I spent 30 days learning shaders to avoid drawing pixel art

[@ManaCoder-h4y]

To be honest I couldn't tell the difference between the first screen shot and second, besides the tree leaves moving and more trees and water, I had to do a few double takes.

[@mel0ndev]

those are the shaders, my man.

[@ManaCoder-h4y]

@@mel0ndev ohhh i see i didnt know, well it looks great then. when i think of shaders i think of the actual color changing not animating like that.

[@snark567]

@@mel0ndev Try adding procedural detail textures to the grass and the tree leaves, they look very bland.

[@smileychess]

@@ManaCoder-h4y - Half Life: Alyx uses a shader to simulate liquids inside of bottles, which move as you hold them. It's an incredible example of how shaders are way more than most people realize.

[@dontaskiwasbored2008]

Try watching the video next time.

[@man-o-valor]

I recommend using a color palette to make your colors look more cohesive. Hue shifting really levels up flat-looking light colored art. You’d be surprised how much it can be improved by just paintbucketing all the colors to similar ones!

[@mel0ndev]

@@man-o-valor thanks for the tip, some of these colors are based on a palette and then I think I went off the rails at some point. As I move towards improving the art, I will shift back into a more clear palette. Thanks!

[@eio4528]

Color is often ignored and severely misunderstood but subconsciously prominent in all of us. Bad color choices will destroy any good project. Developers, especially solo, need a dose of color theory and how it's applicable. The proper application of color will objectively upgrade even the simplest looking games. There is no question.

[@GregieAndMe]

Love these type of gamedev making of videso. Keep up the good work man.

[@mel0ndev]

thanks so much! appreciate it lots, my man.

[@KuroNK]

I don't know what's your game about tbh, but also, from the distance you have the camera most of the time, you cant really tell any of the effects you added. And I wouldn't say your art is bad, but i feel it lacks something, like nothing it's connected. it doesn't feel like a game world, it feels like a sprite collage.

[@mel0ndev]

thanks for the insight, and I tend to agree with you. Lots of things need to be added in order to make the world feel more complete, it is indeed lacking very much in a lot of areas! Sprite collage is a good way to word it though lol.

[@mahvus5586]

​@@mel0ndev Something that might help your pixel art is: Make the colors more vibrant; the colors are kind of muted and somewhat bland. A small tip for better colors: when creating shadows and highlights, use hue shifting (hue shift towards purple for shadows; hue shift towards yellow for highlights) I hope this helps.

[@gweltazlemartret6760]

​​@@mel0ndevthat disconnection often comes from sprites being flat vertical facing, while the world is flat horizontally facing. Imagine actually wandering in that world from a 3d perspective, the trees are at best 45° angled, at worst completely flattened. 😅

[@raphaelschmitz4416]

