Should a beginner learn Blender for making a game or use Pixel Art?

[CodeSlug]

@Billy4184

It’s up to you. But remember pixel games have their own thing. I’m not sure you can just transfer them to 3D and get away with it.

My point was more general. I think you’ll have a much easier time with 3D, but if that’s your dream game you might as well start learning 2D now.

Axion has not been my only goal at developing a game I always loved the idea of a low poly side scrolling game also one that has a similar artstyle to something like resident evil or luigi hunted mansion but in low poly. A shooting game

We have to be realistic, if I utterly suck at 2D pixel art is either I practice more at it and learn traditional art or switch up the style of game to fit something in low poly 3D that can be manageable.

[Billy4184]

@CodeSlug

That’s the thing I SUCK at Pixel art, using colors and doing anything art related. I was better able to make a cylinder as an arm and a square as a head in low poly 3D and using UE4 to do the lighting effects the visual image I got when I did that in 3D looked like it would take me a million hours to get the same effect in 2D.

So what if I change my goal of making something like Axiom Verge, make a cartoony sci fi axion verge or something?

It’s up to you. There’s a lot of successful games out there of all kinds that you can take lessons from and use to see what works and what doesn’t.

In my opinion, if you struggle with 2D art and you’re not set on it, 3D might be up your alley. Ideally you could try both and see what works for you. Especially at the start, you should be experimenting - I tried 2D art (not pixel) myself for a little while before I realized I just prefer 3D and have a much easier time. And I have reasonable drawing ability already.

Different strokes, and all that.

[neginfinity]

@CodeSlug

I assume Unity would be far better for making a Axiom Verge clone than UE4?

Yes. Unity has better support for 2d than Unreal 4.
Do keep in mind that original axiom verge was written from scratch.

@CodeSlug

That’s the thing I SUCK at Pixel art, using colors and doing anything art related.

Sink few hundred hours into it, and you’ll improve.

@Ryiah

No. After watching a quick video of the gameplay I don’t see anything in it that isn’t easily covered by the features available in UE4. You don’t need much more than the basics to make a game like that.

Unreal engine doesn’t keep sprites very high on their priority list, they’re heavily geared towards 3d.

Axiom Verge also heavily invested into replicating NES era feels and NES hardware glitches, and for this kind of game you’d heavily benefit from support of palettes. Neither Unity nor Unreal has this feature.

[CodeSlug]

Alright I am gonna try both 2D and 3D again and see which one I succeed best in

[neginfinity]

@kburkhart84

Now there is a level of skill that I call “acceptable.” It isn’t good, but is “serviceable” in a lower end game. In my opinion, it is much easier to get to that level in 3d than 2d. The reason is that 3d feels more forgivable, and errors in judgement and lack of skill can be better covered up.

However, it is MUCH harder to produce acceptable animation in 3d than in 2d. As animation in 3d is its own separate discipline.

Additionally, there’s 2d, there’s 3d, and then there’s 2d pixel art. Pixel art is easiest from all three due to the limited choices you have.

[kdgalla]

One thing that I find that helps with pixel art-
I will draw something on regular, old pencil & paper and then scan it. Then, In my software, I crop the image and scale it way down to 32x32 or whatever size I want to make the sprite. I don’t actually use that drawing as finished art, though. Instead I create a new, empty layer and draw my pixels over top. The drawing, however, gives me a rough indicator of how to make the shapes in my pixel art, so it’s easier than just starting with a blank screen.

[neginfinity]

@kdgalla

One thing that I find that helps with pixel art-
I will draw something on regular, old pencil & paper and then scan it. Then, In my software, I crop the image and scale it way down to 32x32 or whatever size I want to make the sprite. I don’t actually use that drawing as finished art, though. Instead I create a new, empty layer and draw my pixels over top. The drawing, however, gives me a rough indicator of how to make the shapes in my pixel art, so it’s easier than just starting with a blank screen.

