So, i'm going to be updating this thread with sprites from an RPG i'm going to be working on, well, i'm still torn between a contra style game, or an RPG, but i think i can squeeze out more innovation in the rpg.
What I'd like polycounters (you) to do is help give me crits, pointers and insults at every possible juncture so I can fix stuff that looks too shitty, and polish it up to a level that might make it onto XBLA community games.
When it comes right down to it, indie games do rely more on innovation and refreshing gameplay to succeed, but I don't want people spiking the controller into the ground because it looks like something from 1983. :thumbdown:
The game's style and perspective, for the rpg, is going to be like final fantasy, secret of mana, chrono trigger, zelda etc, etc, etc...
However most of my inspiration comes from Seiken Densetsu 3 (secret of mana 2) on the art style i would most like to follow.
But almost more important to me than quality, is content, so i dont like spending a long time on artwork, so i can make more artwork, the goal of that being avoid heavily repeating tile sets, objects etc.
Replies
i really like it up to the tree with the spider web, then it gets really rushed, imo
i think with following P442 suggestions you can get a lot more quality out of your work !
Look at old sprite-based games for reference and inspiration (Metal Slug!)
Your stuff reads as really flat, you need to add alot more tones and shades to really beef up those sprites. Look at any old 2D console game and you will see that the artist were masters of giving a visual weight and depth to 2D sprites.
When drawing sprites I never do one offs like a rock then a tree then a house unless I am making the initial style guide. You will find that if you are using your perscribed style and make say all the trees then all the rocks...etc etc, then you assets will have a much more cohesive look.
If this is something that your doing on your own and not for work then take your time and make the art look really polished...cheers!
So instead of actually bothering to improve your skill you just want to keep on going and make art that looks like some random guy I picked out on the street could have drawn it if I gave him a wacom. I dont want to be mean but your drawings have 0% style to them, they have absolutely nothing in common with the style you mentioned.
What will you learn from making halfassed rushed work besides how to make halfassed rushed work?
Are you sketching first, to then turn it into pixel stuff?
I don't think your work is considered enough. Pieces have to not only look good in isolation but also work interacting with the other art and the characters themselves. What's your palette? What's your lighting direction? What's your lighting style? The large house appears to have bloom lit effects on the roof and a cast shadow, but nothing else does. The shop's shading suggests its light is falling in the opposite direction to the other props which are variously upper left or directly above-lit (and on one model, both).
The shading needs a lot of work- if you want to create a style such as castle crashers, the nearest I can see to what you're after, take a CLOSE look at the very, very clever use of line and color in that game. You've not really represented any forms- the cloth on the hut thing for example is just a bunch of lines, no sense of folds or shape, and the vines are just green squiggles. In art like this line is going to be massively important, and your look ill-considered and lacking character.
Finally, you've not given any consideration into indicating materials- e.g the stone of the buildings, etc. Cartoons of the 50s and 60s such as the Flintstones show how to simplify materials but still provide good visual clues that are appealing.
The worst mistake is to think that simple means quick or easy- often finding the essential line or shape that communicates a form without any undue detail is the hardest thing of all- ask any traditional animator!
And don't use the defence that quantity is better than quality. You keep shovelling out low quality stuff, you just end up with a pile of crap. Consumers will forgive SIMPLE graphics, and indeed fall in love with unique styles, but crappy graphics are unforgiveable in the current era of gaming art.
Hope this helps
~P~
A while back I had a contractual opportunity to do something similar. One of the first things I did was compile reference material, and here were a few of the sites I bookmarked.
http://www.kennethfejer.com/
http://www.madpxl.com/
http://www.itchstudios.com/psg/
more specifically for you:
http://www.itchstudios.com/psg/art_tut.htm
Mr. Patterson nailed it with the palette statement. When you're going for this low end stuff the art has to pop in a way that the user enjoys what they see, but it does not distract from the game play. To be a bit more specific, distinct palette choices for interactable or navigable versus background and foreground environment. Palette for midtone, shadows and highlights for each material type presented for unique zones.