@@mel0ndev I disagree that the art itself lacks something, but what @KuroNK said, in a _"I can't quite put my finger on it"_ way, is still valuable feedback. The last sentence is most important: _"it doesn't feel like a game world, it feels like a sprite collage"_ . Let me rephrase that to _"this world feels like it is missing purpose"_ . This is something that's amplified by how "zoomed out" your game view is: If you would "zoom in" to see only about a 5th of what you see now (both horizontally and vertically), it would already be way less apparent. BUT that would be a different game. What is really missing here is LEVEL DESIGN. You don't need that for testing if technical features work, but you do when you're asking for feedback on the graphics. Currently, there is a tree sprite randomly copied to several places on the map, so randomly copied is how it feels. All you need to fix it, really, is the ground rule of most types of design: "Form follows function". Weird to say about plants, maybe, but the point is, think about WHY something would be where it is in the real world. Always try to follow rules, even if you sometimes have to make them up yourself because it's a fantasy world. Some quick fixes using only your current tiles, off the top of my head: 💡 Make a dense forest part. (Real world explanation: An area can have nutrients in the ground that are ideal for this type of tree. Your players don't need to know this rule, but subconsciously it will look right to them.) Maybe even so dense that you can't walk through it (but that has game mechanic implications). 💡 Make a river, 4 tiles or more wide, e.g. on the left side, from top to bottom. With irregularities, of course. Also, make a rule that it has a "river bank" i.e. no trees or colourful plants grow right next to the river. 💡 Make a smaller little creek off of the river, and put a pond at the end of it. Maybe surround the pond with trees, but again, same rule, trees not directly next to it, leave a little space. 💡 If you have such a water place, in the real world animals might come there to drink from it. Always taking the same way, and trampling any budding plants that might try to grow there, including trees. So there could be a path to the pond. Which, looks nice, has an explanation, and also works again as kind of guiding the player there a little bit. 💡 Your colourful plants, why do they grow someplace? Maybe they need a lot of sunlight, so they're not growing next to trees, which would take some away from them. So, make a clearing, leave a border of grass where they can't grow - and then further away from the trees make a patch of those colourful plants. 💡 Maybe, hey, it's a magic world, put 6 trees in a clearing somewhere, in a circle, suspiciously evenly placed. Any gamer will immediately tell you there's obviously something magic happening in the middle of that circle 😄 Magic (or goal-oriented horticulture by cult members) being a perfectly valid in-world explanation for why our usual tree rules don't apply in this case. PS In a more abstract way, you're kinda reducing the number of "elements" on screen. Currently, an "element" is a singular tree, and you have like 50 of them, and none of them is more interesting than the 49 others. By combining your current elements into higher-level elements like river, pond, clearing, magic circle, it now feels more like there are something like 5 elements on screen, each with a very distinct "meaning" from the others... even though they're made up from the same components. Damn, this comment got long. Can you tell I like game development? 😂

[@m-hcv]

This feels like it was meant as a tutorial video, but the comments are filled with advice on how to improve the result. I have to agree: I'd spend some time on a lot of basic features before I'd bother with water shaders and the like. The game looks so barebones that it's really not worth getting distracted by that yet.

[@mel0ndev]

hey! I appreciate the feedback! There is a method to the madness of putting the cart before the horse, so to speak, and it was done intentionally. In a visual medium like youtube, videos just do much better if a game looks better, so I chose to invest early in some easy wins (plus, I needed to learn GLSL for later anyways). Ultimately, videos doing better on the platform is the goal, which I hope translates to a larger audience for the actual game when it is released. And you are right, the game is pretty barebones in this video, but that's kind of why I chose to do this when I did. Too much time on backend development with no real visuals make a video that's very hard to market and make thumbnails for. Hope that makes sense!

[@btwfh97]

Uniformity and theme is something I look for when I buy a game (does the art have a consistent style) and one example that really popped out at me was the edges between land and water. The edges that separate the land from water horizontally are thicker than the vertical edges. Another thing I noticed (that might just be because of the first topic) was that those edge tiles are in different perspectives, to clarify, it's as if the camera is looking at the horizontal edges on a slant but somehow also looking at the vertical edges at a right angle, which bugs the eye because it doesn't abide by the laws of our euclidean world.

[@mel0ndev]

ah, that is something that I did not consider. Tile placements are a little borked rn due to not handling some edge cases, but I did not consider the perspective. Thank you for the insight, I will see if I can remedy that for the next video since I am going to be spending a lot of time in aesprite.

[@btwfh97]

@@mel0ndev Honestly, it looks very good though, I am only just now getting into game development and your videos have been really helpful. Extremely high quality for being new to the scene. Keep at it and I'm sure you will find a spot on KZbin to call your own.

[@aaro1268]

Solid results! Shaders for pixel art tend to be pretty limited, but you can get a lot from downsampling higher resolution shader effects and styling them to suit pixel art rules. I think you'll probably benefit from a bit more tree and terrain variety. Particles are a great touch, and you can make some bee or butterfly particles. A bit of vignette or edge fogglow can help the visuals pop a little bit. Some emissive textures, especially in any caves would look great.

[@mel0ndev]

thanks! I was also thinking butterflies/birds would be good later on, still lots to do in terms of gameplay, but trees and more biome work are all planned for the future. I'm not sold on the art style of the game quite yet, so I am avoiding committing too much time to drawing more sprites, but I do agree with you, the biome is quite empty at this point. Also, solid suggestions on post processiong shaders, I too was thinking vignette, but I don't really have any other ideas, besides maybe bloom when I implement lighting, so id you have other ideas lmk. I would like to set up a post processing shader pass when I have enough ideas.

