AuthorTopic: fruity shading experiments  (Read 17365 times)

Offline Ai

  • 0100
  • ***
  • Posts: 1057
  • Karma: +2/-0
  • finti
    • http://pixeljoint.com/pixels/profile.asp?id=1996
    • finticemo
    • View Profile

fruity shading experiments

on: June 27, 2012, 04:11:19 am
I decided I needed to be more disciplined in practicing art skills.. so, I drew a series of fruit and am currently pixeling them (64x64 images); Don't usually post my stuff here, but this looks like it could go for a few more weeks at least, so hit me with any crits you have :)

Basis sketch image (sketched from right to left, not the converse, BTW):
(full size image here)

My aim was... to -effectively- use as many colors as possible. (aside from improving skills with planes).
I have previously experimented with 'iterative/recursive shading' and I'm beginning to get it to work well here.. Also motivated by a desire to find some of the techniques Henk Nieborg has used that make it possible to churn out many sprites with high shading detail, really fast; I suspect really strong understanding and application of planes is a key element.

Here's an earlier stage (just to dispel any suspicions that I CGed anything, heh):



I failed in various degrees on the first two: the cumquat and lemon both needed roughish texture, which confused my attempts to expand the ramps used. Putting that in the too-hard basket for now, I went to an easier attempt, a smooth skinned plum. Latest stage:



Somewhat complicated by the palette restriction I was working with -- 64 rgb intensity levels, ie. VGA palette restrictions. I ended up dithering cause I couldn't insert any more distinguishable colors, in the dark end of the ramp.
Also doing some hue mixing, due to the VGA palette restrictions, too.

I'm feeling happier about this technique now. Back when I remade the graphics for Bub-N-Bros, I used this recursive/iterative shading technique.. bit less happy with that outcome, even though I did like it, it seemed messy at times.

I think I cracked that, though.. one technique I used a LOT in this is to view everything as diagonal gradients. This makes it easier to avoid lumpiness, and rounds things well.. (well, I think the plum is a little too rounded/smoothed, haha. OTOH, it has practically no banding/lumpiness.)..
The other technique for unlumping is, if you must stack strips of color, stagger them in a direction that matches the planes of the object you're representing.
Oh, and make sure color regions taper smoothly throughout the image -- the other two techniques help this, but it's still challenging to me.

One thing that's pretty time consuming is redoing all the tapering when I add another color to the ramp. This isn't too bad as long as the underlying clusters are tightly structured; then it's quick to find the corners.

I'll also be setting up an auto history-saver as I proceed further in this.. I'll make a GIF or APNG when I finish.
Think I'll go for something more textured, like the peach, next.

Stats: cumquat = 15col, lemon = 12col, plum = 44 :), cherries = 24. Time? 7 hours so far..?

EDIT: Nah, went with the cherries. (also, fixed up the depth of the plum.)



</text> :)
« Last Edit: June 27, 2012, 07:57:03 am by Ai »
If you insist on being pessimistic about your own abilities, consider also being pessimistic about the accuracy of that pessimistic judgement.

Offline Mr. Fahrenheit

  • 0010
  • *
  • Posts: 326
  • Karma: +0/-0
    • View Profile

Re: fruity shading experiments

Reply #1 on: June 27, 2012, 01:26:39 pm
I think the plum has an odd shape on your drawing, not too reminiscent of your sketch. Also for the lemon, the pores/dimples are not visible unless you really try to look. Maybe you could darken the dimples up a small amount. Besides that I think they look incredibly realistic

Offline Bissle

  • 0010
  • *
  • Posts: 117
  • Karma: +0/-0
    • View Profile

Re: fruity shading experiments

Reply #2 on: June 27, 2012, 02:02:47 pm
All around wonderfully rendered; very realistic looking. The area on top where the stalk would be on the plum has a very odd shape; it looks bruised. The shape of the plum in general seems rather squished. Also, I've been looking at the reference pic and pics of real plums, and the upper right side has shading that implies multiple grooves, more like a squash or a pumpkin than a plum, which might at most have one groove along the center.

The lemon looks good, though I agree with Mr. Fahrenheit that the texture is not as discernible as it could be due to the coloring of the pores. The kumquat looks great, as do the cherries, although the shape could use a little work, too.

This is a pretty great exercise; never thought about applying the classic fruit study to pixel art!

Offline Ai

  • 0100
  • ***
  • Posts: 1057
  • Karma: +2/-0
  • finti
    • http://pixeljoint.com/pixels/profile.asp?id=1996
    • finticemo
    • View Profile

Re: fruity shading experiments

Reply #3 on: June 27, 2012, 02:45:50 pm
@MrFahrenheit
Thanks for actually replying :) While the plum sketch was pretty rough, I agree that the pixel shape was a bit off.
How about this?



(I also had a go at improving the lemon's texturing.)

@ Bissle:
Thanks :)
I can't even go without reference photos for something I know well, apparently :).
Haven't checked a reference yet, hope this is an improvement:

(.. is newer than the image immediately above this)

The shape of which -- the cherries? -- is off? I felt the shaping of the left cherry was wrong, but I haven't figured out what is correct yet. Is that the particular one you mean, or is another one also involved?
Is this edit

an improvement?
« Last Edit: June 27, 2012, 02:56:50 pm by Ai »
If you insist on being pessimistic about your own abilities, consider also being pessimistic about the accuracy of that pessimistic judgement.

