AuthorTopic: [WIP] Selout challenge from PJ  (Read 26095 times)

Offline bengo

  • 0011
  • **
  • Posts: 597
  • Karma: +0/-0
    • https://pixeljoint.com/p/5787.htm
    • View Profile

Re: [WIP] Selout challenge from PJ

Reply #20 on: October 21, 2007, 02:17:59 am
I was wanting to edit your piece, do you have one that doesn't have a second light source on it?

Offline eghost

  • 0001
  • *
  • Posts: 60
  • Karma: +0/-0
  • Vagrant
    • View Profile
    • My Pixel Joint profile page

Re: [WIP] Selout challenge from PJ

Reply #21 on: October 21, 2007, 02:20:29 am
Unfortunately I don't  :(...Though feel free to edit if you want...
[EDIT] Here's the secondary lighting removed...[/EDIT]
« Last Edit: October 21, 2007, 08:39:33 pm by eghost »

Offline pkmays

  • 0010
  • *
  • Posts: 336
  • Karma: +0/-0
  • dawn patrole mole
    • View Profile

Re: [WIP] Selout challenge from PJ

Reply #22 on: October 21, 2007, 09:09:09 pm
Given that this has stirred up a fairly serious hornet's nest I'm not sure that I should be posting any continuance of this piece here, in spite of the fact that if I want to actually learn something when all of this is said and done this is the best place for it. Considering it's basic premise and my desire not to have to wade through a dozen or so posts bashing the "technique" that is the basis of the challenge to glean the kernels of actual critique and comments on the piece itself...I do appreciate the comments and criticism that the piece has received but I don't see how the current direction of the thread is going to help the piece come to a place that isn't going to offend anyone's sensibilities...
 

Don’t worry about it, these detailed theoretical debates are the reason I, and I’m sure many of the older members joined the original Pixelation years ago. Most people come here to improve their own skills, and threads like this are invaluable. Sorry the direction has been so dominantly theoretical, but you just happen to be the right (or wrong) person in the right place at the right time. After Helm and Baccaman started harping in on the issue I felt this was a good a thread as any to address my thoughts on the subject. I think it’s definitely time for me to make some official corrections and addendums to my old tutorial.


...But, the nature of the beast is that you cannot have an outline color solution for every b/g type so there this sometimes this fails...

...I've often been a little purturbed by the citing of SF sprites to be criticized so harshly as to me they were/are certainly at the pinnacle of the pixel mountain (along with the Metal Slugs and Zeldas...) but I know that's just personal taste and asthetics differ from individual to individual...

Yup, aesthetics do play a huge roll in what we consider the masterworks of game art. We are a product of our environments. I’m really not sure if I’d even have any interest in pixel art if it weren’t for fighters. So I can say that I am definitely biased to the late 90’s arcade fighter type of pixel art. Nor do I see any problems with that. We all have our own shticks, if we didn't, none of us would have any passion for our craft.

That said, there are things I’ve always disliked about the Alpha graphics. The anime style is increasingly not my thing, the color ramps aren’t as imaginative as they could be, the AA on skin is almost always harsher than it should be. A million things I can nitpick, because it's not my own creation. Still, the games are invaluable as reference material. I've learned so much from their treatment on edges and lines, and they are prime examples of efficient animation and well constructed (though still highly stylized) anatomy.

Yes that's what I mean because that's what selout *is*. It's a term you and tsu invented that does a very specific thing. It's been difficult to kill it as a term all these years and I'm stil fighting the good fight and now pixeljoint has a 'selout contest', good god.
selout is different from the others slightly in that it requires tedious application.  don't confuse work with thought, or pattern with description!  often, treating the line properly and attractively is easier than selout which follows a bizarre sort of formula that people end up second-guessing (as well they should!).  No, selout is no better than pillow shading in that it is complicated, ugly, and employs no visual reasoning.

