I have an animated mesh in the .x format I've loaded with D3DXLoadMeshHierarchyFromX and have an animation controller for it. The mesh has two animations, one for walking and one for throwing where the walk animation.
Is it at all possible to blend the two animations in such a way that both animations can run together with the walk taking priority for frames below the hip while throwing animation takes priority for frames above it? If it is will the effect look convincing therefore worth pursuing? Do game developers typically blend animations in such a way to get all the different animations they wish or do they simply create multiple versions of the same animation, i.e. walking while throwing, standing while throwing, walking without throwing?
You can set High and low priority animation tracks with ID3DXAnimationController::SetTrackPriority. You can then blend between them using ID3DXAnimationController::SetPriorityBlend.
Related
In a 2D scene in Godot, I have a factory belt transporting items. The items are physics objects (with linear damping = 1 so they do not accelerate too much), the belts have an Area2D with a gravity force that transports the items. The belts have a sprite animation showing the belt moving.
I can sync the sprite animation reasonably well to the speed the items reach on the belts by adjusting the AnimatedSprite's speed scale. Any divergence looks odd visually of course, as the items will either be moving faster or slower than the belt.
I am worried that differences in FPS across devices may cause the visual correctness to break.
Should I just stop worrying? Is there a better way to implement belts moving items? I am aware that I can implement the item movement myself instead of relying on the physics engine - but this won't solve the problem of syncing with the belt animation.
I'm using Unity 4.6 to develop a 2D game. I want to know if having a lot of GameObjects in the scene (out of the camera's sight) has a considerable influence on performance.
For example, is it efficient to make an scrollable list of names (like 1000 of them)? (each one is a GameObject and has a text, a button etc.)
I mask them in a specified area (for example 10 of them are visible at the same time).
Thanks in advance!
Depends on whether or not the objects have visible components. If they do, the engine will draw them even if they are 'off-camera'. A game object by itself has a pretty light load - a tile based game could have thousands in memory. You'll want to toggle the visibility of sprites if you plan on drawing a large number to the scene off-camera. This is where a SpriteManager comes in. It'll check to see if the sprite is in the camera's rectangle and disabled sprites that aren't. There is a semi-offical exmaple here that is good if a little complicated:
http://wiki.unity3d.com/index.php?title=SpriteManager
I have 3 nice and puffy clouds I made in Photoshop, all the same size, now I would like to animate them so they appear like they were moving in the background. I want this animation to be the base background in all my scenes (menu,settings, score, game).
I'm using cocos2d, I have setup the menus and the buttons so the work but how do I accomplish this?
I was thinking to add this as a layer, any other suggestions?
Can anyone show me how some code how to make this please?
David H
A simple way to do it is with sine and cosine. Have slighly different parameters (period and amplitude) to ensure that the user doesn't realise (as easily) that they are programmatically animated.
You may also want to play with animating the opacity value. I'm not sure if layers have those, otherwise you'll have to add the clouds to separate nodes or images and applying it to them.
It's hard to be more specific without knowing what the images look like.
The simplest way to animate anything is to add the sprite to the scene, set the position and call something like...
[myClouds runAction:[CCMoveBy actionWithDuration:10 position:CGPointMake(200, 0)]];
This will slide the sprite 200px to the right over 10 seconds. As Srekel suggested, you can play around with some trig functions to get a more natural feel and motion paths, but you'll have to schedule a selector and iteratively reposition the elements.
The more difficult part of your questions is about getting the animation in the background of all scenes. Keep in mind that when you switch scenes, you're unloading one node hierarchy and loading a new one. The background cannot be shared. You CAN, however, duplicate the sprites and animation in all scenes, but when you transition between them there will be a jump.
I have a situation where I have many CALayers which animate in a "turn based" fashion. I animate the position on each of those CALayers, but they have the exact same duration. Once all of those CALayers are finished animating, a new "turn" is initiated and they animate changing positions again.
The whole idea is that with a linear interpolation between positions, and at a constant speed, a turn based transition between state to state looks like a real time animation. This, however, is hard to achieve with many different CALayers.
CAAnimationGroup is used to group together animations on a single CALayer. But I was wondering, is there a simple solution to group animations, which are supposed to have the same durations, on several CALayers together?
Edited to include a reply to Kevin Ballard's question
My problem lies in this. I'm creating animations for each of my CALayers, then putting
those in an NSArray. Once I get the callback that the individual animation has ended, I remove it form the NSArray. Once it's empty, I again create animations for them all.
With more than a few layers, there's a noticeable delay between where all of the animations finished and the new ones start.
