Making a Jump You Can Actually Use In Godot 

hi, I have a question, could you modify this code so that you can control the jump height by pressing and holding the jump button? its so that you could make short jumps with tapping the space bar.

@@catotonic1881 You need to apply more gravity on the player when they release the jump button, which will make them fall faster. You could parameterize this in terms of the minimum jump height and the time it takes to reach this minimum jump (both values must be lower than the regular jump height and jump time respectively). You can then calculate early_release_gravity = (min_jump_height - jump_velocity * min_jump_time) / (min_jump_time * min_jump_time)early_release_gravity : float = (min_jump_height - jump_velocity * min_jump_time) / (min_jump_time * min_jump_time) Then, in get_gravity, do something like this: func get_gravity() -> float: if velocity.y > 0.0: return fall_gravity if Input.is_action_pressed("jump"): return jump_gravity return early_release_gravity That should roughly do the trick. If the player taps the jump button for a single frame, they will do the min jump in the min time. If they hold the jump button for the entire jump, they will do the full jump. If they release the button sometime between those two timeframes, their character will start to fall at some height between the min and max. Hope that helps!

@@Pefeper found a different solution, that works well. ass of now I would like to know how to add a coyote time effect as well as having the ability to activate the jump function 0.1 or 0.2 seconds before the character hits the ground. sorry if I'm asking too much I'm new to the whole game dev thing. thank you though, you vids are really helpful.

Hello, I have a question. The difference between global position of the player when standing still and at peak of the jump does not equal to the jump height I set (it is off around 8 pixels). How to fix this

@@Perndoe it's hard to say exactly why this would be happening without seeing your setup, but Godot 4 changes the implementation of character body quite a bit. My guess is that it's an issue converting this code to Godot 4. I'll try to upload an updated version of this script later today on the github gist

Agreed.

Master albert , your videos are also really good

Yes it is 😃

Not because you’re “dull”, it’s because their intention (and most public speeches for that matter) have the unconscious hidden agenda to “look smart” in front of others, thereby making something that could be explained simply explained in a convoluted manner. This video, on the other hand, is actually about teaching you something.

@@pianoatthirty I disagree with you here. If you saw the GDC talk it was very obvious that the speaker had a very limited time allowed to talk. He talked really fast and barely managed to finish before the other guy took his place. Maybe he could have explained things in more depths if there weren't any time restraints.

​@@pianoatthirty Or maybe, and hear me out, it's because the point of the talk was to be a mathematical break down of the subject in as much detail as possible? Don't get me wrong, I'm sure there are plenty of public speakers who are just trying to look like they're smarter than everyone else, but I've seen the talk. It's a good talk. The math is thorough and it's explained well enough, though as the other comment mentioned you can see the influence of time crunch impact the teaching. Both that talk and this video are about teaching the audience, one being the base concept of the subject and the other being the actual real world application respectively.

thank you so much, i couldn't get rid of an error regarding the velocity. Now it works perfectly

Although I find this constructive, I think the first one is a matter of preference. I would much rather see positive values in my exports than having to add that - all the time.

Thanks!!! ♥

so I know that this is 8 months late but how do you actually go about checking if the player is holding the button?

@@noxy539 you could just use if statements like this if Input.is_action_pressed("jump"): # make sure that its NOT is_action_just_pressed() return jump_gravity elif Input.is_action_just_released("jump"): return variable_jump_gravity return variable_jump_gravity

@@noxy539 not sure if you figured it out yet, but instead of checking if they're holding the jump button we should just check if they release the jump button. So we can say "if Input.is_action_just_released():" then run our code. Keep in mind that we need to make sure the player can't spam the jump button to reset their gravity (since every time we release it will return true for the if statement). This means we should keep track of if the player is jumping. When Input.is_action_just_pressed() we can set "is_jumping" (name it whatever) to True. Then when Input.is_action_just_released() we can set that same variable to False. I am a noob but this is how I did it. Hope this helps!