[@manojtgn]

nice work!

[@mel0ndev]

Thank you, and I agree! I plan to add it in when I get to lighting and the day/night cycle in the next month or two.

[@lowlevelgamedev9330]

yo that's a cool thumbnail 👍 the game is of course also very cool looking good stuff

[@mel0ndev]

no waaaay, big fan of your content! and thanks, means a lot coming from you

[@combatplayer]

idk if anyone else has mentioned it, but personally i would add shadows for a quick visual boost, that might also help tying everything together, but i'm no expert on this artsyle, thats just something i feel is missing. also a pet peeve of mine is pixel graphic games that do sub pixel movement. makes the pixel art feel forced. but that might just be me.

[@mel0ndev]

yes, shadows are a must. The current plan is to add in a lighting system in another month or two, along with the day/night cycle.

[@delphicdescant]

Oh I didn't realize at first this was a Zig project. Subscribing for that alone.

[@mel0ndev]

awesome! yeah, zig is my favorite language by far right now, enjoying making something larger scale with it.

[@danielkruyt9475]

You didn't even try to get Zig compiling into SPIR-V shaders, and just went straight to GLSL? Tisk tisk, the Zig community needs your work on this compilation pathway! :P Another way of getting consistent reflections is to upgrade your game to 3D and use screen-space raymarching (a very simple shader that requires you to render your whole screen to a RenderTexture, and then render that texture to screen with the raymarching shader active on the reflective parts; look up "G buffer" and "deferred shading" if this all sounds unfamiliar and wild), but this truly requires the depth buffer to exist (3D land) so it doesn't fit into your current state super trivially. But you may find it less difficult than it sounds at first. The depth buffer will also alleviate the need for weird hacks to get your player to stop "climbing trees" (i.e. not going behind them). You can then use e.g. the stencil buffer to cut a little hole around the player if you wanna be able to see through trees but only with a limit when he's obscured. That stencil will however interact non-trivially with the screen-space raymarching algo... Shader/gfx programming goes *deep* my friend, I hope you enjoy learning about the topic as much as I have in the past. With respect to your question about aesthetics, I have to admit I think you're landing solidly in "amateur pixel art" land. Not incredibly attractive, and it's going to be difficult to break out of. I think if you're not planning on trying to study up on the theory of traditional visual art (I haven't either hehe so maybe I'm not the right one to be judging), the only way I can imagine of making yourself *stand out* is in the integrative aspect, the creation of juicy muchness of interaction in all components of the game, two examples: 1. A gust of wind plays (audio), the trees all get pushed aside in a great wave (shader), your player character falls over (animation state machine), and all the enemies start arriving from the side of the map the wind came from and they're running faster with the wind behind them (game logic)! 2. The big monster arrives with its enormous club. The club swings, it hits the ground, and a shader makes the earth literally shake around the point it hit. Some monsters/minions get knocked over temporarily and have to get up. These things make good Steam page videos that *stand out* . Anything shy of these kinds of things, I don't see how to *stand out* without just... being a better artist. :P No offense intended, I wouldn't be any better than you at pixel art to be fair, and I don't think your art looks bad, just very normal. Also I think pixel art is harder to enjoy when it's being rendered at 1:1 scale to screen pixels on modern displays. Using nearest neighbour scaling with careful discretization of positions to match perfectly with the screen's pixels, to render game textures larger but retaining maximum smoothness of camera- and game object- movements is a popular and pleasing thing in modern pixel art games that keeps the art on display without weird aliasing artefacts. Terraria is an example of this, as is Magicite. I assume you've also seen this kind of thing in other 2D games.

[@mel0ndev]

Oh yes, absolutely the art sucks at this point, thank for you trying to be nice tho. Avoiding pixel art is simply an angle for the video, I will have to eventually muster up the strength to dedicate a month or so to drawing better sprites. If that doesn't work like I hope I will hire an artist. Thank for the suggestions with regards to alternatives to reflections, unfortunately I think transferring the engine to work with 3D might be more trouble than it's worth, but I will keep it in mind if the engine eventually requires more complex effects. And not to worry! The player passing in front of the trees is an easy fix, I am just simply not sorting the player sprite at the moment because the sprite renderer lacks any kind of animation support, and I just like the walking animation lol. But that is very high on the list of things to be done next! Keep the gfx stuff coming, bro, I love this stuff.

