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.


Topics - Ambivorous
Pages: [1] 2

1
2D & 3D / Value Studies
« on: August 12, 2017, 11:19:02 am »
Preface

When I first started doing pixel art I thought that it wouldn’t require much artistic skill. This was obviously horribly incorrect.

The logic being that I can just use someone else’s palette and since I’m placing all the pixels individually it will be easier. And this was partially true.
Thing is, the workflow involved when you’re trying to make pixel art without any artistic knowledge is constant refinement. You draw one thing (and it sucks), then you go over every single detail in it again and again until it looks satisfying. You’re basically trying to brute force an image.
The end result of this is that it was taking me hours; days to finish a single, tiny piece. Now, over time I got better, but the fact of the matter remains I was doing a lot of unnecessary work, and reworking, which was wasting time and still the end results were not as great as I wanted them to be. On top of that if I wanted to change anything about a piece I was extremely reluctant, because of the amount of effort I had already put in. So my pieces were stale and static and sometimes just a massive waste of time. You can also imagine it was basically impossible to animate anything decent in this manner.
Then there comes the palette. Let’s face facts: if you can’t make your own palette (or at least modify someone else’s) you will never get the mood you quite want and you will have to settle.

Luckily for me I was a curious soul and I spotted one of the other pixel artists doing a thumbnail study over in the daily sketch.
I asked what they were doing and the response was that they were taking an image, making it rather small and then trying to copy it - more or less - while essentially blurring their vision so as not to distract themselves with the details. Well, this sounded like fun and I decided to try my hand at it.
Enter the worst thumbnail study in history:



Alright honestly it's not even that bad, and I put a lot of effort into this and even tried out a new brush!
But realistically I now know that I could not see value in any way, shape, or form while doing this image. This is when I was introduced to value studies!

A quick side note on the word 'value':
Basically value means how light or dark something is. Another word for this is brightness. But when I use the word value here in this context I am taking into account one final thing: saturation.
In your digital painting program of choice you will likely come across a transformation to greyscale which will include the word 'perceptual'. What this word means to you is that not only did they take into account the brightness (which is actually called value in your HSV (hue saturation value) colour picker), but also how saturation effects the apparent brightness to the human eye.
You can have yourself a quick google of the differences between all these terms, but the tl;dr of it is that I use value incorrectly, so just bear with me.


Some Theory Work

Alright, so what is a value study?
It's quite simple, all we're going to do is take an image, make it greyscale, and then try to copy it.
The idea here is that colour is distracting and we're only interested in value for now, so we're going to remove any and all distractions as best as possible to focus on what we want to.

But just doing this isn't going to be terribly useful if we don't set some nice rules for ourselves to learn what we intend to learn.

Spoilers: we want to learn how to see value. The entire point of value studies is to teach our brain to see value.
If you don't catch yourself staring at a cup of coffee in the morning looking at how dark the shadow underneath the mug is, and how brightly the light is shining off the top rim, then you're no where near ready to move on from value studies. We need to make your brain obsessed with value and see it all the time in everything we see, even when we're practically unconscious.


So to force ourselves to see value we need these rules:
  • Do not trace: Open up your reference (now greyscaled) image in one program (preferably on a separate screen entirely), and your art program of choice with a blank canvas alongside it.
  • Never use the eyedropper tool: This just defeats the entire purpose of your value study. This actually only applies to using the eyedropper tool on your reference. Using the eyedropper tool on colours you've already used on your study will save you time, but for bonus points selecting each colour (shade of grey) manually every single time you need it will give your art an amazing feel later on and you'll be able to transition to physical media much easier. Also you'll make amazing palettes.
  • Timebox yourself: Stick to an hour a day, per study. Having the time limit open ended will have negative effects like focusing too much on one detail, loss of inspiration, feeling like you're wasting your time, and eventually stopping doing these exercises at all, so make sure you only do this a little bit a day, so it remains fun and never becomes a burden.
  • Don't try too hard: You're going to suck at this at first, that's the whole point. Don't put in too much effort. The  idea here is to teach your brain to do this automatically, so the less conscious effort you put in the better. If you're struggling with something that means your brain isn't quite used to it yet. This is purely an indicator that you need more practice at that specific task, so look forward to things you struggle with because this will give you room to grow!
  • (Optional) Stick to a handful of shades at first: This is what I was told to do, but I don't know what value (hue hue) it actually added, so do it if you want, but I'm not evangelising it. Please do tell me if this is actually a useful step if you try it, so I can update this.
Alright, so now that we know what to do, what not to do and what our intention is, we need our first reference image.
Now this part has sapped a lot of my inspiration in the past. It's hard to find good reference images that ease you into things gently, and indeed finding good reference images is something you want to practice, so it's part of your value studies! If you find yourself exhausted after searching for a good reference image this means you've made your brain work already, so don't feel bad if you're hard pressed to now do your value study.
To take away this daunting task you can separate these two actions. These days I'm always on the lookout for good reference images, so every day I log into deviant art and check the top images of the day and if there are any amazing photo references I can steal. If I find myself google image searching something and a good reference image pops up I stash that away in my references folder. Always be on the lookout for good reference images even if they don't satisfy your current needs, because later on you might need them.

