Self-taught indie dev here, too. I want to go beyond greyboxing, your videos helps a lot. Only formally trained in coding, really. It doesn't get you very far, I'd argue art and lighting is much more important. Depends on the game, sure, but for solo indie projects? Let's be real, art is going to be 80% of the project. Plus lots of code is based on the art, so art becomes inevitable. Thanks for teaching me how to make my art not look bad.
Very inspiring. I feel like am you, just about two years behind. Maybe not as much. I done gamedev as a hobby on and off for years. More off than on. And my current day job doesn't leave much wiggle room. My son is just 10 and already running laps around me in Godot. My only advantage is already being a trained dev with a degree. Enough about me, just saying all these things aloud to keep myself engaged. So thanks again for this inspiring video and now you have one more subscription!
You are genius! Hope to see more talented game devs using Godot, so the engine could improve its reputation a bit, because lot of people perceived it mostly as a 2d game engine
Just as a random idea for the worm for next time: I think the best way to approach this would be taking a page of the classic 2D console games playbook and add the individual worm segments in the game, then just let them follow each other. While this might lead to a bit less smooth shape overall (I bet this could be countered with some vertex shader magic), it's probably way easier and better looking than that Blender thing, that failed.😉
Yeah that's a good idea if this ever evolved into a real game and would definitely be more flexible to work with! Unfortunately 'shader magic' is not my strong suit lol
What a great video and great message! I just moved over to Godot from Unity and am never going back, even if the software paid me to make projects in it. Once I got going, the tool and featureset is pure metal compared to Unity's confusing triangle dinging. There are some really cool things in Unity but I can't wait to see bigger games with Godot!
Your ramble was pretty damn profound, from someone who's been making games as a hobby and following the industry since the late 80's. Blaming an unbroken tool for a shoddy job is like blaming an unbroken controller for getting your ass kicked in SFII: a classic excuse for ineptitude. An artisan will create beauty with the most primitive of tools, and a hack will create excrement with the most advanced of tools. quality = time * resources * skill That being said, I'd like to briefly ramble about "realistic" 3D graphics. The only reason indie devs ever had a chance at decently realistic 3D graphics was because, to date, they're pretty bad when compared to reality. We just don't realize it because we're used to it, and the latest and greatest that amazes us, despite looking like shit compared to reality, is better than what we were used to previously and therefore temporarily spectacular. For an indie game with "realistic" graphics, the user would just assume it was an older game. In the beginning graphical leaps were incredible, but it's long been diminishing returns and so the indie vs. AAA studio graphical gap is getting smaller. The reality vs. AAA game graphics gap, however, will never close completely using traditional rendering pipelines and technologies, but will within the next decade when GANs/"AI" are refined and hardware-supported enough that they can render real time, fed by game asset data. That kind of processing power will be server farm driven, with the indies last in line, and more or less indistinguishable from the 100+ million dollar films the GANs were trained on. Since there's nothing any of us can do about the inevitable bizarre future of game graphics, I can only suggest that indies NOT go for photorealism in their graphics. It is art, after all, and most of the time realism in graphic art is not just boring, but defeats the entire purpose. I've said before, "If I want realistic trees, I'll just go outside." Anyway, your video was awesome...had me laughing. Godot is amazing, in execution, in fundamental design, and in the spirit of free software.
Thanks Kevin, salient points all around! Gotta admit I have no idea what AI will do to the studio scene, but as a hobbyist/solodev I don't see it as an issue. Wacom tablets didn't replace paint brushes/pencils for those who preferred them, and the same will likely be true here. Kinda like how CG took over the anime scene years ago, but then once in a while out comes a film like Redline and your reminded nothing touches the finesse and detail of animating entirely by hand. The best tool for me will always be the one I'm enjoying the most (a luxury of being a hobbyist rather than making a living I know), and frankly trying to get an AI to match what I'm actually imagining is usually infuriating. Parts of the process will speed up, but other parts will likely remain manual to actually achieve your vision.
5:00 basically describes why that 3D game I have in mind for years still hasn't materialized. It would most likely look like a giant 3D turd flipflopping around a 3D scene…
While obviously not the same (set in a house), one of my earliest ideas was a horror game, which remains totally unfinished. Problem was, the scope grew and grew, and eventually just became a huge time sink; so I've largely given up on it. Can see an early incarnation here: th-cam.com/video/h-fu9Nnarkc/w-d-xo.html and a bit of an update/play with lighting here: th-cam.com/video/x9NB-he_M68/w-d-xo.html
Proximity fade seems to be very effective in combating the seam problem. But in many new games I dont see it being used. I wonder why. Is it because of performance?