Offline Bissle

  • 0010
  • *
  • Posts: 117
  • Karma: +0/-0
    • View Profile

Re: fruity shading experiments

Reply #4 on: June 28, 2012, 12:11:00 pm
The plum is much better. It's shaped more like the sketch and real plums. The shading towards the bottom center looks like something is sticking out inside of curving inward.

Sorry for not being clear; I was referring to the cherries, though looking at them now, I don't see whatever I saw before. The difference in the lemon is not very noticeable; maybe add more contrast.

Keep it up!  :y:

Offline Ai

  • 0100
  • ***
  • Posts: 1057
  • Karma: +2/-0
  • finti
    • http://pixeljoint.com/pixels/profile.asp?id=1996
    • finticemo
    • View Profile

Re: fruity shading experiments

Reply #5 on: June 28, 2012, 12:55:07 pm
re: "sticking out instead of curving inwards"
Google images seems very unhelpful in this instance. The plums on the tree in my back yard have always had this kind of bump (as opposed to the small bump or small dent found in shopbought plums or Satsumas). When I think about it, I agree that it is more rounded than my rendering is (though sometimes even more pronounced than I've portrayed. it usually curls somewhat, which I erred on.. my rendering is too pointed.) I'll get to work on that ASAP (probably tomorrow; am currently moving and my computer just got moved, need to adjust the furniture slightly and set the computer up in the new location.)

Google images has given me some good coloration/texture ideas otherwise, though (linked image is of similar type of plum to what I'm used to.. not quite, unfortunately)

Making the texturing more obvious than that on the lemon is going to be a challenge. But it is a challenge I will accept :)
If you insist on being pessimistic about your own abilities, consider also being pessimistic about the accuracy of that pessimistic judgement.

Offline Tourist

  • 0010
  • *
  • Posts: 376
  • Karma: +1/-0
    • View Profile

Re: fruity shading experiments

Reply #6 on: June 29, 2012, 03:26:28 pm
I think the lemon suffers a bit because but lacks a core shadow.  This is hidden by using a pure black background.  The latest version of the lemon has a bit of a shadow, so this is headed in the right direction, but it's not quite there yet.  It's not a problem on the orange because of the difference in light position.

I am interested in the technique you mention.  Do you start with the lightest and darkest points and then simply build a dithered gradient between them?

Tourist

Offline Ai

  • 0100
  • ***
  • Posts: 1057
  • Karma: +2/-0
  • finti
    • http://pixeljoint.com/pixels/profile.asp?id=1996
    • finticemo
    • View Profile

Re: fruity shading experiments

Reply #7 on: June 30, 2012, 04:34:05 am
I think the lemon suffers a bit because but lacks a core shadow.  This is hidden by using a pure black background.  The latest version of the lemon has a bit of a shadow, so this is headed in the right direction, but it's not quite there yet.  It's not a problem on the orange because of the difference in light position.

I am interested in the technique you mention.  Do you start with the lightest and darkest points and then simply build a dithered gradient between them?

Tourist

A core shadow? I've never heard of this concept. Is it a shadow that, in conjunction with the outline,defines the 3d shape of the object?

There are no oranges yet, only cumquats :) It's distinct due to size (much exaggerated here -- cumquats are usually a bit bigger than a gumball.) and oiliness (because the pores are smaller in a cumquat, the shiny points are closer together. This gives an impression of greater oiliness whether or not there's actually more oil.). Oh and taste -- but that's not depicted here;)

The technique is more like.. do the basic shading (3 shades. probably.). then subdivide (insert another shade) until there are more or less no pixels in any given 9x9 sample area that are the same color. Only exceptions being for areas that are flat or virtually flat.
Also, I use a rule that if the new shade doesn't give an impression of increased detail, it is wrongly done.
Dithering doesn't exactly come into that.. I avoid using it whenever texture doesn't demand it. some minor shaped/creative dithering is often a consequence of applying the aforementioned rule though.

BTW, it looks like I won't have solid computer access until 16 july or thereabouts.
I'll keep making updates as I can.
If you insist on being pessimistic about your own abilities, consider also being pessimistic about the accuracy of that pessimistic judgement.

Offline Bissle

  • 0010
  • *
  • Posts: 117
  • Karma: +0/-0
    • View Profile

Re: fruity shading experiments

Reply #8 on: June 30, 2012, 05:44:02 pm
Whenever you get more time, could you take a pic of the plums in your backyard? They sound interesting. Looking forward to more progress on this. :)

Offline Tourist

  • 0010
  • *
  • Posts: 376
  • Karma: +1/-0
    • View Profile

Re: fruity shading experiments

Reply #9 on: June 30, 2012, 06:47:46 pm

A core shadow? I've never heard of this concept. Is it a shadow that, in conjunction with the outline,defines the 3d shape of the object?

Yeah, that's it.  I used to have a very nice picture of an egg with the shadows and whatnot labeled, but I can't seem to find it at the moment.  So here's the same info in a long tutorial:

http://www.conceptart.org/wiki/doku.php?id=tutorials:value

with the most relevant image here (linked for size):

http://www.conceptart.org/wiki/lib/exe/fetch.php?w=&h=&cache=cache&media=sphere_labeled.jpg

Human vision works on contrast.  If you smooth away all the contrast, the result looks a bit smudgy.  Maybe try a fractal approach?  Where you have smooth shading within each light/medium/dark area, but still keep a larger jump between the different areas.  That might give you a vexel style look though, which isn't what you're going for.  Dunno, jut thinking off the cuff.  

I like the experiment and hope to see it continue.
Tourist

Edit: Rats, messed up the link tag on the tutorial.  Should be fixed now.
« Last Edit: July 01, 2012, 02:12:11 pm by Tourist »