I imagine that if I could group all of these animations into a single one, a lot more layers could be animated without a delay between animations. Thereby not ruining the illusions of a contiguous animation.
If you attach animations to multiple CALayers within a single method, they will all commence at (effectively) the same time. I use this approach in a puzzle game with dropping balls, at the end of the animations I attach the next stage of the animation to any ball that needs further animation.
I'm animating upto 60 CALayers at a time and not experiencing any delays between stages of the animation, but I don't cache the animations in an array of any sort, I'm not sure of thge overhead you have there.
My animations are relatively simple, and created and attached to each CALayer on the fly. My sprites are 60px square and use a couple dozen possible images to represent their content.
In some cases there are multiply animations that I can create with different starting times(using beginTime), I bundle them up with a CAAnimationGroup - but you may not be able to precalculate subsequent animations.
If you wrap your animations in a CATransaction, CG will make sure that they all run during the same iteration of the main loop.
[CATransaction begin];
// all your animations
[CATransaction commit];
If you add 3 animations to 3 different layers all at the same time, and they have the same duration, I would expect them all to animate together. What behavior are you seeing?
cp21yos: can you elaborate on your method? I am trying to do something similar, which involves animating several layers at a time, more than once. You said: "at the end of the animations I attach the next stage of the animation ". Can you explain that? When I try to put logic to perform additional animations in an animationDidStop event, only the last animation is occuring, instead of the whole sequence of animations.
What are the different ways to handle animations for 3D games?
Do you somehow program the animations in the modeler, and thus the file, than read and implement them in the game, or do you create animation functions to animate your still vectors?
Where are some good tutorials for the programming side of this? Google only gave me the modeler's side.
In production environments, animators use specialized tools such as Autodesk's 3DS Max to generate keyframed animations of 3D models. For each animation for each model, the animator constructs a number of poses for the model called the keyframes, which are then exported out into the game's data format.
The game then loads these keyframes, and to animate the model at a particular time, it picks the two nearest keyframes and interpolates between them to give a smooth animation, even with a small number of keyframes.
Animated models are typically constructed with a bone hierarchy. There is a root bone, which controls the model's location and orientation within the game world. All of the other bones are defined relative to some parent bone to create a tree. Each vertex of the model is tied to a particular bone, so the model can be controlled with a much smaller number of parameters: the relative positions, orientations, and scales of each bone.
Smooth skinning is a technique used to improve the quality of animations. With smooth skinning, a vertex is not tied to just one bone, but it can be tied to multiple bones (usually a hard limit, such as 4, is set; vertices rarely need more than 3) with corresponding weights. This makes the animator's job harder, since he has to do a lot more work with vertices near joints, but the result is a much smoother animation with less distortion around the joints.
Alternatively, some games use procedural animation, whereby the animation data is constructed at runtime. The bone positions and orientations are computed according to some algorithm, such as inverse kinematics or ragdoll physics. Other options are of course available, but they must be coded by programmers.
Instead of procedurally animating the bones and using forward kinematics to determine the locations of all of the model's vertices, another option is to just procedurally generate the position of each vertex on its own. This allows for more complex animations that aren't bound by bone hierarchies, but of course it's much more complicated and much harder to do well.
Different ways to handle animation in games?
For characters, typically:
Skeletal Deformation
Blend shapes
Sometimes even fully animated meshes (LA. Noire did that)
In the recent Tomb Raider we use a compute job to simulate thousands of splines for the hair, and the vertices are controlled by these splines with a vertex shader.
Often these are driven by key-frame animation which are interpolated at runtime, and sometimes we add code to drive the influences (bones, blendshape weights, etc.) procedurally.
For some simple objects that move rigidly, sometimes the design will have animation control over the whole object without having to deform it. Sometimes simple things like shrubs can be made to move by using something akin to blendshapes where you have a full mesh pose in a few extremes and you kind of blend between those extremes.
As for tutorials... well... Are you interested in writing the math for the animation yourself or are you more interested in getting some animated character in your game and focus on the gameplay?
For just doing the gameplay, Unity is a pretty great platform to play around with.
If you are however interested in understanding the math for animation or games in general, here's what I'd read up on as a primer:
Matrices used for 3d transforms (make sure you understand the commutative, associative and distributive laws, and learn what each row and column represents - they are simpler than they seem)
Quaternions for rotations (less intuitive, but closely coupled to an axis and rotation)
And don't leave out the basics:
Vector Dot Product
Vector Cross Product
Also like 10 years ago I wrote a simple animation library, which is pretty simple, free from most of the more advance concepts so it's not a bad place to look if you want to get a basic idea of how it works: http://animadead.sf.net
Good Luck!