@@Z-713 yeah I figured it out after, I just decided to divide the player velocity by 1.5 if they let go and it gave me a pretfy smooth jump

i think thats because your supposed to set the velocity to jump_velocity and use get_gravity for getting the gravity

@@4per8 probably yeah. didnt have time to check it out but i agree that might has to do with it (im still quite starting to learn gdscript lol)

Do you mean like stalling in the air for a little bit after you reach the peak? You could probably do a check if the player's vertical velocity is close to 0 (which would be the peak) in get_gravity and then return 0 to keep the player steady in the air. Maybe use a timer to control how long they're up there for? I have no idea how this would actually feel to play haha, but the nice thing about this code is that it's flexible enough to experiment with

@@Pefeper Yeah like stalling but I guess it should be smooth, not like hitting a wall. Maybe using variable gravity force? But in the end it would be nice to have intuitive controls like ascend time, hang time, descend time. I've also been experimenting with it, I'll try it out.

@@salarycat Oh, OK, I see what you mean I think. Adding an additional hang_time_gravity variable would probably be the way to go, and returning it as the player's velocity approaches 0 in get_gravity(). You could *probably* calculate it in a similar way to the variables we do here, by defining a height it triggers at and the time it takes for the player to reach the peak and start falling again. The equation should be the same as the other gravity variables.

Thanks a ton for the sub, I'm glad you liked the video! As for the Mario jump, you'll need to apply some kind of downwards force on the player when they release the jump button to make them fall faster. In the get_gravity function, you could check if the jump button has been released, and if velocity.y < 0 (player is moving up), return fall gravity * 2 (or some other multiplier)

@@Pefeper woah didn't expect you to answer so soon, thanks btw :) Btw will this work I have a constant gravity variable?

@@crispodispo You should only apply the gravity multiplier when the player is still moving upwards (velocity.y < 0). That way, the player will lose their upwards momentum more quickly. When they start falling downwards though, you just apply gravity like you would normally. The player won't feel anything besides the usually gravity amount when they fall. If you aren't using a function like get_gravity(), I would recommend implementing one, even if you have constant gravity. It makes doing things like gravity multipliers a lot easier! Hopefully that answered your question!

@@Pefeper ok I'll try this, thanks

It definitely depends a lot from game to game. I think it's good to allow the player some degree of mid air movement, but not so much that they can quickly cancel a jump. What I usually do is treat mid air horizontal velocity change as interpolation from the current velocity to the player's desired velocity. Godot has a built in lerp function that's helpful for this. You would do something like current_hvel.lerp(input_hvel, arbitrary_fraction × delta). Hvel stands for Horizontal velocity. That arbitrary fraction will make the jump feel weightier at lower values. 0 will make it impossible to move in the air. With this method you will infinitely approach the goal velocity, but technically never actually reach it. To solve that, just clamp it at the end to the desired velocity if the difference between the current and desired is small enough. Hopefully I made some sense haha. It's a bit late here, but I wanted to respond as soon as I could.

Good question! It's because the effect of gravity accumulates over time. Think about when an object falls. The longer it falls, the faster it gets. Gravity doesn't move us downwards at a constant rate, but instead it constantly subtracts from our upwards velocity. In our gravity model, we suppose that our gravity variables are represented as units per second. So, every second, our upwards velocity should be lowered by jump_gravity or fall_gravity units. We can do that by just adding gravity * delta to velocity.y. Hopefully that helps!

Same!! I copied and pasted the code, too. Changing the times changes the jump height

Players will definitely be able to tell that they're falling at different speeds with different jumps. In my experience with platformers, every type of jump has the same gravity applied to it, and that tends to feel natural. But, it might be worth experimenting with variable gravities! If you think it feels good, don't feel a need to change it out of convention.

Also of note is that sqrt is an expensive operation, when compared to multiplication. Certainly won't make a difference for one player node, but suppose you tried to control one thousand objects in this way - then it'd start to become significant.

I'm glad you liked the video! For the early fall, you need to apply a more intense gravity when the players velocity is still upwards, but the jump button is no longer being pressed. I don't know a clean way to do this at the moment, but maybe just apply jump velocity * 2 when the jump button is released? (Make sure you only use fall gravity when the player is falling downwards though)

@@Pefeper Aaah, I completely forgot to respond to this, my apologies! Thanks a lot for the tip! I ended up going with something similar and it works like charm, thanks again!