[@CatwaiiYT]

Wow this comment is gold

[@uis246]

Not all OpenGL drivers support SPIR-V

[@tam_tahiti]

Nice one, you mentioned looking at unity water, maybe you already saw it but just in case check jess::codes video about her 2d pixel art water could give you some ideas. I think a lighting system with shadows could give huge overall to your game, some people use normal maps to help shade 2d pixel art, maybe that could be of use to you.

[@megatronusv2215]

Lol I got recommended this video from jess' video on the water shader.

[@mel0ndev]

loool that's so funny because 1) yes, I did see her video and it was extremely helpful and informative and 2) I am planning to use normal maps for some 2d lighting effects. Glad to see someone else mention normal maps for pixel art, I haven't been able to find too much content about it. If you know of any resources for quickly making normal maps out of sprites that would be a big help for when I get around to it.

[@tam_tahiti]

@@mel0ndev There is a tool named Laigter to generate normal maps from sprites. It was presented on game from scratch youtube channel. Maybe that can help you out.

[@samuelbernhardt4030]

Super cool! Also what is the song that starts around 0:15?

[@mel0ndev]

this one! kzbin.info/www/bejne/bqXTdKaHm6iNe80 wanted to put something a little more mainstream mathrocky, but copyright, ya know?

[@samuelbernhardt4030]

@@mel0ndev haha fair. it’s a good one either way

[@Frank_144]

That IS gorgeous

[@fritzytg]

You could also render the whole world last in a single shader with the reflections. You could render the ground to a render-texture, then the tall objects to a texture to be blown around, then the upside down tall objects to another render texture, and combine them all in a single shader that first applies the world texture, then the wind-blown objects, then for every alpha = 0.0 fills in water effects and the reflection render texture.

[@fritzytg]

But you might still want a separate shader pass for each tall/wind-blown item.

[@mel0ndev]

it's a good idea, I'm definitely going to continue playing around with render textures and see what else I can do with them.

[@jacobbb-t7i]

i see zig, i like. i've been building a vampires survivor clone in zig after having jumped around from engine to engine for years and never liking what i've used. zig + raylib has been an absolute joy to work with.

[@mel0ndev]

I could not agree with you more if you paid me. It feels so damn good.

[@DanTurner-x6z]

I've recently been learning shaders myself, to avoid excessive amounts of pixelart-- My big project right now, is a PC98-styled game, and it requires 7 differently scaled copies of each art asset., at a 640x480 resolution. To avoid a ton of manual cleanup, I've switched over to using vectors, with a rasterizer, and a simple shader to handle flood-filling the shapes.

[@mel0ndev]

flood fill is a nice idea, love it.

[@lowlevelgamedev9330]

also to fix texture bleeding you cah either clamp the texture, either add a one pixel padding in the atlas, either use bindless textures and have a different texture per tile

[@mel0ndev]

just watched your video on bindless texture actually, it will help a lot later I am sure. Clamping didn't really work out like I'd hoped, so I added a pixel buffer for now, thanks for the good suggestions.

[@ColdBlueLight]

can you adjust the brightness and contrast to help the colors pop more its a little too pastel and things look to blended together to me. otherwise the shader idea is neat but wont it eventually hit a wall where too many draw calls are being made?

[@mel0ndev]

I think you're right about the contrast. I hoping when I get around to lighting it will help alleviate some of the issues. I am going for a pastel feel, but I have also been considering that it might be a bit too pastel. Post processing should help a bit with contrast, as well, but that won't be implemented for a while. I am not too worried about draw calls at the moment, as long as I can batch everything being drawn with a shader together, but I won't know for sure until I write or implement some kind of profiler. Thanks for the idea about contrast.

[@Doomsdayparade]

Progress! Keep it up!

[@mel0ndev]

thanks! appreciate it, man.