I can’t speak for what Tsu or St0ven or any of the others were thinking at the time, but yeah, that’s how I saw it. I said to myself, “Well, Capcom is outlining in a lot of places, but in other places they are breaking the outlines. So they must be making black or really dark outlines and selectively going in and breaking them for this effect.” It didn’t even occur to me that they were also anti-aliasing and choosing the value of shades based on a specific background value. Chock it up to my lack experience and the initial excitement of when we started to "get it." I started pixeling in the summer of 2001 and wrote that tutorial in one day, less than a year after starting. I'm either going to have to completely revamp it or get Zoggles to take it down completely.

…The actual technique used in various videogames is both a byproduct of just shading a thing and is part of a larger skillset of rendering that cannot be seen outside that skillset's context, nor can it be honed as a disparate 'technique' like dithering or manual antialiasing…

Sure, and if I where to write some new tutorials, I would treat it as an end of book addition to a skillset that should be well established before even attempting it. A good primer in form, color, movement, AA, and the issues of sprite and backgrounds interaction would be prerequisites. Only then would I talk about ways of lightening outlines and anti-aliasing into a neutral background.
But yes, I agree. Selout doesn’t need to exist as a term or technique. It’s a misnomer for a technique that doesn’t work. It needs to die altogether.



Alright, seeing how this was originally supposed to be a critique thread, I’ll try to actually address the sprite in question.

At this point I'm at something of a loss as to where to direct the piece next, and torn between scrapping everything save the torso, sword/spear, and the basic shape of the tail and starting over OR tossing the whole thing into my scrap bin and starting a piece using what I actually understand...

I wouldn’t scrap it, you’ve got a nice foundation. The face and hair are especially strong. Some anatomy and readability issues exist, but nothing that can’t be fixed.



The main changes where giving her some hips and reducing the tangents. There were too many lines and curves that where touching but not intersecting, which made it hard to tell what space her body parts were inhabiting and what direction they were pointing. I lost some of the elegance in the curve from her hips to where her snake belly hits the ground, so you might try a different approach. What’s important is showing that her body recedes to the right and then back to the left, and finally back to the right. I’d also recommend you make the greenish tan of the belly visible when it bends back to the left to keep it grounded.



And then I tried to work the torso area for an example of how I would go about pixelating it. The skin especially needed some work.

The first and most important change when beginning was getting a medium gray background in so I can tell how dark the AA on the edges should go. I added a buffer shade to the skin to smooth things out as well. I completely avoided using the dark blue in the skin as it's just too big a jump in value to integrate attractively.

I also had to raise her left arm as it was too long below the elbow. I believe adding scales will do a world of good for the sprite. They can probably be rendered with the darker skin shades, or by adding a few dark blues, as I replaced your blue with a teal so it would buffer into the green of the skin.

Other than that, I’d say eliminate all the dark pixels on the sword and just block in the shape and form with color. I wouldn’t even outline the blade to keep it looking as sharp as possible.

Offline bengo

  • 0011
  • **
  • Posts: 597
  • Karma: +0/-0
    • https://pixeljoint.com/p/5787.htm
    • View Profile

Re: [WIP] Selout challenge from PJ

Reply #23 on: October 21, 2007, 09:33:11 pm
PKmays did a better job of an edit, so I guess I don't have to do any now. The only thing I've noticed is that I think the blade should be longer, or to save yourself some work you could make it a spear. Another thing to keep in mind is that anything not man-made, especially living creatures, are very uh, loose in terms of shape, so try to not make the lines so stiff, like with her tail for instance, there's a lot of straight vertical and diagonal lines, so yeah.