And now we're ready to begin our very first value study.

Your First Value Study

Here is my very first value study:



Honestly, not all that bad! Once I'd removed all this colour nonsense it was pretty easy to see that there are blacks and whites and a bunch of greys.
So let me go over a few of the things I did in this and then explain how useful or useless I think they were.

I made the canvas the exact same size as my reference image.
This was useful because I could judge the exact distances that things were from the edges of the image. But perhaps I should not have as this is another layer of distraction. If I had not cared about exact sizes and shapes and differences I may have absorbed the value knowledge sooner.
It would probably have been best to first do a composition study to get my alignments and sizes correct and then do value studies, but that is for another time.

I used only a handful of shades of grey.
This may have removed some of the distractions. I didn't have to worry about blending yet, so that was nice.
Again, perhaps blending should be moved to a separate study, or added later (as I did) once you're more comfortable with values.

I only used a round brush.
This was probably a mistake in hindsight.
Again again, removed a layer of distraction since I always used the same brush. I should probably have diversified my brush usage later on once I had everything else waxed, because now I can still only use one brush!

A side note on tools:
As I just mentioned, I still only use a round brush, with 100% hardness. Even once I'd started blending (I use opacity to blend).
What I did learn out of this is that the tools you choose to use are entirely irrelevant to your learning or your eventual ability! Pick the tool you want to use, because eventually you will become good enough with anything to achieve any end result you want.
This means use any program and brush and settings and method you want. Draw outlines, blobs, use photoshop, krita, round brushes, square brushes, draw with a trackpad, mouse, tablet; use microsoft paint if you want. And change what you're using at any point you want! Yes, your art will take a hit at first if you're changing, but that's part of it.
Don't become like me and be reliant on the same brush every time though (or do, I'm pretty cool).


So now I think you're ready to start doing your own value studies, equipped with the knowledge you'll need to become an amazing artist.
As a bonus, I redid the same image just eight value studies later and look at the progress I'd already made:



Now go and become the next [insert famous artist of your choice]!

2
Pixel Art / Textures on Forms
« on: February 29, 2016, 08:00:23 pm »
With the advent of the fox uprising and my recent dive into the world of planets (hur hur), not to mention the ever constant struggle of dealing with pixeling trees and character hair, I have decided that I'm going to put together a thread showing what I mean when I offer the critique of, "You need to define your forms more, because the texture looks like it's on a flat surface."

People often seem to be able to deal with forms when there is no texture, but as soon as we add the texture to our forms they lose all their previous shape. This is particularly noticeable with pretty much everyone's first pixel tree.
This is because they lose all shape of their forms when we use all the highlights and shadows of our object on each, individual piece of texture (each bang of hair, each bunch of leaves, each piece of fur, and indeed each feature on their planets gets the full range of value/brightness). Instead what we need to do is only use the amount of value available to us in any given region and just hint at texture, rather than draw out every single leaf.

So let's dissect this with a simple example.

Here I have a nice, round, beautifully orange ball (using Arne's 16 colour palette):



Beautiful. That is a ball and you can see it is a ball. This is a well defined form.
But we wish to add texture to our ball! In this case some fur (it is orange like a fox after all).



We begin by adding some edges between the light and dark colour. Here we heavily imply the shape of fur.
Notice how I am cutting into the dark area with the lighter colour. This changes our form slightly. It looks like it's being lit slightly differently than our initial ball, because there is now more average light going further around the ball.



We remedy this by adding some darker shade over some of the lighter area.
Again I've used obvious lines where there is fur. I put some shade on the first bits of fur texture we added to give that feeling of depth.
Notice we have changed our form again! We're striving to keep the average value of our transition area of light and dark the same as our original form.



We'll just scatter a little more darkness here and there to give more implied texture. We're not actually defining any more fur, and importantly we have left large, open areas entirely blank!
It doesn't matter though because on the edges between our light and dark we have heavily implied the existence of fur, so our brains will fill in the large areas with imaginary fur for us*.
More form changing.

Side note*: This is a very important part of pixel art. Hinting at details and letting our viewers' brains fill in the gaps rather than trying to use our very valuable, limited resources to draw every piece of fur (which often causes noise and confusion).

