Good day, Sir! The content you're doing is one of the best I've found. Big thanks to you! Keep going! If you ever run out of topics it would be great to see how you do rigging in blender, what tools are you using and stuff like that. I'm also interested in best practices of active weapon animation and blending it with different poses, like, eg rifle with recoil in stance, kneel, prone. Anyway, thanks again! Cheers!
Thank you so much! it means a lot~ I had originally been against doing videos on subjects other than directly Godot, but honestly the longer I've gone the more other people and yourself have asked for specifically some Blender tutorials, so I think I probably will if nothing else just kinda present my method for making creatures or something. I'll add it all to my backlog for when I get back to tutorials though!
Creating a Twitter account? You're making me look sane! I recently started thinking about a little project of my own. An idea so bad that it might loop all the way around to being good again: Reimplementing Euphoria Engine style animated ragdolls in Godot. Why? Just play Max Payne 3 (or any Rockstar game developed with RAGE) and you will know why. Now it might not sound like that bad of an idea, seeing as there are tutorials on UA-cam on how to do it for Unreal Engine. The idea would be to take the theory of those videos and then try to do the same in Godot. Trouble is that I am terrible at coding and just thinking about puts me in a state halfway between cautious excitement and debilitating despair.
Duuuudeee! you should do it! I love the euphoria engine animation, I wish more games would do something like that. I think it's doable with less code than you think, if you take an active ragdoll system like this: ua-cam.com/video/0MHY2TDeMLM/v-deo.html and then on top of them use IK for the animation reference, you could probably get pretty far already, like handle the character animation dynamically, and use the active ragdoll to make it properly physics based on top of that. And besides, it's always a good time to learn some coding xD
@@Bonkahe I love Crigz! I've already watched the tutorial several times. According to the comments, I could get away with even fewer lines of code using Jolt physics, which has functioning spring bones and is also more performant. All it'd need now is a pain variable which would determine whether the ragdoll goes fully limp or tries to mimic the animation. Each hurtbox would have its own pain variable and determine the rigidity for the attached bones. The local value would then add to the overall pain value. Once that reaches 100, the ragdoll goes completely limp. If I add a multiplier based on movement speed, I could have it that if the character is running and gets hit they fall immediately. If both legs are at 100 the ragdoll goes fully limp everywhere except for the arms and head. The arms would also try to grasp areas at 100. The whole thing would maybe become part of a Max Payne 3 style prefab for people wanting to make a similar game. Better come up with a not lawsuit induicing name though...Got it: Maxwell Throe! That ought to keep them off my back.
Lol never underestimate lawyers to make the world worse in general xD Yeah that could work, kind of almost like black limbs from escape from tarkov, that way on a limb by limb basis you could have them fail. I wonder how you could have the grasping system though, I mean of course you could have the usual raycasting, but something that might be interesting for performance side, is have a sphere cast, where it checks all the points on objects nearby that it first encounters, this theoretically should be any corners pointed towards the character, then it saves this information and the limbs just access it and find the nearest point on their "side" of the character, this data is only updated if the ragdoll moves too far away from the last position it checked (like half the radius of the sphere check or something), this way the ragdoll is constantly trying to grasp at any nearby rail or wall corner, but it's not got a million raycasts to determine that. Now your getting me interested in trying it!
@@Bonkahe I think just going with canned animations mixed with a bit of IK would do the trick. Have one generic animation of them grasping the middle of the hurtbox (i.e. the middle of their forearm) per hurtbox and the use IK to readjust the final resting position to size.The IK bones position is set equal to the hit decal's position and the switch from FK to IK is made about 85% through the grasping animation. Would be no different than manipulating the IK bone in Blender when working with a rig that has FK-> IK snapping. No need for raycasts or anything too performance intensive/complicated for my math averse brain. To deal with edge cases: 1. If the player is directly behind the enemy, play a different grasping animation (or just turn it off entirely, it's already an edge case). 2. Only have direct player damage cause grasping (if necessary) 3. Once a wound is created, have there be a timer which prevents new ones from being created. That's pretty much how it works in Max Payne 3. What is often forgotten about Euphoria is that it heavily relies on hundreds of hand crafted animations to work. Honestly though, the IK readjustment is almost optional. If there is anything I have learned while doing 3D animation, it is that anything rather than nothing already gets you 90% of the way there. Don't feel like baking a complex softbody simulation of a hand grasping around an arm? Make a crappy shapekey and call it a day. Any kind of feedback is better than none at all, so just leaving it at the canned FK animation + ragdoll physics would do the trick decently well enough. Ones you consider hand size and the fact that it's near impossible to grasp the inside of your opposite arm, just using one generic hit reaction animation that might slightly miss the precise target should be fine.
