Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Hatch
Pages: [1] 2 3

General Discussion / Re: Online Pixel Image Spec Tool
« on: March 15, 2010, 01:56:14 pm »
Nice, Hatch! It's fast!
Thanks! I have nothing to compare it to, so I'll take your word for it.

The Palette Sorter tool is worth mentioning here then.
Great tool! I was not aware of it.

It would be great if your Pixel Image Spec (PIS) tool would output a palette, too. How to sort it, though? You alreadyThis easily begins to feel like feature bloat, so it's just a suggestion. I get in trouble for feature creep all the time.
This is certainly possible, and a good idea, but I'd say it's a fairly low priority for now. Incidentally, that's an unfortunate acronym :<

-I fed PIS a web address by accident and got "Exception 450: no data returned `'" It ought to just say something less programmerish like, "Sorry, no valid image found" or something like that when there is no actual image specified

-If the color count exceeds 256, it's not an indexed image and therefore most likely not pixel art. Once 256 is exceeded the operation could either abort or PIS kicks into some other mode of operation. I see no reason for a crash.
Fixed! Although I made the limit much higher than 256 (5000, to be exact). I've tested it with 1000+ and it did fine. I disagree that being over 256 makes something not pixel art. Lots of large iso scenes, for example, will exceed that number. Anyhoo, this perhaps has uses beyond pixel art. It supports partial trans, for one thing. Not that something with partial trans automatically makes it not pixel art... but I digress.

-If you don't output a palette, imagine someone wanting to extract the palette for storage in their palette library or use for a project. Besides manually selecting each hex in their graphics program of choice, one by one, all they can do is screenshot PIS's "hex table". The hex values being printed right on the color swatches is a problem at this point. The space between each swatch is also a problem. So how about a toggle switch to hide all text and reduce all cell margins to 0, as maybe make each cell square as well. A toggle, or just another instance of the palette displayed per the previous criteria. Doubt you agree with all that, but you still see what I'm getting at here.
nodnod. I didn't really design it for people to extract palettes from images, but this potentially enhances usability and doesn't harm anything, so it's a good idea in my book.

-One big field for multiple pic addresses is rather odd. Why not multiple single line fields?
Flexibility. One can easily paste large lists if they want, and they can do as many as they want with no arbitrary limit of available fields. This does not harm usability much in my opinion and I vastly prefer it over multiple boxes. Will probably keep it this way unless I'm shouted down.

Thanks for your input!

General Discussion / Re: Online Pixel Image Spec Tool
« on: March 13, 2010, 12:50:18 am »
Hey, thanks for your input! Alpha transparency does work, but yes, that image has too many colors indeed. I'd actually intended to put a check on it to skip images with a crazy number for the sake of my cheap hosting plan, but I plain forgot :P
I'll do that as soon as I can. No more stress tests in the meantime please :)

Not sure how I'll display decimal values in an uncluttered way. I was gonna make a JS show/hide button for the hex, so maybe I can just have two.

General Discussion / Online Pixel Image Spec Tool
« on: March 12, 2010, 10:36:50 pm »
Hello! I made an online image specs dumper thing. It gives you all the specs of an image, like colors, width/height, frames, etc. Please use it if you find it helpful and let me know how I can improve it.

General Discussion / Re: Collabs are back at PixelJoint
« on: February 22, 2010, 07:19:31 pm »
Quote from: Gil
the current system looks insuficient to me

Yes, it is. That's the whole point of the post: the changes to the site are not forthcoming, so we had to develop these rules to work within our existing infrastructure.

Quote from: Gil
Quote from: Hatch
No collab accounts. We feel that these will ultimately create more problems than they solve.
I still don't get what the problems are... Collab accounts seem far more user-friendly.

No solution is going to be perfect while we lack the proper infrastructure. Collab accounts were rejected primarily for the same fairly obvious reasons that multiple accounts are ubiquitously prohibited everywhere: sock puppets, ballot stuffing, spamming, etc. It also seems hackish and impractical for every combination of two or more pixel artists to create a new account when they want to submit something. It puts a large moderation burden on us and creates opportunities for exploitation.