[@TheSensei88]

"After 30 days of learning shaders, game got music and we finally killed that cricket"

[@mel0ndev]

lmfaooo, in the full game I'm adding extra crickets sounds now during day cycle

[@TheSensei88]

@@mel0ndev It's the cricket's family looking for revenge

[@ckpioo]

FINALLY someone who is using zig, im familiar with rust and i have been trying to find some videos which use zig for more complex projects rather than the simple api, cli, etc. Thank you for making this video and just a small favour but if you have any creators in mind who also heavily use zig or even rust for their projects please please tell me!

[@mel0ndev]

hey! yeah zig is actually in a really great spot right now, despite not even being a 1.0 release yet. As far as creators go, I'm afraid zig is few and far between, though there is a guy called quantum developer who is writing a voxel game in zig atm, and the primeagen is also writing a game in zig on stream right now. for rust, my favorite is tantan, who is (also) writing a voxel engine. Hope you find one that you like!

[@delphicdescant]

I'm writing a game in zig right now, but I don't make videos about it. That sounds like way too much work. If you're interested in using zig for game dev, there's a great collection of libraries I'm also contributing too called zig-gamedev.

[@SolathPrime]

I actually do this a lot whenever someone asks me about games and game development I tell them to start with the mother of it all aka "OpenGL" and mainly shaders cause if you can get a shading engine running you have like 50% of the graphics done, I usually tell them also to make their own shader IDE more points if they can give it a 2D boxish blender like shader editor

[@mel0ndev]

it was a very useful learning experience for sure, helps get a better grasp on the graphics pipeline

[@bamboozledsoftware]

I tried shaders. I just don't get them yet. Good luck! Good old parralax shadowns and other tricks are fine for me now lol.

[@mel0ndev]

definitely a shift in thinking to go from cpu programs to gpu programs. Biggest thing for me was wrapping my head around how to pass data between the two to get things that I already had cpu side into the gpu. I didn't really go into it in the video, but there was a solid week of banging my head against the keyboard before I finally had the "aha" moment and kind of understood how the basics worked. Keep trying, you will get it eventually, they are not that bad once you get the hang of it!

[@baranjan6969]

On a sidenote what's your desktop window manager called? It looks awesome.

[@Starflight44]

Looks like it's berry

[@mel0ndev]

running the classic Hyprland + waybar combo. Hyprland is awesome, you get some sweet animations out of the box.

[@foxtrot000]

imo awesome is pretty awesome too

[@mathijswy]

​@@foxtrot000 especially if you use a nice compositor like certain picom forks

[@foxtrot000]

@@mathijswy I haven’t used it yet but it looks well.. awesome and I’ll definitely give it a go with void when I get myself a thinkpad in a week!

[@Big_Theft_Auto]

The only bad mistake is the shadows of different elements are of different intensity but they should all be close to the same intensity for all the elements because they are all light by the sun. The green trees and the pink trees look like different things. The pink trees are difficult to see because they look the same intensity as the grass. If i was drawing this, i would copy paste the pink and the green trees next to each other and try to somehow change the shading to match one another. And is good to do it with everything or else it looks "like a collage" as KuroNK said.

[@mel0ndev]

good points, thanks for the insight. I am still experimenting with aesthetics atm as it is still very early in development. A lot of people seem to think the colors aren't quite right, and I think I agree overall. Collage feeling I've pinpointed to lack of shadows in the world I think, makes everything look flat and lifeless. This will be addressed when I get to the lighting system. But first, building and gathering mechanics are needed!

[@Big_Theft_Auto]

In my opinion it's not the colors. Any colors are ok, but more like the style for the game is still undefined, as you said. I would like to give you a little hand but KZbin won't let me write my email adress in here. I don't know how to send files to you. If you have a way to send and receive files I would like to help you out if you want

[@delphicdescant]

Finding creative ways to use programming to avoid doing game art by hand is my favorite part of game dev.

[@mel0ndev]

I definitely agree! I had seen people do some really, really nice wind effects for their games using sprite animations, and I wanted something similar, but less work lol. They're much more talented than I, and it was an excuse to learn some stuff about shaders.