When confronted with your situation it's going to be technique that's going to make things happen. Let's be honest, getting better at painting art takes time, but you need results. So the object here is to devise a speedy technique to output the best results for you skill level. Drawing up individual bushes, rocks and trees isn't the answer. You're going to end up with a lot of different colors, quality levels massive gaps in detail, and exactly what you wanted to avoid "heavily repeating tile sets".
Step 1, layout your entire environment in Silhouette form. The amount of which you want to do is up to you. This could be the entire area or a mock screenshot, whatever floats your boat. The idea here is to get an exact on screen representation of what the user will see. These silhouettes should be a solid color to represent their midtone palette (as discussed above). Things with different colors don't matter right now.
Besides presenting an accurate end results of your project, you'll have a visual reference of when reusing a stone, tree or building is ok without looking repetitive. Additionally, you can then break it all up into organized layers in photoshop of each asset. This gives you a pseudo-schedule. It's really up to you how organized you want to be about all this. Nevertheless, you now know everything you need in an area, and can begin the artistic assault on you area.
Again, this is all opposed to randomly making a few trees and shrubs, and actually making something useful for the long-term of your project.
Step 2. First, I'd like to point out, this is unique to YOU. I'm basing it on the quality of work and project your present to give you the best results possible. Other artists would get different assessments, but this is how project planning works. Ok back to Step 2, Paint all of your sprites in black and white. I want you to make FULL use of the 0 to 100 brightness range in your work whenever possible. Your new favorite tools will be Levels, Histogram window, Contrast, Equalize, etc. Working in this method will allow you to dodge and burn all you want without looking like a complete tool since it won't effect the final colors.
Now a quick step back here, as you are painting, remember those silhouettes we created earlier? Those are important, ALWAYS keep those as base layer. Why? Because you ctrl+click them before you paint so you keep all your painting in the lines! This is your all important MASK layer. Secondly, you need to create a mask layer for any other material presented per sprite. For example, the first brush would have two, one for the green leafy area and one for the brown roots. Again, these are simple, single color silhouettes. I keep most of the solid black or white. More of their importance will be realized later.
Step 3 Your palette. Assign a midtone, shadow and highlight color for every material you want to present in your zone. It's important to note now that pure black is not a shadow, it's useless as is pure white. This is your opportunity to do very exciting things artistically with your work such as deep purple for shadows or whatever. As long as it's consistent all around around looks cool. You can make very exciting things with color choice alone.
Back to all that greyscale art, it's time to assign the palette. Once you've done a couple of these you'll realize just how much power and control over color you're about to get, and down the road, this will teach you a lot about how to handle color painting in the future.
Take one of your sprites and use the base mask to select an area you want to color on your greyscale and go to Gradient Map. You can click through the defaults to see where we're going with this. That's not important, but may be a good base to start with one of the three color gradient maps to build off. From here, simply drop in your color palette of your shadow, midtone and highlights for that material. Getting their placement correct will take a bit of eye-balling based how much advantage you took of your original levels. Once something looks right, SAVE your gradient! Now you can select all of your silhouette masks for everything of that material, and load the gradient map you just made to it, and everything of that material in your zone will be colored. I suggest going a bit more asset to asset at first to ensure consistency of your original black and white painting.
Gradient Maps are very powerful, and can allow you to smoothly transition between your palette or do something more like toon shading. It's really all up to you. The whole point here is to get all your work clean, colorful and consistent. I've used technique to do full portable game environments speedy and efficiently as well as mass palette swap assets for elemental variations.
Hey remember those masks we saved from the start? You can select them, go to the hue/sat editor and make them anything now. This is perfect for making those shutter green, door red and so on.
@P442- Meh, yeah, 10 minutes is about right, okay, so you want better lines and more detail. It depends on my programmers, but hopefully there will be real time lighting in this game, so i should actually stick with no particular light source and just vary values for interest?