I have been a bit swamped with work, but the last bit of this comment reminded me of old halo 1 days, there was only like a couple key frames of flailing, and whenever a body was on the ground but moving it would cycle between them, with a falling/flailing animation for in the air, and it was incredibly solid on selling the feeling, with little to no interpolation. So long winded way of saying I think your probably right, you could get most of the way there with some simple tricks and a bit more of a dynamic way of selecting animations to play.
I'm forget why i make game. , with this video you inspired me! Before on child time i want make same game which i play and get all off this items. Now , i want make this for just art and atmosphere, thanks man
First comment. Holy Smokes, will you do a tutorial on how you made your world in the video with the grass. Was that made in Godot or imported? Looks so good. Also love your tutorial videos. I appreciate them a lot. 🔥 This channels a gem.
Hello there! thank you so much for the kind words~ The terrain was made all in engine with a set of tools I made myself, it takes stamps for various mountains or roads or what have you, and then uses a layer based system to have customizable and modifiable procedural generation on top of that generated terrain. Kinda lets me do procedural generation + hand crafted stuff all in the same system. I will be making a tutorial in the future but we're still a ways out, got a ways to go before I'm really pleased with it enough to be willing to put out the toolset, though when I do it'll also be just an addon that can be downloaded. Thanks for watching and welcome to the community!
It crossed my mind, and my brother brought it up. But at the moment I think the audience isn't really there, twitter while dumb in a lot of ways, has a fairly large number of my subscribers (at least that have reached out to me), as well as a simple method to convey data to them in small bursts, significantly easier than the community posts here. Probably in the future I will have to as twitter descends further, but so far I'll try to keep the number of places I gotta keep track of to a minimum xP
Hey can you by chance point me to some resources for boids implementation in godot? I'm having a hard time finding something for my needs. The number of boids seen here is quite low so i'm wondering are these mesh instance? I'm looking for something in the 2 to 3 digit ballpark so it would need to be optimized, maybe with a particle system? Again, i couldn't really find something useful on this especially for 3D.
Hello there! yeah I probably will have to put together a video, this one is just a couple mesh instances so not a big deal, for something like what your talking about I'd either be going with a compute shader or a custom gpu particle system (which is actually very similar anyways), I wanted mine entirely on cpu so I could swap out their behaviour on the fly if a player attacked them, and also just their small number didn't require anything too complex but yeah. As for resources, depends on what you do, if what your looking for is the simplest implementation while still looking solid, you can hit up this guys video, although it's in 2D the 3D implementation is almost exactly the same: ua-cam.com/video/iGvbb2HlCdQ/v-deo.html If you don't mind C#, this guy goes over it probably best of anyone I know: ua-cam.com/video/6dJlhv3hfQ0/v-deo.html And then of course there's the legendary Sebastian Lague: ua-cam.com/video/bqtqltqcQhw/v-deo.htmlsi=RTqgPJicS01AszXl Thats roughly in order of more to less Godot specific, and less to more complex. I will be touching this subject in a future video though to be sure. PS: 2-3 hundred is actually totally doable with just some gdscript and multimesh btw, really don't need particles or compute shaders unless your doing thousands, and one of the benefits of keeping it in cpu space is you can directly check the collisions using raycast for collision avoidance, just a thought~
Good day, Sir! The content you're doing is one of the best I've found. Big thanks to you! Keep going! If you ever run out of topics it would be great to see how you do rigging in blender, what tools are you using and stuff like that. I'm also interested in best practices of active weapon animation and blending it with different poses, like, eg rifle with recoil in stance, kneel, prone. Anyway, thanks again! Cheers!