For a final step let's just add some fur around the edges for more implication of the textures.



And there you have it. With two simple colours I have created a readable, fur-covered ball.
Attentive readers will notice I have decreased the average value on the brighter side of our furball through all this, so honestly our ball is slightly less bright than it was originally, but we have kept our form and that is the important part.

A bonus byproduct of this process is that we were able to break up the solid, boring line between light and dark on the original ball using texture. This means we can, and should, use textures to add interesting details around edges when possible.



I shall continue to update this as I think of more things to add/change (or come up with better examples) and would like the opinions of other members with regards to the topic and this post.

3
Portfolios / [Available] Pixel Artist
« on: February 25, 2016, 02:33:26 pm »
I am currently only available for part time/short term work.

Greetings!

I am a pixel artist who primarily works on sprites, most specifically edits (new interpretations of another's pixel art) and replicating the style of other artists. I can also do animations, tiles and backgrounds if needed though.
Glossing over my work on PixelJoint and DeviantArt should give you a good idea of the quality of work I can put out.

Contact me via PM here or by email (geoffrey@ambivorous.com).

Portfolio Sites:
Other Work:
(ie. not shown on dA or PJ and mostly unfinished)

Practice Pieces/Personal Projects:





Notable Edits:

=>
for Tsej on a native american archer.
=>
for Azuyre on a witch.
=>
for Gentu on a forlorn knight.
=>
for Seiseki on planets.

4
Devlogs & Projects / SoaGA (2D, sidescrolling, action platformer) [WIP]
« on: February 20, 2016, 09:50:32 am »
Sort of a Grey Area (working title)
I am making a game. This will be my devblog.

Preview of gameplay:


Introduction:
I am a programmer (and becoming a pixel artist), but I have never put out a full game, and certainly never by myself. Carpe diem!

Inspiration:
Storyline - The Black and White series (but mostly the first one).
Art direction - Indigo's Chasm.
Fighting Mechanics - Mortal Kombat and any other fighting games I've played.
Platformer mechanics - Sonic the Hedgehog, Megaman, etc.

Tools:
Language - C#
Graphics Library - openTK
IDE - Visual Studio Community 2015
Pixel Art - Aseprite
Graphics Tablet - Wacom Intuos

Resources:
BitBucket Git Repository - I push regularly. After pretty much each new added (or fixed) feature.
Storyline Document - All the spoilers.

5
Pixel Art / Chibi Archery Chick Shooting
« on: February 02, 2016, 09:00:23 pm »
Making a shooting animation for a chibi-ish looking archery girl I made to blatantly and unashamedly copy the style of a sprite Conceit made, which I later found out was actually in the style of some Brazilian guy named Fe or something.



So that's the base.
Now I'm trying to animate.



Original.
Initial reception asked for more action during release.



Added terrible-technique hand flick after release. Her foot even moves for no reason.
I thought they'd be happy with this. I was wrong.

Current Advice for next animation:
TWICE AS MUCH SWOOSH.
Shaky hands while fully drawn.
Bow's arms swinging in front of the riser after release.
Add recoil.
Push hips out even more for less reasons.
More hair flow (not just during release).
Add foreshortening sweep to ponytail.



Forgot to mention: the reason for doubling the swoosh is because it's easier to tone down than up.
Here's the result:



I'm gonna move on to detailing now.

6
Pixel Art / Fairy in the Forest [WIP] [C&C]
« on: January 30, 2016, 10:16:53 pm »


This piece is a lot of things to me.
One is of course practice. I'm also trying to get my head around some styles and experimenting with a few things.
I do also actually have the intention of making this into a game, but it will be very basic and will likely take me ages. It would be more of an art demo than a game really.

I'm keeping Astraldata's composition in games thread in mind while creating this as well, so do keep an eye out to crit on that.

So I'm looking for crit on everything really. Some things I've done intentionally, but there are a lot of things I'm doing just wholly wrong.
It's still very early WIP, so it might be difficult to critique, but I always read all posts and welcome any ideas, so please add whatever you think might help.
Worst case I'll just carry on posting my updates with descriptions and hopefully that'll help people in the future.

7
Devlogs & Projects / Scanning (working title)
« on: January 30, 2016, 10:05:41 pm »
A friend and I are busy making a small game while I'm on leave. Just for fun and so we can both learn.
I've been focusing on doing the art and he's been handling the programming.

So I'll post the assets I've made. Everything is still in the placeholder stage of course, and I've basically just used Arne's 32 colour palette for everything, because it saves a lot of time not having to create your own palette and I can always change it all later once we've decided on more things.


The progression of the base



I started out designing what I thought a small hold would look like given today's technology on a foreign planet.