@Acumen- Yeah, its notably funny how things come across, the 3 trees took me by far the least amount of time, the most time is going to be the house with the bloom on it (saved on seperate layer), also this house was traced from a seiken densetsu 3 house. I agree i got too excited...
@Zenarion- Welll... I'm kind of fixed on this idea, not only for just a general style, but also to make the animated sprites easier for me. I wont be using 3d to make the animated sprites.
@Canadian Ink- Yes, you've addressed maybe my biggest problem in this art, and all art in general, i'm not good with making things pop, or read non-flat, probably the biggest thing i will have to overcome for this, but i'll just trace a bunch of real rocks and trees if i have to.
@Kwakkie- Well i guess the whole goal of working on this is to improve, but not in the sense of quality or speed, but both. Thats whats going to be most hard, getting good and fast at the same time.
@JohnnySix- I am heavily reffing that game, or tracing from it, but i'm not converting to pixel, i'm not making things pixel by pixel, i figure since this game is going to be played in 720p or in a scaled down 720p aspect, these size sprites were small enough, pixel art from what i've seen on websites can have great results, but takes forever.
And again my point here is, you can have a game with 2 monsters that look really good, varied slightly throughout the game (gears of war, grunts and boomers) or you can have a game that doesnt look as good but has quite a bit of variety Mass Effect (few types of geth, zombie guys, humans, krogans, asari, other races etc)
@praxedes- wow, thanks for the feedback,
Firstly, i dont think i have a limited pallette, unless i should? A limited resolution certainly.
Looking at castle crashers what you compared me to, actually looks like something to strive for
-everything is two-tone shading
-every line is appears vectorized
-very simple
Thats also a good word of advice to look at old cartoons for indicating materials, thanks.
@cholden- WOW, I feel this was extremely helpful ( i read the whole thing )
My one question/concern on what your telling me is the mask thing, is it possible for me to do the greyscale painting, and then do a colorize layer over it, instead of the masking thing? I have seen this technique done quite a bit in some digital painting. Colorize will only hit rendered part of the images, not the blank space, but without the mask i'd have to zoom in and make the line between materials very defined, i think that would work out though.
I'm striving more for small art, rather than pixel art. Like castle crashers i guess.
That being said your 4th link seems the most helpful with just general information on values and making things have weight.
Again, thank you to everyone, i did not expect this level of critique and support, i'll continue adding new sprites to this thread so you guys can keep me in line. Thanks.
Here's a VERY old tutor I did showing pretty much exactly what I was talking about. http://chrisholden.net/tutor/masklayer.htm
But again, the thing is, if you're painting sprites in photoshop, this base layer happens automatically. So your point is invalid. :P
I think what you have there is fine. I think it even has a sort of charm.
if you're not a pro artist, then get a partner... if you post here, you're going to get people who are dedicated artists comparing you to their highest standards... and pardon me for saying so, but..
a small indie game made by a dude.. mayybee shouldn't be concerned with the highest possible art standards.. and if you're not an experienced artist, it's going to be difficult to pick yourself up to the level that these guys would expect.
I think the stuff that you have there is clear, and might make for something actually fairly interesting if you compose it into a scene correctly.
I guess it depends on what your goals are though. I suppose what i'm saying is.... keep going, and learn and improve, fine.. but make sure you don't get caught up on re-doing shit until it's at the level that these guys will suggest.. because that'll probably take for ever, if it happens at all, and when i do that type of shit i get discouraged.
you made a tree. fuck it. use the goddamn tree.
Anyone know of a tutorial for masking in painter? (i use 9) a brief googling and google video'ing didnt result in much...
Anyone here have any experience with 2d real-time lighting in xna? I stuck on whether i should just give all the sprites shadows from 1 angle, or leave shadows off them so we can do day night transition effects etc in xna...
your going about the shadows all wrong, the are ending up being drop chadows and making your sprites looking like they are floating in mid air.