Thank you so much! it means a lot~
I had originally been against doing videos on subjects other than directly Godot, but honestly the longer I've gone the more other people and yourself have asked for specifically some Blender tutorials, so I think I probably will if nothing else just kinda present my method for making creatures or something.
I'll add it all to my backlog for when I get back to tutorials though!
Looks great..❤
Creating a Twitter account? You're making me look sane!
I recently started thinking about a little project of my own. An idea so bad that it might loop all the way around to being good again:
Reimplementing Euphoria Engine style animated ragdolls in Godot.
Why? Just play Max Payne 3 (or any Rockstar game developed with RAGE) and you will know why.
Now it might not sound like that bad of an idea, seeing as there are tutorials on UA-cam on how to do it for Unreal Engine. The idea would be to take the theory of those videos and then try to do the same in Godot. Trouble is that I am terrible at coding and just thinking about puts me in a state halfway between cautious excitement and debilitating despair.
Duuuudeee! you should do it! I love the euphoria engine animation, I wish more games would do something like that.
I think it's doable with less code than you think, if you take an active ragdoll system like this: ua-cam.com/video/0MHY2TDeMLM/v-deo.html and then on top of them use IK for the animation reference, you could probably get pretty far already, like handle the character animation dynamically, and use the active ragdoll to make it properly physics based on top of that.
And besides, it's always a good time to learn some coding xD
@@Bonkahe I love Crigz!
I've already watched the tutorial several times. According to the comments, I could get away with even fewer lines of code using Jolt physics, which has functioning spring bones and is also more performant.
All it'd need now is a pain variable which would determine whether the ragdoll goes fully limp or tries to mimic the animation. Each hurtbox would have its own pain variable and determine the rigidity for the attached bones. The local value would then add to the overall pain value. Once that reaches 100, the ragdoll goes completely limp. If I add a multiplier based on movement speed, I could have it that if the character is running and gets hit they fall immediately. If both legs are at 100 the ragdoll goes fully limp everywhere except for the arms and head. The arms would also try to grasp areas at 100.
The whole thing would maybe become part of a Max Payne 3 style prefab for people wanting to make a similar game. Better come up with a not lawsuit induicing name though...Got it: Maxwell Throe! That ought to keep them off my back.
Lol never underestimate lawyers to make the world worse in general xD
Yeah that could work, kind of almost like black limbs from escape from tarkov, that way on a limb by limb basis you could have them fail.
I wonder how you could have the grasping system though, I mean of course you could have the usual raycasting, but something that might be interesting for performance side, is have a sphere cast, where it checks all the points on objects nearby that it first encounters, this theoretically should be any corners pointed towards the character, then it saves this information and the limbs just access it and find the nearest point on their "side" of the character, this data is only updated if the ragdoll moves too far away from the last position it checked (like half the radius of the sphere check or something), this way the ragdoll is constantly trying to grasp at any nearby rail or wall corner, but it's not got a million raycasts to determine that.
Now your getting me interested in trying it!
@@Bonkahe I think just going with canned animations mixed with a bit of IK would do the trick.
Have one generic animation of them grasping the middle of the hurtbox (i.e. the middle of their forearm) per hurtbox and the use IK to readjust the final resting position to size.The IK bones position is set equal to the hit decal's position and the switch from FK to IK is made about 85% through the grasping animation.
Would be no different than manipulating the IK bone in Blender when working with a rig that has FK-> IK snapping. No need for raycasts or anything too performance intensive/complicated for my math averse brain.
To deal with edge cases:
1. If the player is directly behind the enemy, play a different grasping animation (or just turn it off entirely, it's already an edge case).
2. Only have direct player damage cause grasping (if necessary)
3. Once a wound is created, have there be a timer which prevents new ones from being created.
That's pretty much how it works in Max Payne 3. What is often forgotten about Euphoria is that it heavily relies on hundreds of hand crafted animations to work.
