Solving a BIG problem (GRAPHICS) ... if it is actually a common problem

So I just started 3D modeling and yup it is 10000X times easier and more natural to me than 2D

Pretty weird if you ask me

No clue I would love modeling this much.

 

Not really, 3d modelling is more analytic and less gestural, it' appeal to different skillset. But still having basic understanding of 2d (ie how to see a form) is important to improve, I would say keep doing the 2d exercises, especially obsevation and measuring, they will amplify your 3d skills faster than your 2d one, you will see.

 

Thanks you are right

 

Yep that's it exactly. A game doesn't need to rely 100% on just the graphics to provide context & meaning. I don't why so many people seem to think it does.

Give the game a description. Create a story. Make the blob on the screen feel like what it is supposed to portray. Provide some audio so support the illusion we're creating.

 

That means you might do fairly well in programming possibly. I also found 3D modeling to be much more logical and technical in nature than 2D art which makes it easier to create objects.

Anyway, that is AWESOME you find 3D modeling much more comfortable. It certainly would be idea for building vehicles like the Metal Slug. I guess you will still make the game in 2D and render out your 3D models to use as sprites. OR are you considering making the game in 3D (even if 2.5D)?

 

Since we are there, document and practice "edge flows", "edge loop" and "articulation topology" as early as possible, these are the critical element imho, I learn that and everything made sense after. Learn the importance of pole and quad too. And learn typical solution to recurring model, like character and all. Even though you won't make teh best model, having this as a basis would allow you to go fast with very decent model.

When you start I would say to stay around 3000-8000 polygon per character, that's around ps2/gamecube level graphics. ps3 model are in the 50 000 poly to give you an example, ps4 are in millions. 8000 allow you to get enough details to train from realism to toon, yet be wastful and still be decent and have no impact on performance. It also allows you to look at old model and learn from their topology, lesson scale to bigger poly so you aren't missing anything. Modern computer don't have much trouble pushing polygon. What's really difficult is low poly, especially below 1000, then the next level is extreme realism (think cinematic facial expression), in which you must be good in all thing art.

I would also advise you to stay away from fancy effect and shader until you are confident enough, no normal map any fancy extra, it's because finding your own workflow with texture will scale pretty well once you get basic diffuse. Also when you start having optimal uv isn't important YET, just get the job done, after that it's optimization. If you use blender use the automatic unwrapping, it's good enough most of the time to start.

 

I with you guys that find 3d easier than 2d as far as graphics creation goes. 3d objects/polys are much easier to move around and change than simple pixels. It is also much easier for me to get shading(etc...) along with materials, shadows, and similar stuff. With 2d art(pixel or otherwise), shading has to be done manually, but with 3d pre-renders, you can basically just move a light around and re-render things. The same thing happens if I need to make changes, be it clothing to a character, texture/material changes, and similar. I can just re-render it. Sure, it takes more work to get it started, but since I can't get what I call "acceptable" results with manual drawing, this is a better method for me.

 

All of this talk about graphics... had me wasting... spending time tonight on graphics again. The years leading up to a decade or so ago I used to just make graphics so much. Because it was fun really. No other reason than that. I mean I told myself I was making them to make a game but I finally realized it was just for the fun of it.

Oh well it doesn't hurt to have some fun. And I am having a lot of fun with this project. First I touched up the graphics again. Didn't spend much time on that. Just adjusted colors again on the environment graphics. Decided to make the first layer be in front of the player, enemies and interactive objects so I can put some grass, rocks, etc there.

Then I went on to adjust the barbarian helm and hair.I tricked myself into thinking this was needed... had to be sure the barbarian is final before I do all of the other animations. And I fell for it. Anyway these things were easy & quick.

Then I told myself to think of a story... just a beginning... just one thing... and I started thinking... and saw a lone barbarian walking down a hill he had been away traveling for weeks on the other side of the distant mountains. When he returned he found destruction had come to the little settlement. Of course, there is always that one person who has a few gasps of breath left in them who explained a strange band of misfits, bizarre creatures and powerful sorcerers had arrived killed the men and taken the women and children.