That’s a good advice. Same thing can eb done on computer by downscaling a high resolution image.
Basically you can draw a thing - even if it is horrible looking, and then downscale it.
Then convert it into pixel art - either by placing it into bototm layer or by keeping it at your second screen for the reference.

I heard that Capcom used high resolution scans to create their sprites. But it is unknown whether this is true or not.
Related:
http://gas13.ru/v3/tutorials/sywtbapa_making_fighting_game_sprite.php

[kburkhart84]

@neginfinity

However, it is MUCH harder to produce acceptable animation in 3d than in 2d. As animation in 3d is its own separate discipline.

Additionally, there’s 2d, there’s 3d, and then there’s 2d pixel art. Pixel art is easiest from all three due to the limited choices you have.

I agree that animation is generally a separate discipline(and in fact mentioned that just above here). But I’m not convinced that animation at an “acceptable” level is easier in 2d than 3d. I can see how the point could be argued though. I’d say it depends on too many factors to be able to properly argue even.

And yes, pixel art as a subset of 2d is indeed far easier in my opinion…especially at the acceptable level. I’d still say it is very difficult and time consuming to become expert at it just like any other type of art, although I’d agree it is easier than other forms due to the limitations.

[neginfinity]

@kburkhart84

But I’m not convinced that animation at an “acceptable” level is easier in 2d than 3d. I can see how the point could be argued though.

The main issue with 3d animation, when it is bad is movement being wooden and robotic.
In 2d it is much easier to avoid this problem.

One of the techniques I saw being used, I believe, by creators of “Robots” is to make a 2d stick figure animated preview of movement, and if it looks right, transfer it to 3d. That’s because 2d lines are more expressive (that’s why gesture drawing is a thing), you are not constrained in any way, and one frame takes few seconds to draw. This absolutely isn’t going to work in 3 dimensions, where you also need to consider alternative angles of view.

[kdgalla]

@kburkhart84

And yes, pixel art as a subset of 2d is indeed far easier in my opinion…especially at the acceptable level.

Yes. One thing I’ve noticed is that NES graphics tend to have 4-frame walk cycles and that looks perfectly fine for those type of games. If you imitate that style you can easily get-away with the same thing. When you try to make high-resolution sprites with the same amount of frames, though, it starts to look very choppy and unfinished, so you need to draw more interpolary frames for the animation to look acceptable. Smaller sprites also can mean far fewer frames to render.

[CodeSlug]

Alright thanks I decided to use Pixel Art 2D and sink some 100 hours or more into it, I really want to make that Axiom Verge clone so I really am just gonna stick to that style.

Will post some work here when I am done with some acceptable amount of work on it to get some feedback

[BIGTIMEMASTER]

If you are practicing art and want feedback, go to polycount forum and post in the 2d art section. That’s a forum just for digital artist. There is some good artist here too, though.

[CodeSlug]

@BIGTIMEMASTER

If you are practicing art and want feedback, go to polycount forum and post in the 2d art section. That’s a forum just for digital artist. There is some good artist here too, though.

Thanks I will

[angrypenguin]

@CodeSlug

I suck at both, in this case should I also get a tablet and pen to draw? or just use low poly?

Are you trying to make the whole thing yourself, or are you willing to use off-the-shelf art to speed things along? I’d look at the Asset Store, OpenGameArt.org, and other places which sell game assets to see how much stuff I could avoid making on my own.

Chances are that you’ll still want some key pieces of bespoke art, ie: stuff made specifically for your game. For instance, you probably want your player character and some key props and environmental features to be unique to your game, and designed specifically for your needs. Most games have a lot of stuff that is pretty common, though, and while I’d certainly do stuff like tweak the colours to match my own look I’d be hesitant as a solo or small team developer to make all of that stuff from scratch.

[CodeSlug]

@angrypenguin

Are you trying to make the whole thing yourself, or are you willing to use off-the-shelf art to speed things along? I’d look at the Asset Store, OpenGameArt.org, and other places which sell game assets to see how much stuff I could avoid making on my own.

