.
Interesting experiment. Agree with Lupus on the loss of details.
Your vector traces are way more refined than mine. Then again, I only invest less than 3hrs for most of my works. I need to implement shading gradients and better line-work. Time to start on more complicated images…
I got here from google images “whats the difference between a yukata and a kimono. I noticed the picture of Shana from shaukugan no shana and was like “hmmm… might as well check it out.”
Love you drawings of shana btw.