[@RosemaryWilliams49fruits]

The shader looks nice on the water, and I like your overall colors they go together nicely but I think they're lacking a little bit of contrast. I recommend adding either a more green/turquoise color of a different shade to the water and possibly to the green areas as well to not only add some contrast with the color pop, but also maintain unity. Maybe that's not quite the right color choice, probably experimenting is a good idea, but as is everything is roughly the same value so it blends together a little too well and lacks definition. A good rule of thumb is to push shadow colors towards the cooler side (if green shift the hue towards cyan or blue, if an orangey warm red, shift the hue towards a cooler red that's closer to magenta), and for highlights color shift towards warm because the sun gives off a warm sort of yellow look. That makes the art stand out more/feel more dynamic and like it has depth and three dimensionality with very little changes. Your color palette right now is very much on the warm side, with warm greens and pinks, and the range of values on the trees is very nice, so just a small adjustment in the shadow colors (by color shifting towards blue) I think will make a huge difference, while the water just needs a little bit more contrast through adding another value to help it stand out and feel like it has depth (currently the shader you added is difficult to see because of lack of shade variety). I hope this advice makes sense and is helpful to you :)

[@mel0ndev]

thank you for the suggestions! I believe someone else also suggested shifting the colors for highlights and shadows, I will definitely be taking that into account as I continue experimenting with the sprites.

[@すべてに興味を失う]

1:31 It's genius!

[@gizel4376]

thing seems a little too random, depending on what your game is about i would consider having the tree density a little more consistent, with road, or trails and open space, and i hope the default zoom gonna be a little closer

[@mel0ndev]

it's procedurally generated! Tree density will be tweaked as development continues, but as far as roads or trails, those will be up to the player to construct however they see fit

[@ivanstepanovftw]

Which DE are you using? Looks beautiful.

[@mathijswy]

Hyprland (on Arch)

[@mel0ndev]

Hyprland! and the theme is catpuccin if you like the colors, both for vim and for kitty (terminal emulator)

[@Shadowthevampire]

Everytime you said you changed stuff with the water I couldn't really tell a difference either you change it to slightly or its because the camera is super zoomed out I have to squint when I watch your game. I would advice you to have the viewport a bit more zoomed in so the players can enjoy all the effects you have made 😊

[@mel0ndev]

haha the old water was just a solid color! Maybe I didn't make a clear enough distinction between the old water tiles and the experimental shader I teased at the end of my last video. You are also not the first one to mention the viewport, so thanks for the input. I do think while it looks fine on an actual screen, when put onto youtube, it doesn't scale as well, which means that I do need to scale it up when recording/uploading. In game you can zoom in and out to your heart's desire. I'll definitely be more conscious of that when recording clips for the next devlog, so thanks!

[@foxtrot000]

Looks very nice! Never seen a yt gamedev use Linux or Zig. What distro u using btw?

[@mel0ndev]

@@foxtrot000 arch btw. gotta love the aur.

[@foxtrot000]

@@mel0ndev nothing beats the nix repo

[@jeffdunehew]

are we going to be able to zoom out though?

[@mel0ndev]

yes, I didn't do a very good job showing it off in this particular video, but you can control the zoom both in and out quite a good distance

[@jeffdunehew]

@@mel0ndev I couldn't resist sorry

[@mel0ndev]

​@@jeffdunehew😂

[@chillpauly]

There is something you can fix in your process which may fix how the art is going at an overall level. This is the disparity between the level of zoom you are working at when creating assets or effects, compared to the level of zoom the game camera at. You're spending time painting in each pixel in these sprites so they look good to you when you are zoomed all the way in looking at those individual pixels, but when you see the same sprite in game, it's tiny and you can't see any of that. The result is way too small level of detail, way too dense looking sprites, invisible effects, and things looking empty/plain/bland while simultaneously looking over-detailed. When creating visuals for games you always need to keep what it looks like from the game camera as the first thing in your mind, because that's how everyone besides you is going to see them. Generally you can try to think in terms of large shapes, medium shapes, and small shapes. For your game camera's level of zoom, all our brains can really see is the large shapes. However, currently when you're making assets/effects, you're making them almost entirely out of small shapes. So all we see in game is a big blob of indeterminate detail that our brains filter out in favour of the big blocks of colour. I would recommend zooming your game's camera in a bit, and then adjusting the placement of objects to reflect the new zoom (closer together). Then review your sprites from your game view - spread those small details out in favour of large shapes > medium > small. Also review the scale of objects in the game. Compare the scale of everything to your character. The trees are kind of big but not too bad. The flowers are massive. The water/waves are tiny weeny.