Shocked he hurried on to his home. And found someone very important was missing...


and even using a reference photo of a real life random brunette this took too long but I think it turned out good especially for being basically 4 colors. First had to find the right kind of photo. Had to have long hair. Definitely needed to be brunette because those show highlights the best I recalled from past experience doing this years ago. Then find one that looked easiest to reproduce not a ton of variation in skin tone, etc. Then on to bulking out the shape, cleaning up and details but obviously not much.

And now... for real... the ONLY graphics work is for the barbarian combat and an enemy and / or object for him to attack.

Anyway, I just shared this to say I don't have much of a problem with spending a little more time on stuff like this "cinematic" thingy reason being because it is not animated and will likely be used many times throughout the game. In fact I can change hair color, open the eyes, and slightly change face/head shape, change the hair and have another gal or two. And that is something the way the game design is coming together I will need probably. Anyway, if it was animated it would be a silhouette it what I am getting at.

 

I really like the new helmet

 

Thanks, I should have posted a screenshot. When I added the parallax scrolling of course the gif size increased by a huge amount over previously so that is why I shrunk the gifs down.






Anyway, it is all workable. More than what I need for sure to make a good game. Later I will get in another dev session and add an enemy. Probably just use the barbarian, recolor, change helm and weapon and presto there is an enemy for melee combat.

Hmm... change the helm. @neoshaman mentioned that viking helm. Now see I could do an interesting thing with the helms on enemy humanoids. Very simple or no helm for weak enemies. The bigger / fancier the helm is the more powerful the enemy. And it be something the players can easily read. I think I like the idea.

 

After trying 3D I will never go back to 2D hand drawing. In 3D no matter how much crap you make you can easily go back and just mold it correctly, can't do that in 2D you have to redraw it from scratch

 

IMO: The style of the portrait is very different from the style of visuals, once again. Your sprite is flat. The portrait has shadows on it.

I think you could try to take a look at .... Legend of Kage.

Alternatively, you could also check valhalla (cyberpunk bartender game), see their artwork and see how it compares to their sprites.

Spoiler

 

That is very true. Also you can build up some cool stuff fairly quickly completely out of primitives. And animating stuff is easier providing you have your skeleton set up right. You could build the metal slug vehicle just out of cubes actually. Just something to consider.

When I was doing a lot of modeling (a long time ago) I would find a reference image or even draw a sketch myself and then use that image and model on top of it. One of the best places I used to dig in a lot is The Blueprints. You may find some cool vehicles there to practice with.

 

Well in this case that is a completely different scene (the cinematic box on top of the actual gameplay window). I just honestly don't think it matters. I cannot imagine any gamer sitting there saying "hmm... odd... it looks like the lighting in that cinema box thingy that popped up was different than the lighting in the actual game." lol Why would anyone even care if the lighting was the same or not?

Now in the actual game (as in the gameplay area) that would be bad to have some things lit from the top, others from the top right, others from beneath all randomly. Then I could see it being a problem. Although still mainly I think just the nit-picky folks would think that unless it was really majorly ruining the display.

See I *think* that is what is going on. Here we are all devs. A lot of people are from the art side or at least very interested in the art side. So they notice all of this lighting, shadows, inconsistencies because it is the thing they are really interested in. It's like when I play a game I notice the controls feel off or the AI seems really bad, etc.

 

I think you may be underestimating people ^_^.

Well, how about this, then:

There's perspective switch going on for boss(?) fights or for trader screen.

 

Can't draw a circle to say your life? no problem 3D got you covered create an automatic circle. Can't draw a body the correct size? or to look human? no problem trace over an existing random pic of a human found on google images and mold it into place and you now have a perfect outline already filled in where you now mold in muscles etc don't have a wacom tablet to draw with and just a normal mouse? no problem 3D got you covered.

There is one thing for certain, learning some basic traditional 2D art gives you a better idea on how things are supposed to look off the bat so you don't have to trace over real humans. So you could now create something more artistic, think saturday morning cartoon like Team Fortress 2 art work.

But overall 3D is incredibly easy to work with compared to 2D art but it isn't unique as 2D art, for example take Street Fighter 5 for example, no matter what you do, the 3D models in SF5 looks like a cross breed between a chimpanzee and a buffalo's buttocks.