I'm not saying you should tile everything down to a 16x16 like oldschool RPGs, but even castle crashers has reused assets and tiled sections. The ability of the artist has a lot to do in hiding the seams, but things are tiled and overlayed several times over.
look at your screenshot and squint. see how everything, except the house fades into nothingness? it's because everything in the scene has got the same value and saturation levels.
go up to that castle crashers screenshot and squint. completely different thing, right? see how the ground is desaturated, and the character pops? he's also being framed by this nicely composed scene that sort of "hugs" the important elements.
when you compose a scene, keep in mind where you want the viewer's eye to go. give that contrast, in terms of color, value, and detail. when i see the castle crashers scene, I immediately notice my player, my life bar, and the girls who are tied up. that's all the important information.
IMO i would
desaturate the ground and remove some detail
Pick one light direction and light the sprites from that direction. pick a warm color like yellowy orange for the sunlight, and put the shaded side of the object towards more of a purple.
You can create shadows easily by mirroring a sprite and turning it black, and stretching it out away from the light source from the base of the original object. then, blurr the end of it. baym. instant shaddow.
I would bring the edge line of that forrest down more. See hwo it's ending in line with the house? that's a little awkward. try and bring it down so that it snakes off to the right a little bit.. and maybe even add a few trees behind the house on the top left corner.
frame stuff.
but it just isn't there yet.. something is off...
Outlines shouldn't be black. Leaving black lines will kill any sprite you put in the game. It's simple, quick, and will make a ton of difference. It's good that you're getting a lot of help here, and that you're willing to improve.
But take a look at John's last screen - notice how his 2 second mailbox thing seems to fit into the environment so much more than the house or the leafless trees? If you have to, just make the outlines a slightly darker color than whatever the most nearby color is, but black is just asking for bad presentation.
Keep at it.
ps, I have to ask: why painter for this?
@rooster- Its my favorite, and i cant get photoshop to blend like i want it to without slowing my computer down considerably, i've messed with everything scattering, size dynamic, shape dynamic, turned it all on or off, varying inbetween, dont like it, cant make it work for me.
Thanks again for all the input and help with the paintover, guess we'll see how much of this sticks and helps me improve on my next update.
http://www.2dplay.com/nevermore-3/nevermore-3-play.htm
The style's a bit different, but I think it'll help you out with seeing how you can flesh out shapes and use color
I didn't say "never", I simply said they shouldn't be black. Typically. I guess we can look at it this way, as Rooster pointed out, there are certainly exceptions. But as you noted, black outlines will make things stand out. You do not want every aspect of every character, tree, house, or other environmental prop to stand out. And given the games he mentioned explicitely as references, as being similar - and what he gave us as a screenshot... The black outlines do without a doubt kill a good portion of the sprites. Even in a limited resolution, there's a whole lot more to "darker edges" than pure black.
True Dat
4 minutes... it will even take more time to say whats wrong in this piece compared to any other 2d game
no offense, but if you want anything maybe offer something at first?
Spend and hour making the best looking rock you could possibly make....then come back and ask for crits. If it looks like you cant be bothered with putting the effort in then people here wont want to waste there time giving you the benefit of there experiences in the form of constructive crits.
lFirst animated sprite test, used flash, with the new bone system, combined with standard pivot tweening, took a character made in painter, broke it apart, exported each layer (moving part) as png, imported, connected with bones or animated it with just pivoting.
Would it be better to do it thisway, or with vector graphics (within flash) so i can scale it up and down as much as i want?
Is there, another better way i should do animation? I feel that hand drawing frame by frame, although better looking, wouldnt end up as smooth, and would take longer.
Order.
Read.
Try again.
am i a loser if i trace it, and then alter it into my own character,
this took me about 30 mins, in the final version i think i would change the boots and hands more...