I don't think so, or at least I didn't notice a performance hit. I think for bigger studios they probably employ a selection of other techniques that are subjectively better but take longer to setup (AO and texture blending with object intersections for example, or more advanced water shaders that simulate edge waves/foam etc.). For me though, this is a 'cheap trick' alternative ^^
@@RADkate True that, though for water it's already transparent^^. I personally think a baked AO/texture blend is the way to go for general object intersection, it's just time consuming (could talk about this in another video maybe). This is more a cheap trick that situationally may be helpful (e.g. game jams, water, simple scenes etc.)
I did make sure to align the small waves with the tree/grass swaying and particle sandstorm to hopefully indicate it was windy ^^. In reality it would probably be smaller wavelets though . . . and no giant sand worm.
Very interesting, anyway, I think that the water of a lake should not have many ripples, I would like to know how to make a produral terrain. All the best.-
I am telling you, just like blender to maya... godot will become the main rival to uniity. Only thing godot lacking right now is a good 3d level editor. if godot 4 comes with a even rudimentory level editor, godot will jump to the sky...
Totally agree re. Godot being like Blender but I don't really see the lack of 3D level editor being an issue as you would almost certainly want/need a custom implementation for each use case anyway considering how varied 3D games are. The plugin ecosystem is where I want to see more love but it'll come with time ^^
@@BeauSeymour May be for small game its not a big issue. But Yes its a big issue, manually setup a 3d stage or if you are thinking about a openworld... you absolutly need a level maker. otherwise it would be very hard to do your job. Both unity and unreal has its own stage maker.
A level maker for what? 3D platformer? Space exploration? Large tillable terrain? To me if you want a detailed open world for example, you would be generating it in third party software anyway (e.g. World Machine/Blender) and importing it, or you would run a bespoke generative approach (e.g. Marching Cubes or height map displacement from procedural noise). Then there's the question of how you want to manage materials/shaders? Triplanar mapping or baked? Procedurally or manually placing objects? When you think about it, the scope of "A 3D level editor" is ridiculously broad and varies hugely by use case. Trying to distil this to a single tool is not trivial, and I would argue borderline impossible. These tools are rarely trivially re-usable, as each implementation may need to be entirely custom, such as with space exploration you may want all objects to move relative to the player to avoid float precision issues at scale and Octree LOD Marching Cubes for world generation. Which may be completely unnecessary if you wanted to make a defined region (e.g Skyrim) where you could get away with heightmap tiles or modelled assets. Which may be overkill again when thinking of an endless runner where terrain could be entirely flat, but with procedurally placed objects that unload once behind the player. My point is the needs of a "3D level editor" are literally infinite, and can be significantly different in scope between games. If you watch any GDC postmortem (I'd recommend th-cam.com/video/ToCozpl1sYY/w-d-xo.html), they almost always talk about how their game tooling ended up being bespoke, with the built-in engine tools relegated to prototyping or just where they started initially (like CSGMeshes in Godot), before customising and writing their own code tailored to their needs. Given the broad nature of said tools, it makes far more sense to me to split into specific components and maintain as addons dependent on demand. For example, the Riverways addon, or Spatial Gardener, or the various Voxel/Heightmap plugins. The same is true of Unity, which has literally hundreds of bespoke addons for exactly this reason. Trying to put this all into core is how you turn what is currently THE lightweight 3D engine into bloatware. Hence my statement I think what Godot lacks is a mature Add-on ecosystem with a large variety of such tools to chose from. Bloating core with fringe case tooling is not the way.
Brother, a year has passed, would there be a way for you to pass the project in git? If you can't do it too, it's just that my scene is very similar to yours and I would take yours as inspiration.
That's a beautiful scene by the way.
Self-taught indie dev here, too. I want to go beyond greyboxing, your videos helps a lot. Only formally trained in coding, really. It doesn't get you very far, I'd argue art and lighting is much more important. Depends on the game, sure, but for solo indie projects? Let's be real, art is going to be 80% of the project. Plus lots of code is based on the art, so art becomes inevitable. Thanks for teaching me how to make my art not look bad.
I love the game Dev process, such a difficult journey. I am starting mine with a computer science class in school and am working on my own small games
Love this and good luck with your classes! I wish I'd studied it properly haha ^^
Really appreciate the breaking down of your scene. Great video!
Awesome scene. Great points. Nice video. I enjoyed every second of it. Thanks.
Very inspiring. I feel like am you, just about two years behind. Maybe not as much. I done gamedev as a hobby on and off for years. More off than on. And my current day job doesn't leave much wiggle room. My son is just 10 and already running laps around me in Godot. My only advantage is already being a trained dev with a degree. Enough about me, just saying all these things aloud to keep myself engaged. So thanks again for this inspiring video and now you have one more subscription!