Now look street fighter 2 in HD sprites hand drawn, its one of a kind its beautiful its elegant, its precious it is beauty in every sense it is timeless it does not age. You could see the comic book 2D art sprites in SF2 HD remix. There is nothing like it on earth in 3D.

So as you can see 2D has its clear advantage over 3D but only when you have he best of the best in the industry like Udon comics in this example. For the rest of us, who do not have practice or time to learn how to hand draw anything there is no replacement for 3D its cheap and quick. But if you are doing pixel art its far easier to get better results in 3D especially when the engine does all the shadow and lighting for you free.

 

The first paragraph there.....my thoughts exactly! I have actually tried learning up on pixel art, shading, etc... and if I take the time(lots of it) I CAN get OK pixel art. But I take more time with pixel art than what I can take doing it in 3d pre-renders, and the 3d one is more "acceptable" in my eyes. I'm not saying 3d looks better than 2d, rather that with the specific results I personally get with each method, the 3d one comes out better for me.

I agree with the rest of your post as well. But as HD as things are anymore, it is too time consuming even for a AAA like capcom to hand-draw sprites. AAAs aren't willing to draw low-rez and scale up like many indies do, and in order to make budgets end up going 3d. I understand that though, HD pixel art costs lots of time and money.

 

Yeah see that is a good example. Even the gameplay area is not lit the same in this case. The sprites are all flat with no depth. The environmental objects / structures are 3D-like. The light seems to be coming from the top left for the most part and most of the objects are shaded that way but not all. But I think they are doing that to show what is the active play area and what is the inactive area. And the close-up battle mode the light seems to be coming from the player's eyes themselves. But that is a different scene so makes sense in that context light could be coming from anywhere.

But heck you can find a lot of games out there with inconsistent lighting. Sometimes artists just change the lighting on certain objects because doing so can make it look better.


I still think the ideal target for graphics is along these lines...

Proteus (probably ideal graphics style for any kind of survival, rpg, FPS, TPS game etc just set mood via lighting)

Devil Daggers (probably ideal for about anything else)

Gun Rocket (simple yet unique visuals... an ideal)

Dungeon of Zolthan (again simple distinct visuals)

You Have 10 Seconds (uber simple... and in doing so... unique)

.........

And in fact these are actually beautiful examples of minimalism all around. Not only did the devs choose a visual style with a very low cost (effort & time) but the games themselves also seem to be very focused tiny experiences. I'm not saying they are great games or terrible games. Just that the visuals were a smart choice. And the scopes are a smart choice.

On the other hand, what I am trying to achieve is a very simple yet unique visual style BUT I am doing that so I can make a game that is probably 5X the scope of the games above maybe even more. So I NEED to keep the visuals way down.

And I do not know but I think it is likely one or more of the 5 games above were created by a team of more than one person and perhaps people who have more time available to work on games as well. In contrast I am doing this stuff solo on a very part-time basis. Unless a person has just tons of free time or has a team of many people I don't understand why they would be doing things any differently.

As simple as my visuals are now I still have it in the back of my head that I am going way overboard on visuals. However, that is something I also want to test. Can I use a very simple style and then make it more interesting with programming (parallax scrolling, color glows, perhaps particles, etc).

 

Has anybody seen "risk of rain" ???

That is one of the most BASIC graphics yet looks and plays so well WTF?? so you don't even have to be good at 2D drawing you could literally have 1 rectangle box as a head with no eyes and its fine.

Pretty Amazing

 

The visuals look pretty good to me in Risk of Rain. The main thing I see watching this video is this looks like a fantastic game. I don't mean "looks" as in the visuals but the game itself. It looks pretty freakin' awesome honestly. Leveling up, a variety of playable characters to choose from, cool weapons to buy, an interesting interactive environment. I'd say they did a great job putting the focus on the fun.

The thing is most of the graphics are tiny which is probably a great choice for Indies as well. Kind of like Blasting Agent and BroForce. Little dudes (and dudettes I'd guess) running around.

I considered going tiny for my game as well for the same reasons. Quick graphics look better when they are that tiny and you can make the levels feel pretty massive even if in reality they are about one screen's worth of space.

But so many 2D Indie games have went the route of tiny to small character size I wanted to have something bigger and bolder a character about 1/3 the height of the screen and some enemies that will be much larger.