Also, does anyone know of a nice walk cycle sheet for facing the camera, and walking away (i know i can use sprite sheets as a ref, but these nice walk cycle pictures tend to be better)
Also, i did obtain that book, it seems pretty helpful especially regarding information of inbetweening,
also, in case this ever becomes a sticky showcasing what not to do, omg this guy sucks,
then here is a link to help other 2d artists,
http://www.davidhellman.net/braidbrief.htm
its the art archive from braid,
edit: fixed the leg in first frame.
um on a sidenote. i'd just suggest you don't post how many miniminutes you worked on what piece, just because....well, people will always take this as first crits. i think you can save this
I think the only problem with that rock is that the yellow was way to fucking yellow, and looked more like a yellow coating than sunlight.
I mean, he's not going to be able to hit a super high quality bar on this project --?? am i an asshole for saying that? I mean instead of expecting SOM2 from him, why not just lower the bar to his level and give him some art direction?
And honestly the best crit for just wanting to complete a game would be "Spend more time on your game design and stop caring about the art and go hire an artist or choose a style that even a braindead monkey could do well".
So I don't know, what sort of crits do we really want, here?
I think the main problem with the rock is that it looks like quick sloppy scribbling... probably because it is. That could be turned into a style if you wanted: just make everything consistently sloppy and it would actually work.
Here's something I did pretty quick (like 10 minutes) based on the tute Holden posted, I also used some secret of mana 2 shots as ref.
holy saturated colors batman!
oh yeah, read up on the art of braid
http://www.davidhellman.net/blog/the-art-of-braid-part-1/
http://www.davidhellman.net/blog/the-art-of-braid-part-2/
http://www.davidhellman.net/blog/the-art-of-braid-part-3/
I personally thought that yellow stuff was supposed to be moss, not sunlight.
It is true to say that art doesn't necessarily need to be realistic, or follow any particular style in order to be really badass. But ideas and visuals should by all means be recognizable as what they're meant to be. Moss or sunlight? Mansion, or shack?
Presently, pieces seem very rushed. I am not expecting the best sprites I've ever seen. But I am expecting some semblance of care and effort. At least, this is what I'd expect to see from anyone who wants to be taken seriously.
It is not to say the artist in question is without talent or potential, merely that neither of these things seem very important in 4 minute rocks covered in what could be moss, sunlight, or mustard.
To the OP: Keep at it. There are two ways to salvage what we have here. The first being, to take some serious time and put up the best possible sprite you think you can taking into consideration all of the advices given here. The second being, to keep posting regardless. Even if not from an artistic standpoint, there is some level of pride to be taken in doing it for the sake of doing it, not for the sake of supportive responses.
Cheers and best of luck.
everything looks rushed - taking the time for just 1 thing is the most important part of experience imo.
Okay, you're getting it, when nobody else is.
When making an indie game which you wish to sell you have several directions you can go.
A) take a long time to make it, not really introducing any new ideas or gameplay mechanics, and make it high quality (Machinarium(took 3 years))
take a normal amount of time to make it, make it look and sound good enough, but introduce new, interesting fun and dynamic gameplay. Turning a genre on its head, or showing people something they've never seen before which is also fun. (Crayon Physics Deluxe)
--So basically, I'm the braindead monkey. My two friends are the programmers, and we're all the designers.
The goal with the graphics is "Good Enough"
We want to sell the game with the gameplay and story, but not alienate the customer with the graphics.
What you want to do is earn easy money, and thats a thing that doesn't exist, especially not with games, even the "small" productions that sell a lot like Castlecrashers, Braid or World of Goo took a lot of time to develop and polish it to the final state, you already read the articles about braid, so i guess its not worth saying it again.
If you want to make money, go for a simple iphone app that a lot of people would buy for not so much money (games are definitely not the cashcow, sell shit like that beautycheck app that compares your photo to standards and gives points and a ranking, hope a magazine like cosmopolitan sees that and has an article about it, and bang thousands of stupid chicks will buy your app), and maybe you are lucky enough to be one amongst the thousands that are trying it that way.
Maybe even think further and work on the new google OS for Mobile phones, if it really gets as big as it sounds, you'll have a giant platform to sell shit to, but then again, rushing won't be the way, do it simple, do it good, do it with dedication. It just looks as if you dont really want to do anything, but hope that no one will see it.