Soon after the initial art was put in I quickly realised 32x32 tiles just wouldn't look good, so we increased it to 48x48.
At the same time we decided on some of the driving force behind the mechanics. We went with stranded on a planet, so we'd start off with your spaceship that crashed and you're trying to get home.



Later realised we needed a place to spawn our drones and maintain them, so a depot was in order. Figured I'd flesh the base out to 4 tiles and have a progression be shown through the development of this base. The idea being eventually your have a small holding; a nice, fancy depot and a fully built rocket to take you home.


The slaves



Kept the drones at 32x32 themselves, but they sit on 48x48 tiles. Just to give them a sense of scale compared to the base and such.
More complicated drones will get bigger later, which is another reason I wanted to increase the size in the first place.


The game itself



So the details of the game are as yet undecided, but the gist of it is that you're stranded and you need some kind of resource and all you have are some small robots to scout the area and collect things for you, but you don't just get vision around your drones - they cannot be communicated with while they're not at base plugged into your laptops as it were.
The rest is us just trying to figure out how to keep this mechanic while making the game feel good, and honestly right now it feels pretty good, but there are plenty of other issues.


Anyways I'll keep you guys up to date, but do give suggestions if you have any.

8
I've long struggled to get into the PixelJoint weekly challenges. Sometimes they are too complex for me to attempt, sometimes I am disheartened by how well the other members are doing, and sometimes they just feel like a waste of time.
Not that I have complaints about the system itself though, and I know it encourages a nice plethora of subjects.

I would very much appreciate a more practical version of the system though. Something aimed at helping people who are newer to pixel art an opportunity to build up assets in an inspired manner.

The idea I had to solve this is as such:
We have weekly challenges, but on a far more practical/usable scale, so instead of having subjects that encourage more artistic -if you will- pieces we encourage the creation of sprites, tiles and scrolling backgrounds.
We could potentially have overarching themes, such as "creating a small mario clone" where the entries from several weekly challenges add up to create the assets needed for a complete game. Could even have a generic program written where you can just shove your art in and then play your own clone.
For those who don't, or can't, participate in every week's challenge the resulting assets could just be used to pad their portfolios, or just for the practice.

Further examples of challenges could be one week having something like, "Create a sprite of a character for a 2D, side-scrolling, platformer using either a 16x16, 32x32 or 64x64 canvas." and the following week, "Create an idle animation for a sprite created in the previous week's challenge (can use someone else's sprite if you did not participate)." with the idea of creating a fully animated (running, jumping, dying, etc.) sprite at the end of a certain number of weeks.

I feel like this kind of challenge will be more valuable to growing members as functional pixel artists and help transition them into the working world or wherever they wish to take their craft.


There are downsides to implementing the idea of course. It requires quite a bit of planning and dedication, and I know there are already a lot of ideas in the pipeline, but I just figured I'd put it out there for you guys to think about.

Any thoughts welcome.

9
Pixel Art / Dat sweet butt wiggle
« on: December 30, 2015, 03:26:05 pm »
So I'm trying to get that sweet butt wiggle on an adorable sluggy-creature thing (called a gastrosnap from some person on DA), but really this is my first time actually attempting real animation for my own cause.
Which means I suck. :D



What I want to achieve is a puppy like excitement. You know, when they get so excited because you're about to throw a ball for them and they start wiggling their whole bodies instead of just the tail. I want that.


Btw, the subject is a pun, because the creature is supposed to eat a lot of sugary stuff.  8)



So I had forgotten rule 1 of animating. "Stick figures."



Think my little blocky creature has way more of what I'm going for.



Added my base colours and outlines to help smooth out some of the movements.





Man this progress thread is fun.



Something is not right in that tail, but overall I think it looks much better than the first version.





Cleaned up the tail directions and added some of those swooshy lines people add to animations.
Also changed the duration of some of the frames. One was uniform the other not.



So I've slept and come back with a nice fresh perspective and I think I am happy without speeding up the frames or adding the particles.



Higher tail with the fixes. The lower tail wiggle looks better because more SWOOOSH, but this one gives me more of the feel I want, so sticking with it.


Gonna leave this for now, but if anyone can give me some advice on making this better I'd probably be more than happy to start all the way from scratch, so don't be shy.  ^-^

10
Pixel Art / Archer chick [WIP]
« on: October 03, 2015, 02:24:51 pm »
Inspired by some edit I posted I decided to try make an archer to go along with the swordsman in my avatar.



And the two next to each other for comparison:


Obviously the most important thing is that they go well together, both stylistically and thematically (are those the same thing?).
Any opinions on the colours would also be greatly appreciated because I am never happy with the colour palettes I end up making. :P

I might even end up changing the guy in my avatar as well, so any CnC for him would also be welcome.  :y:

Pages: [1] 2