Chances are that you’ll still want some key pieces of bespoke art, ie: stuff made specifically for your game. For instance, you probably want your player character and some key props and environmental features to be unique to your game, and designed specifically for your needs. Most games have a lot of stuff that is pretty common, though, and while I’d certainly do stuff like tweak the colours to match my own look I’d be hesitant as a solo or small team developer to make all of that stuff from scratch.

I want to do everything on my own I have a unique art I want a vision in my mind.

I decided on 2D pixel art

[hippocoder]

I started my journey with pixel art, I wholeheartedly recommend it.

[CodeSlug]

@hippocoder

I started my journey with pixel art, I wholeheartedly recommend it.

TY I shall do my best with it

[Billy4184]

@neginfinity

However, it is MUCH harder to produce acceptable animation in 3d than in 2d. As animation in 3d is its own separate discipline.

Additionally, there’s 2d, there’s 3d, and then there’s 2d pixel art. Pixel art is easiest from all three due to the limited choices you have.

Really? I thought having to create a dozen different sprites that look like the same object from different angles was harder than creating a few keyframes in blender. If you have IK bones set up it’s really not hard at all to get a decent result.

Besides, if it’s a humanoid character there are a lot of animations out there to use.

The way I look at 3D you have more of a time expenditure of learning the tools, but it requires less skill and in the long run you can produce content very fast if your workflow is good. 2D is more approachable, but requires more skill and it’s much harder to optimize in the long run.

You should really learn whatever it takes to make the game you want to make though.

[neginfinity]

@Billy4184

Really? I thought having to create a dozen different sprites that look like the same object from different angles was harder than creating a few keyframes in blender. If you have IK bones set up it’s really not hard at all to get a decent result.

A practical example:

Take a photo.
Produce a copy of the photo using digital painting.
Then make a render that produces output identical to the photo.

The timeframes will be vaslty different.

---

Unless you’re making an isometric sprite game, you will not be making different angles. You also have one less dimension to worry about. You also are not constrained by structure of the model in any way. The sky is the limit regarding what you can do. The character can grow wings or second head. Then there’s matter of transformative animations like this:

In case of trying this in 3d, you’ll be in world of pain, because this is something that needs to be planned ahead supported by topology and blend shapes, while in 2d you can just draw it.

In case of pixel art you also have greatly reduced complexity, because limited choices allow you to experiment and find out what looks right.

@Billy4184

The way I look at 3D you have more of a time expenditure of learning the tools, but it requires less skill and in the long run you can produce content very fast if your workflow is good.

I believe this is false. In my experience, working in 3d is both significantly slower and requires more skill than 2d. Making something 3d dimensional look 2d is ridiculously hard. Watch Guilty Gear Xrd Sign GDC video:

https://www.youtube.com/watch?v=yhGjCzxJV3E

However, 3d wins in asset reuse. Displaying something from different angle is easy, changing their clothing is easy and so on, changing camera angle is easy, in-between frames are easy and so on.

But, something that is trivial in 2d is going to be ridiculously hard in 3d and require greater degree of skill and in some case advanced technology.

[neoshaman]

I beg to differ, having good 2d mean you have to master proportion, volume, line and composition in all separate frame. You take care of them only once in 3d. 2D is free but it’s expensive as hell, and very high skill.

But the truth is that the continuing line of 3d vs 2d criss cross along the difficulty level.

1. 3d has more overhead, so for simpler art like straight square and cube, 2d win over easily
2. low rez 8 - 16 bit resolution pixel art with low animation win
3. low rez 16 bits art with a lot of animation, I would say it’s equivalent, think fighting games
4. decent rez 2d medium complexity, 3d win all the time
5. high details 2d character, okay you need to be a master 3d is better, even 2d guilty gear used 3d template, even 2d animation like steamboy use 3d scaffolding …
6. everything above a level of complexity is 3d all the way, cheaper, higher quality, faster

and being good a 3d mean at high level being good at 2d to begin with