General Discussion / Collabs are back at PixelJoint
« on: February 22, 2010, 06:20:04 pm »
This is sort of old news at this point, but I realized that a lot of people here at pixelation probably visit PJ infrequently enough that they missed the post.

Collabs are back!
Don't go frantically grabbing for your calendar—it's not April 1st. Yes, after a long two-year wait, collabs will be allowed into the PixelJoint gallery again.

Why the sudden reversal after so long? Read on...

Collabs were originally disabled as a temporary solution while some changes were made to the site to properly accommodate them. Unfortunately, the changes haven't been able to be made as quickly as we'd have hoped. It now looks like they're still a long way off, and having reevaluated the situation, we've decided that a new solution is needed.

To minimize confusion and conflict, we've prepared a set of rules specifically for collab submissions. Please read over these points carefully:

  • No collab accounts. We feel that these will ultimately create more problems than they solve. Existing collab accounts will be kept and their pieces re-enabled, but we ask that the artists who own them no longer use them.
  • No multiple submissions. All artists involved must decide on a single one of them who will submit the piece. There must be unanimous agreement.
  • All artists involved must give their consent and be properly attributed. This is supremely important. If even a single artist retracts his/her consent at any time, the piece will be immediately sent back for revision. If there's any doubt as to the consent of all parties when the piece is submitted, it will be sent back.
  • Collaborators should be linked to. If they have a PixelJoint profile, link to it. If not, link to their profile on another site—wherever you think they're most active.
  • No big community collabs with many participants. We're relying on squishy common sense for this rather than a firm number
  • Who did what should be explained in the description as clearly as possible
  • No collabs in weekly challenges!

We want to hear your feedback, particularly regarding the rules and potential problems they don't address. Please let us know what you think.

So feel free to submit if you have any you've been holding back. And please let us know how we can improve the system as we have it.

General Discussion / Re: I need to rant!
« on: February 09, 2010, 10:21:50 pm »
I'm the New Guy on the PJ mod squad, and I just wanted to point out some things about the submission and approval process:

  • Art is highly subjective.
  • All of the rules regarding PixelJoint gallery submission are subjective for that reason.
  • When a mod decides whether a piece should be accepted into the gallery or not, they're making a subjective judgment call. Some other mods may disagree with them. Some users may will disagree with them. This is unavoidable due to points 1 and 2.
  • If we were completely objective and dogmatic and said, for instance, "ALL WHITE BACKGROUNDS WILL BE REJECTED, PERIOD", then we'd be required to force the creator of this beautiful piece to completely ruin it or take it elsewhere. Similar edge cases can be found for any of our rules.
  • If we didn't have this flexibility, we'd be branded "pixel fascists" even more frequently than we are already.
  • One of the tradeoffs for this flexibility is that we have many people highly critical of the moderation. This would be the case no matter who was on the mod team. There will always be users who disagree with some of the decisions in such a subjective process, and there will always be a highly vocal subset of them. In spite of this, it's a good tradeoff, in my opinion. I'll take a few lumps in order to avoid rigid dogmatism.

Just please understand that we have to make these judgment calls dozens of times a day. If each user looks out for his or her own interests and calls our attention to mistakes we may have made with their art, they're thus deputized as regulators of our moderation; the PJ Internal Affairs if you will. All we ask is that you treat us with respect and courtesy, and you'll be treated in kind.

Yes, this means calling us out on our mistakes on our own turf, please ;)

Pixel Art / Re: [WIP]Boo-Boo~~from YOGI BEAR!!
« on: August 15, 2009, 04:37:15 am »
I think everyone has misunderstood Scribblette. I believe he/she is saying that the thick black outlines convey the same sense of fat pixel banding--not that banding is a good technique that should be applied.