All good, I'm juggling a day job too but no kids! As long as you enjoy it keep at it! It's a rewarding hobby :)
You are genius! Hope to see more talented game devs using Godot, so the engine could improve its reputation a bit, because lot of people perceived it mostly as a 2d game engine
I really like that you makes 3D looks possible, great videos
This is a very helpful and well(!) narrated video. Well done.
thank you a lot for this video. It was really helpful!
Thanks for the feedback ^^
Just as a random idea for the worm for next time: I think the best way to approach this would be taking a page of the classic 2D console games playbook and add the individual worm segments in the game, then just let them follow each other. While this might lead to a bit less smooth shape overall (I bet this could be countered with some vertex shader magic), it's probably way easier and better looking than that Blender thing, that failed.😉
Yeah that's a good idea if this ever evolved into a real game and would definitely be more flexible to work with! Unfortunately 'shader magic' is not my strong suit lol
This tutorial is amazing! Thank you!
Glad it was helpful!
Incredible tutorial -thanks for working on it ^^
Thanks for watching and letting me know ^^
What a great video and great message! I just moved over to Godot from Unity and am never going back, even if the software paid me to make projects in it. Once I got going, the tool and featureset is pure metal compared to Unity's confusing triangle dinging. There are some really cool things in Unity but I can't wait to see bigger games with Godot!
Thanks glad you like it! And good luck with your endeavours in Godot. It's not perfect, but what is, and it's improving at a huge pace!
This was a good video! It really makes me want to make one of these scenes myself
Excelente!!! Gracias por compartir.
*Amazing Graphics.*
Your ramble was pretty damn profound, from someone who's been making games as a hobby and following the industry since the late 80's. Blaming an unbroken tool for a shoddy job is like blaming an unbroken controller for getting your ass kicked in SFII: a classic excuse for ineptitude. An artisan will create beauty with the most primitive of tools, and a hack will create excrement with the most advanced of tools. quality = time * resources * skill
That being said, I'd like to briefly ramble about "realistic" 3D graphics. The only reason indie devs ever had a chance at decently realistic 3D graphics was because, to date, they're pretty bad when compared to reality. We just don't realize it because we're used to it, and the latest and greatest that amazes us, despite looking like shit compared to reality, is better than what we were used to previously and therefore temporarily spectacular. For an indie game with "realistic" graphics, the user would just assume it was an older game. In the beginning graphical leaps were incredible, but it's long been diminishing returns and so the indie vs. AAA studio graphical gap is getting smaller. The reality vs. AAA game graphics gap, however, will never close completely using traditional rendering pipelines and technologies, but will within the next decade when GANs/"AI" are refined and hardware-supported enough that they can render real time, fed by game asset data. That kind of processing power will be server farm driven, with the indies last in line, and more or less indistinguishable from the 100+ million dollar films the GANs were trained on.
Since there's nothing any of us can do about the inevitable bizarre future of game graphics, I can only suggest that indies NOT go for photorealism in their graphics. It is art, after all, and most of the time realism in graphic art is not just boring, but defeats the entire purpose. I've said before, "If I want realistic trees, I'll just go outside."
Anyway, your video was awesome...had me laughing. Godot is amazing, in execution, in fundamental design, and in the spirit of free software.
Thanks Kevin, salient points all around! Gotta admit I have no idea what AI will do to the studio scene, but as a hobbyist/solodev I don't see it as an issue. Wacom tablets didn't replace paint brushes/pencils for those who preferred them, and the same will likely be true here. Kinda like how CG took over the anime scene years ago, but then once in a while out comes a film like Redline and your reminded nothing touches the finesse and detail of animating entirely by hand.
The best tool for me will always be the one I'm enjoying the most (a luxury of being a hobbyist rather than making a living I know), and frankly trying to get an AI to match what I'm actually imagining is usually infuriating. Parts of the process will speed up, but other parts will likely remain manual to actually achieve your vision.
This is a very well made video.
Thanks ^^
5:00 basically describes why that 3D game I have in mind for years still hasn't materialized. It would most likely look like a giant 3D turd flipflopping around a 3D scene…
Can't improve without starting ^^
Beautiful!
True 😊
Can you make a dark mideaval castle in Godot set in a horror setting? I want to see something like return to castle wolfenstien style environment.
While obviously not the same (set in a house), one of my earliest ideas was a horror game, which remains totally unfinished. Problem was, the scope grew and grew, and eventually just became a huge time sink; so I've largely given up on it.
