Custom Animations

Ok, i will just drop the question. I aint perfect, i need help here. Who knows?

What i want is:
1) The ability to select different body parts, and items, etc... Or have a base animation, on which the items are attached. This way i do not have to sprite 101 variations, because i can JUST change the hat, or the suit, or the pants, or the shoes, or the gloves.
2) A way to animate it. This should be compatible with 100 or so instances. Animations i should be able to make varray from jumping, runnning, salto's, etc..

We are talking about an RPG style setup. Preferable, editable using external files. So people can add their own items and stuff without the need to have MMF2. But if that would make it impossible, you can forget it, and we just work with internal objects.

How do i do this? I have had some setups, but i got really lost, or end up having a heck of a slow engine, as i added more "players"...

I've found this problem too.
My general idea was to have 10 or so active objects per character, which of course contain the different layers of clothing which you can change. (Loading the external png files at runtime into frames)
For myself, this turned out to be a slow process as i needed thousands of images... and more ram than sanity allows.
I'm still looking for a better solution, apart from cutting back on my unreasonable ideas.
I'm naively hoping a new 3d sprite object might come out or something... Something that will let you quickly and easily apply the same animation to different objects.

Any one of ClickTeam that can provide some insight on this topic?

Quote:

---

Originally Posted by Patrick

Or have a base animation, on which the items are attached.
2) A way to animate it. This should be compatible with 100 or so instances. Animations i should be able to make varray from jumping, runnning, salto's, etc..

---

The best solution might be a bones system like e. g. the anime studio animation software has one.
So you build a rig (similar to 3d programs) and connect the different body parts to a skeleton.
Then animate the joints, the object will follow them with predefined rotation restrictions and friction.
This works in 3D, in anime studio and so i see no problem to implement such a feature in mmf2.

And try to duplicate that 100 times, working, and still have your game run at 50+fps on a 1GHZ PC.

Im not asking an impossible thing, But maybe it is impossible with MMF2. Altough i love the tool, im yet again faced with the frustrating limitness of its engine. How faster it is then MMF1.5 its still not ready for this kind of thing. That is what i think. Ofcorse i hope to hear ClickTeam say im all wrong, and they come up with a solution.

Yes, that would be a challenge. But I am of good hope.
MMF is strong enough to handle the physics extension and somewhere in the forum is a example of inverse kinematics.
It's just a question of "how to", and clickteam has already become attentive to it.

i don't see why this would be a problem.

fast loop's and fixed values tests can keep Heep's of characters with Heep's of variable objects together easily.

the only problem i can see is positioning them all with only one action spot, but that can be over cum by setting individual objects hot spots relatively...

I'll make an example for you, check the archive forum.

Yes, but by custom animations i meen not using the build-in animation feature of MMF2. Basicly, the animations are done by moving the individual body parts. Like with ragdolls. See you need to keep track of all the positions, and save them, to play them later. Try and do that with 100+ objects.

I'm tired... check the archive.

It is good to see people who are interested in animating with MMF post about it. This is an area that could use some attention.

As we all know, there are many factors involved. The whole process can get rather complex, so we are trying to break it down to essentials and work our way up.

I think we will be able to make MMF more suitable for real-time animation early on via extensions. If users find that useful and take advantage of it, perhaps future versions of MMF will take that into account, as well.

As I may have said before, we should be able to create a recording deck that will allow you to record what objects are doing during runtime, and play them back. I have already done some rudimentary puppeteering using Actives and an array, but that can be tedious and is not very flexible, nor are the data files generated as compact as they should be.

Along with the record/playback, we will be looking at some form of connectivity between objects that will facilitate easy grouping and control to go along with the whole scheme.

Those two features look very promising as 3ee had done a very basic object, (Modifier) that accomplished both to a lesser degree than we are planning to do.

From there, we can look into expansion of the idea and more advanced features perhaps. Our intent is to make this growable and as complete as is technically possible with V2 of MMF. We will be laying the groundwork for future potentials.

The main reason I am posting here about this is to get some feedback and ideas from users. This is kind of new territory for MMF, so your responses and enthusiasm are valuable. Users who are excited about MMF becoming more real-time animation oriented will appreciate this new venture most. I think there are many of you out there who would like to see this happen.

This project would benefit from knowledgeable testers who are familiar with animation and even some with skills in creating graphics for testing with. We are taking applications via PM and will invite users as per their qualifications.

I dont really think puppet animations are necessary, but what would make the MMF2 animation editor about 5,000 times better would just be a simple layers system.

I have to agree. Lets say I have a character and its a 2d platform guy and he's right arm holds objects like weapons, i will have to create the arm as a seperate object because the weapon is between the arm and the body.

The quesiton is, do you want every and any type of real-time animation to be a reality that would free you from eventing every, minute aspect of it as you go along?

In the end, I feel that the questions posed will be answered by getting to square one first. From there, maybe the answers you seek will be more obvious in practice without need for specific solutions or answerers.

Once you get where the Cartoovie project is coming from, you might see that it could possibly serve you in many other ways that you had not yet imagined.

Only time and lots of interaction and innovation will tell.

i dont even mind doing that, i wish i could do it. what is the issue is the speed and duplication of such a thing. so, i vote yes. I defently would want to have MMF support it.