Pixel Art / Re: Tank
« on: April 03, 2008, 03:04:06 pm »
Conversely, it looks like the rounded plates on the turret have gotten very flat. I think another buffer shade might be in order. If the extra color bothers you, you could merge the brightest blue and the brightest gray (I would do this regardless).

Pixel Art / Re: [WIP] Avatar
« on: February 27, 2008, 04:47:15 am »
There is a ton more work to be done on the face before you start worrying about neck and shoulders.

The human face is an incredibly complex thing with scads of little muscles and details. You need to think more in terms of these little details. Here's a quick edit:

Very quick, and I'm not saying it's perfect (my strong speculars are pretty inappropriate :P), but let's go over a few things in no particular order, despite the fact that they're numbered:

1) Proportions. Human's don't generally have noses that short or upper lips that tall. Always consult references of what you're drawing; Google Images search is invaluable for this sort of study. Pick up a facial anatomy book if you want to get serious; these usually have wonderful proportion guidelines.

2) Nose. The human nose alone is an incredibly complex thing! It can't be simplified into a rectangle if you're trying to draw realistically. Think carefully about nostrils, not just the holes, but the way those passages shape the nose as a whole. Think more about how it blends smoothly in with the facial plane rather than being simply plugged in. Think about how it interacts and blends with the brow (something I didn't address in my edit but meant to)

3) Masses and levels of depth. When shading anything, think about breaking it down into basic shapes--primitives--and shade accordingly. This will help you define your masses clearly. Your shading around the chin and brow, for example, is very illogical. And like tocky was saying, use all your colors to the fullest! Use them to define clear levels of depth. When drawing a face, you can almost think of each shade as defining a new plane.

4) Proportions again. Remember that the human head is divided into two roughly equal halves of cranial mass and facial mass, with the the center being the eyes (technically the bottom of the brow). With this in mind, I think his forehead needs to be a tad taller (though, again, I didn't address this in my edit).

5) The little things. The depression under the nose that separates the upper lip into two halves (there's a technical name for it that I can never remember), wrinkles and folds (like those in the edit originating from the nose), cheekbones, jaw line, brow ridge, etc., etc. Again, an anatomy book can be your best friend, but even just looking at photos of faces is a great way to study.

Pixel Art / Re: [WIP]Dude - Animated, Isometric & Breathtaking.
« on: February 21, 2008, 03:32:21 pm »
Ho there, Pixelation! I'm going to try to get serious about critique.

My initial reaction to the swamp mockup is confusion. The ground tiles are very busy and distracting. Some things, like the logs and the green rock (?) are half submerged, while others, like the trees and the person, are fully on top, making it look like they're walking on water. If the other elements are meant to be half buried rather than submerged, well... That's not really how it reads, at least not to me, especially in the context of a swamp, which I think of more as water than mud.

Speaking of the logs. At first glance, they look inverted--like textured, log-shaped depressions in the ground. To me it looks like one of those 3D optical illusions where the image seems to 'pop' in and out if you stare at it long enough. But I may just be tired :P

Also, the trees seem really disjointed and somewhat inappropriate. I know leafy and evergreen trees can and do appear in swamps in real life, but of course the more typical appearance, the one most people have in their heads when they think "swamp", is that of dead, broken trees, like the one you have on the left. That's why they also think of rotting, half submerged logs--the artists formerly known as trees.

Lastly, the whole thing seems to lack cohesion. I'm sure you had a blank patch of ground tiles onto which you simply pasted various elements, and that's exactly what it looks like. I saw a great post by Ben2theEdge that I think is very applicable.

SO, in conclusion, what I would do is:

  • Simplify the ground tiles with fewer colors and less contrast.
  • Decide once and or all if it's mud or shallow water, and redraw everything accordingly. Or have patches of mud and patches of water, I guess.
  • Maybe revisit the shading on the submerged logs--I'm still not sure if it's just me being slightly nutty.
  • Ditch the leafy (deciduous?) and evergreen (coniferous?) trees in favor of more typical dead trees.
  • Think more in terms a full scene rather than individual elements.

Pages: [1] 2 3