[@mel0ndev]

I can see your point, but the camera zoom is adjustable by the player, so there is not going to be a single zoom that for every player. I can say that the scale looks more strange on video than it does when actually running the game, but I think it's a good point to bring up, esp. regarding the lod in the sprites. I will consider this moving forward with defining the aesthetic of the game, thanks for bringing it up!

[@chillpauly]

@@mel0ndev yeah all good about variable zoom. I'd probably then ask, what do you think is the 'intended' zoom where your game is meant to look best? When someone's playing a game they're likely to find a zoom they like best and stick with that 95% of the time. If your game visuals aren't designed to look awesome at that kind of zoom, then it's missing the mark. That being said, I'll rephrase part of my initial comment - you might wanna limit the max level of zoomed-outedness to a lesser value. Find a value that you consider core, then design to that. Large, medium, small still applies no matter what the zoom is. For a great example of this done badly, check out a comparison of the original visuals of Warcraft 3, to the 'Reforged' visuals they released. The originals were well designed to look good at intended game zoom (max zoomed out, 99% of game time), so the proportion of L to M to S shapes is spot on. The Reforged visuals were badly designed, like the artists created the new assets while only looking at them fully zoomed in, so when you play the game, all the textures look bland, the models look noisy, etc. I am stressing this recommendation because I've learnt it the hard way. I've made the mistake myself, and things have gone into a released game that I have then looked back on and realised how bad it is once I played it.

[@faradayawerty]

what is that based pixel art editor ??

[@mel0ndev]

it's called libresprite, a free fork of the very popular aesprite.

[@faradayawerty]

@@mel0ndev thanks

[@magni319]

I don't want to shock you, but the difference between the first one and the second is the amount of trees. No player would notice or care about some twinkling leaves

[@mel0ndev]

maybe you don't, and that's ok. I wanted to learn about shaders and add these effects in and so I did.

[@misterguye]

👍

[@mel0ndev]

👍

[@doce3609]

@@mel0ndev 👍

[@btarg1]

Based programmer refuses to do art Godspeed

[@mel0ndev]

lmao, glad someone appreciates it

[@mehmeh8883]

He uses arch btw

[@mel0ndev]

@@mehmeh8883 btw

[@cadaba9849]

Should def learn to draw ✍️

[@mel0ndev]

oh yes, I will need to at some point!

[@boxhead-zk7sn]

Great vid my advice is can you zoom in

[@mel0ndev]

yep, doing it in the next video! thanks for the feedback

[@Stufftonotruinmysearchhistory]

What OS are you using?

[@mel0ndev]

I am using linux, specifically arch linux.

[@A_I_X]

Nice video!

[@mel0ndev]

Thanks! Much appreciated.

[@g1gabyt3]

add cloud shadows

[@mel0ndev]

this is actually a planned feature in the lighting system, along with a bunch of other stuff

[@matheussylvestre2819]

in fact you are learning art.

[@mel0ndev]

shhhhhh, don't tell anyone tho

[@luccadeahl5340]

:)

[@nanthilrodriguez]

Your priorities are inverted. Braid had no art implemented until the month before its release. Source: The creator of the game Gameplay can be refined and polished without focusing on the art

[@mel0ndev]

as much as I would love to do that, unfortunately this is youtube, where thumbnails and visuals are king. Braid was created in a different time and landscape and under different circumstances. I want the aesthetics in this game to be an iterative process, and the more eyes and feedback I get early on, the better I can shape the way the game looks/feels visually as I progress further into development. I understand your point, though, and it's a good one. However, I am relying mostly on these videos as my sole form of marketing, so the better the game looks earlier on, the easier it is to get people to click. Hope that makes sense, there is a rhyme and reason to doing it 'backwards'.

[@TimelessGamingSoftware]

U made it worse

[@RedBackBite0677]

The shaders dont add anything if the art is still bad, learn pixel art.

[@mel0ndev]

@@RedBackBite0677 yes, that is the joke