Can see an early incarnation here: th-cam.com/video/h-fu9Nnarkc/w-d-xo.html
and a bit of an update/play with lighting here: th-cam.com/video/x9NB-he_M68/w-d-xo.html
What a cool "sand storm" shader. Can you make video about it or may be link for the source? It's exactly what i needed for my small project
Love it
Proximity fade seems to be very effective in combating the seam problem. But in many new games I dont see it being used. I wonder why. Is it because of performance?
I don't think so, or at least I didn't notice a performance hit. I think for bigger studios they probably employ a selection of other techniques that are subjectively better but take longer to setup (AO and texture blending with object intersections for example, or more advanced water shaders that simulate edge waves/foam etc.).
For me though, this is a 'cheap trick' alternative ^^
most modern engines are deferred so they avoid as much transpareny as they can
@@RADkate True that, though for water it's already transparent^^. I personally think a baked AO/texture blend is the way to go for general object intersection, it's just time consuming (could talk about this in another video maybe). This is more a cheap trick that situationally may be helpful (e.g. game jams, water, simple scenes etc.)
I know how to create a full game now it took me 3 long years to learn now I got it my main engine Godot
good video informative
Looks gorgeous, you've done a great job here. But wouldn't that water be still? 🤔
I did make sure to align the small waves with the tree/grass swaying and particle sandstorm to hopefully indicate it was windy ^^. In reality it would probably be smaller wavelets though . . . and no giant sand worm.
Very interesting, anyway, I think that the water of a lake should not have many ripples, I would like to know how to make a produral terrain. All the best.-
What if I want to add a paint tool for a player to draw lakes or rivers? How can it be made in runtime?
Nice👍
Cool!!!
extremamente lindo!!!
Hello, what is this "Dust storm" year? I can't find such a node.
It's using particles, so effectively It's just lots of mostly transparent flate planes that look like dust and move across the environment repeatedly
Is this the gamdev journey guy?? Voice sounds eerily similar
Hahaha nope! not me ^^
I am telling you, just like blender to maya... godot will become the main rival to uniity. Only thing godot lacking right now is a good 3d level editor. if godot 4 comes with a even rudimentory level editor, godot will jump to the sky...
Totally agree re. Godot being like Blender but I don't really see the lack of 3D level editor being an issue as you would almost certainly want/need a custom implementation for each use case anyway considering how varied 3D games are. The plugin ecosystem is where I want to see more love but it'll come with time ^^
@@BeauSeymour May be for small game its not a big issue. But Yes its a big issue, manually setup a 3d stage or if you are thinking about a openworld... you absolutly need a level maker. otherwise it would be very hard to do your job. Both unity and unreal has its own stage maker.
A level maker for what? 3D platformer? Space exploration? Large tillable terrain? To me if you want a detailed open world for example, you would be generating it in third party software anyway (e.g. World Machine/Blender) and importing it, or you would run a bespoke generative approach (e.g. Marching Cubes or height map displacement from procedural noise). Then there's the question of how you want to manage materials/shaders? Triplanar mapping or baked? Procedurally or manually placing objects? When you think about it, the scope of "A 3D level editor" is ridiculously broad and varies hugely by use case. Trying to distil this to a single tool is not trivial, and I would argue borderline impossible.
These tools are rarely trivially re-usable, as each implementation may need to be entirely custom, such as with space exploration you may want all objects to move relative to the player to avoid float precision issues at scale and Octree LOD Marching Cubes for world generation. Which may be completely unnecessary if you wanted to make a defined region (e.g Skyrim) where you could get away with heightmap tiles or modelled assets. Which may be overkill again when thinking of an endless runner where terrain could be entirely flat, but with procedurally placed objects that unload once behind the player.
My point is the needs of a "3D level editor" are literally infinite, and can be significantly different in scope between games. If you watch any GDC postmortem (I'd recommend th-cam.com/video/ToCozpl1sYY/w-d-xo.html), they almost always talk about how their game tooling ended up being bespoke, with the built-in engine tools relegated to prototyping or just where they started initially (like CSGMeshes in Godot), before customising and writing their own code tailored to their needs.
Given the broad nature of said tools, it makes far more sense to me to split into specific components and maintain as addons dependent on demand. For example, the Riverways addon, or Spatial Gardener, or the various Voxel/Heightmap plugins. The same is true of Unity, which has literally hundreds of bespoke addons for exactly this reason. Trying to put this all into core is how you turn what is currently THE lightweight 3D engine into bloatware. Hence my statement I think what Godot lacks is a mature Add-on ecosystem with a large variety of such tools to chose from. Bloating core with fringe case tooling is not the way.
Brother, a year has passed, would there be a way for you to pass the project in git? If you can't do it too, it's just that my scene is very similar to yours and I would take yours as inspiration.
The video is the inspiration 😅
Hoe did you get the lighting like this?