You are right though! Another good way to go would create the images tiny as people do for normal pixel art games and then not even scale them up at all and just keep them the actual size in the game. Going that route we just need to make sure all of the supporting stuff is in (the FX when the characters move around, shoot, etc) that help gamer's mind to associate this little box of pixels is actually a little humanoid armed with a weapon. Of course, we should be doing that supporting stuff in any case.

And heck you know that is an option for you. I know you like 3D modeling much better but you may well be able to draw and animate a metal slug vehicle that fits in 8x8 or 12x12 pixels and have the game in miniature.

Thanks for mentioning Risk of Rain. I heard about it many times but never checked it out before. I like the RPG aspect tied in with the run n gun platforming. LEVEL UP. That is very much like I want to do. Great stuff.

 

Simple doesn't necessarily mean easy though. You may not have to be good at drawing but I'd argue that you have to at least understand art. Because while some people may find meaning in the jumbled pattern of shapes and colors that programmer art tends to be most of us need it to at least resemble something we would expect it to look like.

 

^ Yeah the guys who made that have a good sense of art they know their stuff.

BTW has anyone seen this as yet? made with Unity, bloody amazing.

 

Graphically it looks very well done. Enemies seem to have the same basic behaviors as previous games. I think what it is missing is there seems to be no challenge other than the boss battle. But maybe that is because the dev is playing game and knows exactly where to stand, exactly what is comng up next, when to fire, etc.

But it might be level design too. Like I think only one spot needed the ball. Enemies are few. It is impressive for one person though.

 

*cough* *cough* Guilty gear xrd

 

Yeah, I'm liking the Metroid 3 remake there....it would need to be more complete though to fully make it, as right now it is more of a tech demo.

Yup, that game shows that 3d can indeed mimic 2d styles in a way, and pretty well done at that.

 

Okay, after working on enemies today I decided to just completely throw out the idea of creating animated characters that have some shape to them.

Of course, the main reason is because I just can't be spending all of this time on graphics. But the reason I chose this specific style well...

1. I truly believe that a fun game, a great game, will be fun regardless of how it looks and have said so many times. So this can be a put my money where my mouth is so to speak. Note: I'm not saying that graphics cannot add something to a game. More that I think a game should be able to stand on its own regardless of what it looks like. Then if you give it great graphics on top of that you've got something very special.
   
2. This is how I am used to making games. First just using uber simple shapes usually solid color rectangles. Just enough to express the basic ideas for size, animation, etc. Then I can focus on the graphics at the very end. Either keeping them exactly like they are, redrawing them all myself or hiring an artist to do it.

So... meet the barbarian...


Once I have this rectangle behaving like / projecting the essence of a barbarian which will primarily be programming & audio work then I know I've done good work.

This decision should make things easier on me so I can focus more on the game. I was wrestling way too much with graphics up to this point.

The take-away for non-artists or anyone really who doesn't want to grind endlessly on graphics work is simply often you just need to throw them out. As far as attempting to make any kind of humanoid shape and animate them. Just use the simplest graphics you can get away with to implement and test the mechanics. And focus on the essence of the animations needed to bring things to life.

 

My thoughts....if you want the barbarian to be square like that, I think the trees, trunks, sky(clouds), and ground will need to be converted to a similar style. The gray square there just looks "off" compared to the rest. The ground could be just some think lines, maybe some circles for "shrubs." The trees should be either squares or ellipses(or some of each), and the same for the clouds in the sky.

 

This is a good start. You are close to his one


It reminds me there is a successfull 8 bits kickstarter

 

Yeah Barbarian on the C64. I played the heck out of that game. They made a sequel that had more to it a flip-screen game. But the original was cool especially being able to chop their heads off. Very rare for the tim.

I've seen Unknown Realms on one of the retro Indie sites I visit. That looks great!

 

It definitely looks off but I don't care about that. It would be foolish for me to take the time to convert the background to be rectangles at this point. Now later on when close to release sure I may very well do that. But to spend all of this time just doing nothing but altering the graphics is IMO insane. lol

Completely get what you're saying. I do wish I had just went with rectangles from the very beginning because it would have time on all of the iterating I went through. But at this point I am not going to worry about it. Actually just about to get started working on this for tonight's dev session.