Godot 4 added a "velocity" variable to the CharacterBody class, so you can just remove it from your script

If you're talking about the gravity function, I don't think you would need to change it. All it does is check if the player is moving up or down, the height shouldn't matter

@@Pefeper I meant the onready var fall_gravity(). Since it's using time to descent, I thought the time would need to change if the person double jumps (e.g. double time since the jump height doubled).

@@tonyramirez593 When you jump a second time, just call the jump function again so it sets the player's y velocity back to jump velocity. You shouldn't need to do anything to gravity, that should remain the same regardless of how many times you jump

Godot 4 implements move_and_slide differently than Godot 3 (which is what this tutorial was made for). Now you need to modify the Character Body's velocity variable (instead of creating your own velocity variable in the script). Then, call move_and_slide with no arguments

@@Pefeper Alright, how do I modify the velocity variable and all that to get it correctly, unless it's really simple and I'm just dumb mindedly clueless rn lol.

Are you putting the logic that moves the character in _physics_process?

You should add a "velocity = move and ______(......)" instead of just "move and _____(......)" it self

Tell me if you found this helpfull

It gives me an error saying that it cannot convert a bool into a vector2@@DFM605

I made it myself in Aseprite! You're not the first person to have asked for it, I'll upload it to Google drive and share the link later

I would try starting a timer as soon as the player is no longer on the floor. Then just let the player continue to jump until the timer times out

The player setup is just a KinematicBody with a capsule collider on it and no other settings altered. I would've put the whole project on GitHub, but most of it is just graphics and the single script. I figured a gist would suffice. If you're sure you got it set up correctly, take some screenshots of the code and your player and put them in a comment on the GitHub gist and I'll try to help how I can!

Do you mean Coyote time, where you can jump for a brief period of time after falling off a ledge? If so, just make a timer, start it when the player suddenly stops being on the floor but didn't jump, and let the player jump if they press the button before the timer runs out.

No, it doesn't, but that should be pretty easy to implement

Time to peak should not affect the jump height. Have you tried copying the code from the Github gist to make sure the equations are correct?

@@Pefeper ok good so it was something with my code then. I haven't tried using your entire player template. I just ended up not using this method and going back to the guessing game. I do like where my player physics is going tho rn.

@@Pefeper also thanks for still replying to comments years after the videos release. really nice of you gotta appreciate that.

@@JuhoSprite of course!! Good luck with your project!

Nvm, found the solution. Thank you for the tutorial

I actually made it myself! Let me find a good place to upload it and I'll add a link in the description

@@Pefeper wow didnt expect you to actually reply.... thanks man and very nice video i learn a lot from you man

After 2.5 years I have finally uploaded a copy of it to Google Drive, thanks for reminding me! drive.google.com/file/d/1iJ2LDDIQTHzo6HJT_YGEeLugkr-1_w0Q/view?usp=sharing

You'd want to do everything just the same, except when the player releases the jump button, return a higher gravity value in get_gravity. So, when the player releases the jump button, they start to fall faster, eventually moving downwards at which point you can apply normal fall gravity

@@Pefeper Thank you!

What line was that on? Did you make sure you set all the variables?

@@Pefeper it was on the first line and it turns out I'm a dunce. I instanced the scene into my world to test, but was then trying to change the variables in the main scene instead of the instanced one, so all the variables on my instanced scene were still zero. I'm good now XD thank you for the tutorial!!!

@@zacrenner5121 No problem, mistakes happen! Glad you got it figured out, and thanks for watching

The delta time between physics updates isn't guaranteed to be perfectly consistent, so it's best to still multiply time-dependent variables by delta

@@Pefeper oh, thanks for the fast answer! Did not knew it wasn't perfectly consistent, thanks for the information. Ur videos are great :>

for my game, i did this: direction is either -1, 0, or 1, depending on the state of the input axis for movement in the ready function, set accel_rate to be top_speed / acceleration_time, and the same with decel_rate and deceleration if direction != 0: velocity.x += accel_rate * direction * delta velocity.x = clampf(velocity.x, -top_speed, top_speed) else: if velocity.x > 0: velocity.x -= decel_rate * delta velocity.x = max(velocity.x, 0) elif velocity.x < 0: velocity.x += decel_rate * delta velocity.x = min(velocity.x, 0) move_and_slide()

@@edgarvas2108 tysm bro🙏

@@pipacombate393 ur welcome bro