Offline eghost

  • 0001
  • *
  • Posts: 60
  • Karma: +0/-0
  • Vagrant
    • View Profile
    • My Pixel Joint profile page

Re: [WIP] Selout challenge from PJ

Reply #24 on: October 21, 2007, 11:15:10 pm
Thank you both very much...
The edit definitely cleared up some things that were giving me fits...I'll see about tossing a progrss piece up as soon as I can get the suggestions implemented...:)
Thanks again...:D

Offline Scuba Steve

  • 0010
  • *
  • Posts: 151
  • Karma: +0/-0
    • View Profile
    • Incredible Baron

Re: [WIP] Selout challenge from PJ

Reply #25 on: October 22, 2007, 01:01:57 am
I know I've used the 'selout' technique in the past to decent effect.  I know it was mentioned... but if your characters are almost always going to be viewed on dark backgrounds, they provide a great faux anti-aliasing effect.  I did a lot of Doom illustrating back in the day and you were almost guaranteed to always have darkness around you... so you could draw characters or weapons with a 'selout' style and it would be effective in blending in with the background.  And yes, if you happened to come across lighter squares and they crossed... it is not attractive... but it was effective for the other 90% of the cases they interacted.  I would say that, depending on what you are trying to accomplish, it can be effective... but as numerous people have iterated, you have to know what it's intended effects are and how it can certainly be detracting and ugly.
Glub Glub!

Offline eghost

  • 0001
  • *
  • Posts: 60
  • Karma: +0/-0
  • Vagrant
    • View Profile
    • My Pixel Joint profile page

Re: [WIP] Selout challenge from PJ

Reply #26 on: October 22, 2007, 06:57:03 am
Ok...Fairly heavy edit mostly working in pkmay's suggested lines as the tail lines he put up make a lot more sense by way of readability...I'm still tweaking the tail to a point that I like and trying to AA to 50% grey...
Also took the suggested tonal changes from the color edit as well...It still needs a lot of work but I'm definitely feeling better about the direction it's heading in...Thanks again y'all...


Offline Indigo

  • Administrator
  • 0011
  • *
  • Posts: 944
  • Karma: +0/-0
  • Artist, Indie Game Dev
    • DanFessler
    • DanFessler
    • http://pixeljoint.com/p/849.htm
    • DanFessler
    • DanFessler
    • View Profile
    • Portfolio

Re: [WIP] Selout challenge from PJ

Reply #27 on: October 22, 2007, 04:04:33 pm
you know what's odd about the definition of selout?  The term is slowly shifting to redefine itself as "colored outlines", but helm and others insist keeping it in it's negative form.  Often times tutorials teach of the "good selout" - as defined by colored outlines.  Many people also refer to Kenneth's work for Selout.

example:
http://www.derekyu.com/extras/pixel01.html

If it wants to shift to something good, then why stop it?
« Last Edit: October 22, 2007, 04:06:58 pm by Indigo »

Offline ndchristie

  • 0100
  • ***
  • Posts: 2426
  • Karma: +2/-0
    • View Profile

Re: [WIP] Selout challenge from PJ

Reply #28 on: October 22, 2007, 04:27:59 pm
that tutorial has most of the problems of the rest because it still separates line from fill instead of treating everything as shape.  we don't see line and fill, we see shape's that form edges, and shapes that form regions! the only way to produce good work is by dealing with shape!  it's also considering dithering a mandatory step 0.o

for some more proof that selout is tainted, just look at what turned up for this challenge.  half as many pieces, half as good, twice the bitching:

http://pixeljoint.com/pixels/poll.asp?id=1058
http://pixeljoint.com/2007/10/22/2354/Challenge_Voting-_Selective_Outlining.htm

even mash's entry is only subpar compared to his/her other work, and most of the better users didn't even touch it.

you might be abel to make an argument for it, true, but when a community that really enjoys video game art just drops it like that, I get a pretty clear message.
A mistake is a mistake.
The same mistake twice is a bad habit.
The same mistake three or more times is a motif.

Offline Indigo

  • Administrator
  • 0011
  • *
  • Posts: 944
  • Karma: +0/-0
  • Artist, Indie Game Dev
    • DanFessler
    • DanFessler
    • http://pixeljoint.com/p/849.htm
    • DanFessler
    • DanFessler
    • View Profile
    • Portfolio

Re: [WIP] Selout challenge from PJ

Reply #29 on: October 22, 2007, 04:56:27 pm
sure, but don't throw out colored outlines all together.  heck even my grunt uses it, but doesn't disregard form. (http://www.spriteart.com/indigo/images/art/pixelart/grunt_done.gif)