 

Okay, I'm trying a new thing where instead of trying to work on this game dev a bit each day I will instead do it every other day 3 to 4 days per week. Figured that will help with burnout. Of course it will also double the amount of days, weeks, months whatever that is needed.

The barbarian block is starting to feel pretty good.

• First is the normal jump straight up
  
• Then the moving jump / leaping (which cannot be controlled in air... I am still debating on that one. I do have something in mind for this leap I will talk about a little down below)
• Then the normal attack (which actually this is only a beginning... I have something greater in mind for this)
  
• Then you see his dash attack but I think the word smash fits a barbarian better than dash. But I might call it dash attack in the end. Who knows.

The leaping I am thinking of not allowing the player to control the character in the air which may make it very tricky for the platforming aspect of the game. So I might do that just depends on when I get to the platformer prototyping part.

One thing I do want is to have a leap attack. Basically how this will work is when the character is at the top part of the jump (say 1/5 of the way from the jump expiring and maybe 1/5 of the way into falling) and the attack button is pressed the barbarian will slam down on an enemy.

Right now I am considering doing it like this... if there are no visible enemies he simply smashes straight down to the ground. BOOM!

Otherwise he automatically targets the closest enemy in the direction he is facing and performs a slamming down attack on them. Mega damage. Reason for this is because I think many devs overcomplicate controls in their games.

Sometimes you need to press down and attack when a character is in the air to perform a ground slam attack. I see no sense in that unless controls mastery is one of the objectives of the games. Just press the attack button and he will do the slam attack. Of course, those guys probably allowed you to jump up and hit stuff with a normal attack. So hmm... I might need that. I was leaning toward a jumping attack controlled similarly to the way the slam attack works. BUT haven't though much about it in detail. Right now I am literally just adding what is in my head and then adjusting code til it feels right.

I won't post any more of this here because I don't want this to turn into a dev log for this game (thinking I will probably set up a dev log over at GameJolt and the AGK forum) and also I am not creating this game in Unity.

But I wanted to follow-up on the discussion of graphics to at least bring it to a conclusion. And I find myself thinking even more yes you really could make a cool game like this using just rectangles. I might just go with the rectangles ultimately instead of replacing graphics. I will figure that out at the end if I get there.

 

I still want to see some horn on that rect just for fun

 

You people and the graphics stuff. LOL

 

Trying to create bosses for my little game, sigh one frame equals 60 minutes (and still needs so much refinement):

​

 

That looks very good... it's highly readable. It just has a more "technical" than "artsy" look to it. Animated and seen within the context of the game it could be excellent.

It is crazy how much time it takes to do graphics.

 

I almost qualified it in same manner but what is "acceptable"? It is a vague term with no real meaning. Gamers (including myself) play games with graphics ranging from solid colored rectangles and untextured cubes to AAA DOOM quality graphics.

That's why I left it at "graphics".... referring to actually attempting to draw or model a real shape and animate it. Where just using rectangles and cubes sure they are still graphics but in a sense they are not as far as no time spent on trying to give form to them to create a "real" shape of the object they are representing.

 

IMO, it is when you see it two days after you made it, and you don't hate it.

 

Vegas News:

'One armed bandit' escapes casino, mob pursues it in attempt to retrieve it's cash deposit and ignores the police warning of it being armed and dangerous.

Take control of the 'one armed bandit' and fight for your live, you have no friends!

 

Ha ha! I wonder how often that happens? From what I've seen most people (particularly artists) seeing their own work 2 days later would see many things that need to be improved.

My theory on the graphics thing is that it all works kind of like this... gamers view games and note the graphics in them based on 2 things: 1. how the graphics compare to their own art ability (higher ability results in higher requirements to please them) and 2. how visually oriented they are meaning how much value they place on how things "look" in general.

For example, I am aware of the art in games. I notice it as far as I know the "art" is there. And at the same time regardless of how artsy (just how the graphics appear completely judged as "art") a game is I also interpret it very differently when playing not even noticing the "art" because it is completely irrelevant to me when playing.

Like when I play this...

I am aware of this...


But I basically only "see" this...


Which has the same meaning to me as this...


