Jul 28 2007

Vector Graphics

Published by §oL at 12:40 pm under Lucky Star, Trace

Zetia For Sale Buy Chitosan No Prescription Melatonin No Prescription Buy Online Plavix Buy Accupril Online Innopran Xl For Sale Buy Cla No Prescription Retin-a No Prescription Buy Online Endep Rumalaya For Sale Buy Aricept Online Buy Toprol XL No Prescription Aceon No Prescription Buy Online Trimox Buy Arava Online Omnicef No Prescription Green Tea For Sale Buy Elavil No Prescription Buy Online Rimonabant Buy Mevacor Online Trimox For Sale Stretchnil No Prescription Buy VPXL No Prescription Buy Online Zetia Buy Clarinex Online

After seeing Marmot and NegativeZero’s vector traces, I suddenly had the urge to do one. And no kidding, they are fun!

Vector graphics are unlike normal images in the sense that instead of using pixels, they use vectors in mathematical form to define lines. This makes them scalable to different resolution while maintaining its clarity, whereas bitmap images like JPEG and GIF lose their clarity when enlarged. Thats how i see it, correct me if i’m wrong.

Enough of the technical crap though. After around 6 hours, I finally finished it. My first vector trace lol.


Click link for 1024×768 version

Overall, i think it looks like shit LMAO… The finer details are not there, the colour’s a bit off, the eyes look weird and i can’t create the rosy cheek effect for nuts. NegativeZero’s Aaeru’s hair is incredible compared to mine.


Traced it from the original picture here

It took me 6 hours probably cause its my first time. Now that i’ve gotten familiar, its time to start on the other characters. I think i have too much time.

If you guys have any suggestions on how to improve this, please post it in the comments, thank you.

12 Responses to “Vector Graphics”

  1. Marmoton 28 Jul 2007 at 12:58 pm

    Hmmm, for the colors, it doesn’t seem like you used the dropper tool. Try #C65C6A for the skirt (#AC4151 for the shadow) and #F18A9B for the necktie (#CC7176 for the shadow). But I know that giving a hungry man a fish only feeds him for a day, so be sure to look for the dropper tool; it’ll let you lift the colors off of whatever part of the image you click on (ex: the skirt, the eyes, etc.). It looks like… well, a dropper, but here: http://courses.washington.edu/dmwork/images_W1L3/eye_dropper.jpg It’s the one right above the magnifying glass–hope it helps! >w<

  2. §oLon 28 Jul 2007 at 1:04 pm

    actaully i did use the dropper tool, but do you lift the colors off the original image?
    cause i used the original as a template, so its dimmed 50%.
    in the end the colour seems dull…
    or is it my imagination?

  3. Marmoton 28 Jul 2007 at 1:13 pm

    No, it’s not your imagination, but the image shouldn’t be dimmed when you’re getting the colors off of it. It’s got to be at full opacity for you to get the right colors. So yeah, you do get the colors off the original. ‘ 3 ‘

  4. §oLon 28 Jul 2007 at 1:18 pm

    oic… well, cause the image automatically gets dimmed for me.
    ok! now im gonna go remake this…

  5. NegativeZeroon 28 Jul 2007 at 2:04 pm

    Not bad for a first attempt. :)
    If you’re using illustrator, then it’s quite easy to get the eyedropper to work intuitively. First, instead of dimming the original art, select it and drop its opacity to 60-70%. That has the effect of dimming it and also will lighten dark areas and make your lineart easier to see.

    Then you double-click the eyedropper tool and it will bring up a menu with options. Tell it not to pick up opacity. Also for most stuff switch it to a 5×5 average, you get more accurate colour. This way you’ll get the actual original colour without ever playing with your original image again - you can lock its layer at that point to make sure you don’t accidentally miss it.

    The reason that my Aaeru’s hair looks good is a bit of trickery, actually. The lines are all full paths, so for every line, there’s actually a black shape describing it. This means that the lines don’t have regular width, giving it a more natural look. That in turn will make the lines seem to ‘flow’ more realistically.

    Attention to detail is really important. You need to make sure you get every bit of shadowing - there’s a few bits of shadow that you’ve missed, and they will go a long way towards improving the overall trace. Try brightening the green in her iris, it’s not nearly bright enough. Also the shadowing in her skirt needs to be darker. I imagine the eyedropper will help here. :)
    The final trick is a compositing trick. Open the vector in photoshop (gimp I imagine will work fine too). Put your background behind it. Duplicate the trace itself, and guassian blur to a radius of 4.0 pixels. Drop the opacity of the blurred layer to 30%. This will soften the whole trace and when you’ve used the right colours will give stuff like hair a shine to it that you can’t get normally.

    Also, thanks for reminding me I need to finish making that trace gallery. Doing it manually was giving me a headache, so I got halfway into automating it then stopped.

  6. §oLon 28 Jul 2007 at 2:17 pm

    loads of useful stuff, thanks, i really need it.
    i’ll post my corrected one after i finish it…

  7. §oLon 28 Jul 2007 at 3:25 pm

    hrmm, when i opened a 1280×800 new file,
    placed the vector tracing of konata inside,
    enlarged it and rasterize, the image appears jagged…
    any ideas?

  8. NegativeZeroon 28 Jul 2007 at 4:01 pm

    Rasterise the trace at high resolution (300dpi or better) and don’t resize it upwards once it’s rasterised. Make sure you’ve got anti-alisaing enabled too.

  9. §oLon 28 Jul 2007 at 4:05 pm

    umm, did all that, but it doesn’t make a diff.
    Nvm, turns out quite ok. check my latest post.

  10. Marmoton 29 Jul 2007 at 7:44 am

    (LISTEN TO NEGS, HE KNOWS INFINITELY MORE THAN ME)

  11. Titus Mcfaddenon 13 Nov 2008 at 8:37 am

    2e64svn22t8ua2tl

  12. Keiraon 26 Oct 2009 at 1:07 am

    Are all free templaes are poor quality, and the costly ones cost?

Trackback URI | Comments RSS

Leave a Reply