Honestly though, the IK readjustment is almost optional. If there is anything I have learned while doing 3D animation, it is that anything rather than nothing already gets you 90% of the way there. Don't feel like baking a complex softbody simulation of a hand grasping around an arm? Make a crappy shapekey and call it a day. Any kind of feedback is better than none at all, so just leaving it at the canned FK animation + ragdoll physics would do the trick decently well enough. Ones you consider hand size and the fact that it's near impossible to grasp the inside of your opposite arm, just using one generic hit reaction animation that might slightly miss the precise target should be fine.
I have been a bit swamped with work, but the last bit of this comment reminded me of old halo 1 days, there was only like a couple key frames of flailing, and whenever a body was on the ground but moving it would cycle between them, with a falling/flailing animation for in the air, and it was incredibly solid on selling the feeling, with little to no interpolation.
So long winded way of saying I think your probably right, you could get most of the way there with some simple tricks and a bit more of a dynamic way of selecting animations to play.
I'm forget why i make game. , with this video you inspired me!
Before on child time i want make same game which i play and get all off this items.
Now , i want make this for just art and atmosphere, thanks man
Thank you! I'm glad I helped in any way, go for it!
First comment. Holy Smokes, will you do a tutorial on how you made your world in the video with the grass. Was that made in Godot or imported? Looks so good. Also love your tutorial videos. I appreciate them a lot. 🔥 This channels a gem.
Hello there! thank you so much for the kind words~
The terrain was made all in engine with a set of tools I made myself, it takes stamps for various mountains or roads or what have you, and then uses a layer based system to have customizable and modifiable procedural generation on top of that generated terrain.
Kinda lets me do procedural generation + hand crafted stuff all in the same system.
I will be making a tutorial in the future but we're still a ways out, got a ways to go before I'm really pleased with it enough to be willing to put out the toolset, though when I do it'll also be just an addon that can be downloaded.
Thanks for watching and welcome to the community!
Cool stuff! Maybe a mastodon account as well? :)
It crossed my mind, and my brother brought it up.
But at the moment I think the audience isn't really there, twitter while dumb in a lot of ways, has a fairly large number of my subscribers (at least that have reached out to me), as well as a simple method to convey data to them in small bursts, significantly easier than the community posts here.
Probably in the future I will have to as twitter descends further, but so far I'll try to keep the number of places I gotta keep track of to a minimum xP
played squad with u when it drops im down keep up the work
yeah for sure, send me a friend request on steam.
@@Bonkahe i already did
Hey can you by chance point me to some resources for boids implementation in godot? I'm having a hard time finding something for my needs. The number of boids seen here is quite low so i'm wondering are these mesh instance? I'm looking for something in the 2 to 3 digit ballpark so it would need to be optimized, maybe with a particle system? Again, i couldn't really find something useful on this especially for 3D.
Hello there! yeah I probably will have to put together a video, this one is just a couple mesh instances so not a big deal, for something like what your talking about I'd either be going with a compute shader or a custom gpu particle system (which is actually very similar anyways), I wanted mine entirely on cpu so I could swap out their behaviour on the fly if a player attacked them, and also just their small number didn't require anything too complex but yeah.
As for resources, depends on what you do, if what your looking for is the simplest implementation while still looking solid, you can hit up this guys video, although it's in 2D the 3D implementation is almost exactly the same: ua-cam.com/video/iGvbb2HlCdQ/v-deo.html
If you don't mind C#, this guy goes over it probably best of anyone I know: ua-cam.com/video/6dJlhv3hfQ0/v-deo.html
And then of course there's the legendary Sebastian Lague: ua-cam.com/video/bqtqltqcQhw/v-deo.htmlsi=RTqgPJicS01AszXl
Thats roughly in order of more to less Godot specific, and less to more complex.
I will be touching this subject in a future video though to be sure.
PS: 2-3 hundred is actually totally doable with just some gdscript and multimesh btw, really don't need particles or compute shaders unless your doing thousands, and one of the benefits of keeping it in cpu space is you can directly check the collisions using raycast for collision avoidance, just a thought~