The game could be completely represented with cubes or other abstract shapes and play exactly the same. Providing the story was still there providing a "mood" and the audio and graphics worked together to provide the same level of feedback (knowing I hit them, knowing they hit me, etc), the controls were the same, etc it would basically be the same experience for me.

 

don't forget that good graphics are also a way for players to not have to "interpret" symbols, serves as feedback etc. ergo can make it easier for someone to immerse into a game, if it's easy to identify everything on the screen.

 

Which is good.

However I said "don't hate". Never said anything about it having "nothing to improve".

 

I completely get that. Reason I made the rectangles representing the enemies red in the final illustration. Graphics have to be readable. Quality of graphics as in detail, "beauty" etc doesn't go hand-in-hand with readability. Color choices can and behavior definitely does though.

Unless there was some kind of legend at the start of a game for example when I saw a red character (rectangle or otherwise) my first thought would be it is "bad". If I saw another that was say yellow I might think that represents enlightenment/knowledge/"good". Grays perhaps neutral neither good or bad.

However if the yellow character was rushing toward me shooting bullets I would use the behavior to override how I interpreted the colors. And in that case I'd think hmm... things are not necessarily as they appear. And would likely move over next to the red characters and perhaps discover from their behavior they are actually quest givers, healers, etc.

 

yeah, I have to especially agree with how graphics detail can also even have the opposite effect of readability....good examples I personally experienced is the difference between Borderlands 1 and 2, while 1 was focusing very much their detail on where the player should pay attention to, therefore "leading" the player in a wanted direction quite often. Borderlands 2 was so convoluted with details or "noise" that even with a sniper rifle aimed at an enemy that i saw moving in front of me, I had trouble "recognizing" the enemy character infront of the background....made Borderlands 2 very unpleasant for me...and took me a while to understand whats going on, because on the surface those games didn't look any different.

And yes color choices aswell as shapes etc. I have often thought about the readability of icons as example, because we are so used to using symbolic for it, but I know that symbolic can differ from culture to culture and from each generation aswell. Ever thought about how a greek temple as icon for "knowledge" could become at some point completely not understandable anymore in the future? or thinking about as example how to display health if not in form of a heart or cross?

Also, like the example you mentioned, we are pretty much "trained" to perceive red as danger, the enemy, something bad or "attention!" and if something falls out of that we immediately perceive it as less intuitive, even if we CAN get used to it.

oh and....now I'm just rambling about my thoughts XD

 

All great rambling though!

Completely agree on shapes as well. The enemies could have been triangles and that may help with their readability because pointy things are "bad". More rounded shapes are happier/safer/friendlier. Rectangle is strong. Tough. Solid. And so forth.

That is the kind of stuff I have an interest in for graphics in games. Readability, feedback communicating with the player not simply art for art's sake simply out of focus on artistic skill and creating beauty.

Not that I don't enjoy art and just looking at things for the sake of art. Not at all I look at a ton of game screenshots, game art for WIPs, nature photography etc and photography is one of my hobbies as well. My FB for example has many albums of digital photography for friends and family to enjoy. I used to take little road trips fairly often just to find something interesting, a cool view, cloud formations, animals, a sunset etc to take photos of.

 

Yeah I get that. Just that a lot of people seem to have this problem where if they notice anything.... as in there is something, anything, that can be improved... it seems to almost drive them crazy and they can't stop thinking about it... so they kind of do hate it if there is even the slightest bit of anything they notice that can be improved. And you know as well as I do that is a long road... I think a person could look at anything in any game AAA or not and think of ways to improve it.

Yet ironically, it might be much easier to simply accept these as they are...

 

Any information to share about the art styles you have tested and where you are now as far as your modeling workflow and the end result you're going for.

If so please feel free to post screenshots, animations, sketches and so forth in the interest of helping all of the folks struggling with the huge workload graphics represent.

 

On topic of pre-rendered 3d: is it just me or it is actually used in modern "pixel" games more often than we think?

I'd like to know your opinion. The first video most certainly uses flat colored lowpoly models rendered with antialiasing off. The second example must've involved heavy 2d post processing, but still suspicious.

 

I wouldn't be surprised if pre-renders were used more often than advertised. I'm sure they do a lot more "cleanup" on them then I do with mine, even if just to get things into the right palettes.