A:M Step by Step

[Rodney]

We are embarking on a project here in the New Users Forum to identify, resource, fully illustrate and thoroughly reveal areas of Animation:Master that users of Animation:Master want to see fully explained. The perspective we are seeking is that of the New User. From opening the box on the first day to adding detail to your animated film successful we want to document it all.

 

This is not a new idea and there are many who have contributed to the wealth of knowledge already available. We are greatful for those that have worked so hard to help everyone understand the power and ease of animating with A:M. We hope you'll join in and help with this project.

 

Shortly a mysterious gentleman with the moniker triath5147 will hopefully introduce himself here and we'll get underway. He has volunteered to help get this project moving.

 

While this particular project isn't specifically linked with the 2005 CD Project or other Tutorial works it is hoped to be compatible with them and freely available for all . If we do a good job perhaps it'll find its way onto a future CDs or Hash Inc's Technical Reference. In any case we hope you add it to your storehouse of A:M related knowledge.

 

We encourage new users to add their questions and suggestions to ensure we address each area thoroughly and to make instructions clear, concise and thoroughly document each case with examples that show what success looks like.

 

Whether new or old we encourage you to add your 2 cents to help produce the absolute best possible How to documentation available.

[Rodney]

A good place to start the project might be to lock down the basic but optimal format for documentation.

While the format might evolve the style that best seems to fit at the moment is none other than Hash Inc's own online HTML Technical Reference.

 

Here is an example from Surface Properties

 

Clear and concise with examples right? Very nice.

Now what we'd like to do is take it to the next level.

 

How can we demonstrate the same information in a way that is simple yet easy to remember and useful?

 

The user is really the one that must determine the final answer to that question.

If the example is useful and easy to understand for you that sounds like success.

 

If we could add a project file or two with fully illustrated examples that would help, right? That would be a nice start.

 

If anyone would care to comment or contribute we could use this area (Surface Materials) to refine and set an example to use as we start on our journey.

[vf124]

If we could add a project file or two with fully illustrated examples that would help, right? That would be a nice start.

 

Hi Rodney,

 

Looks to me like you have come up with another "really great ideal" ... in the eight months that I've had A:M ... my biggest problem (thus far) is putting all my small clips together. Not the editing in another program, but getting multiple actions in one chor.

 

The new section on "reuseable actions, etc" is where I'm at in this struggling phase ... it seems most of the A:M pros don't have time to put together a coherent and concise tutorial.

 

I do understand how valuable time is ...

 

Peace

[Rodney]

it seems most of the A:M pros don't have time to put together a coherent and concise tutorial.

 

Not to mention it does take effort to make those coherent and concise tutorial.

It sure isn't easy that's for sure.

 

We'll try to hit some of the highlights rather than just work strickly in chronology.

Reusable motions hit two birds with one stone (such violence!) so those would be great to do.

 

It certainly helps if the person writing the tutorial is familiar with the subject.

That could be a problem with me doing some of the more advanced action write ups.

 

I tend to gravitate toward esoteric and little used features which could be problematic as well. Undocumented and unused features sometimes disappear without people even knowing they exist. So we'll try to tread cautiously in those areas.

[triath5147]

Hey everybody, I am agreeing with everything Rodney has already said, and am just adding my 2 cents.

I approached Rodney for help with this, in the hopes that we could make documentation for A:M that doesn't quite exist. I would like to make these docs, very visually intensive. I don't know about you, but artists act emotionally, and learn more visually, than just reading some boring text, with no examples.

Let's face it, reading some of those Reference pages is like reading the manual that comes with your appliances. The information is all there, but it fails to keep your attention.

I also know that most people learn by doing,... not reading and hopefully the series we create will inspire people to start building from their imaginations with the confidence that only comes from experience, and satisfaction of doing it yourself.

Theres nothing more gratifying than starting with nothing and watching your work come to life, and we're going to give you the nudge you need to make that a reality.

I plan on making the series, based on what I would have liked to have seen when I was just starting out with A:M. Which means alot of detail, alot of pictures, and above all, alot of examples of different settings and what they actually do for you, so it will take some of the guess work outta this magic software, and allow you to really tap into it's power, which eludes many of us.

Lets see some more topics that are really buggin you guys.

We plan on basing the series on the PWS, and going from start to finish, to try and cover everything.

As the project matures hopefully we can get support from Martin, and possibly even develope these tutorial sets on new features before the CD comes out. That way you have a clear doc. on the feature and what it can actually do to help you.

OK thats all for now.

[vf124]

I plan on making the series, based on what I would have liked to have seen when I was just starting out with A:M. Which means alot of detail, alot of pictures, and above all, alot of examples of different settings and what they actually do for you, so it will take some of the guess work outta this magic software, and allow you to really tap into it's power, which eludes many of us.

 

Sounds promising ...

[higginsdj]

Well I don't understand CP Weights, how to use them or modify them.

 

Cheers

[luckbat]

Well I don't understand CP Weights, how to use them or modify them.

 

This page helped me get me started on CP weighting:

http://www.hash.com/am2004/Modeling/CP%20Weights/

[higginsdj]

Thanks,

 

Just found it in Wills Technical Reference as well (I was looking under CP but couldn't find it and it didn't even cross my mind to look under Control Point). I'm going to have to sit down one month and read this book then practise each item...

 

Cheers

[Rodney]

Not to sidetrack too much but it's always good to get the thoughts and direction of the creators of Animation:Master. If there is anything that can be considered 'authoritive' it would be the words that come from those at Hash Inc.

 

Concerning CP Weighting there is some revelation into the history of CP Weights in A:M. The follks at Hash said:

 

------------------------------------------------------------------------------------------

Weighted Control Points

Hash Inc. has resisted embracing weighted CP's for a very long time. There are several reasons for this. The technology is fairly good at softening hard joints especially when the geometry is dense. However, it has serious limitations.

 

Weighted vertices have a tendency to reduce the volume of the joint as it blends. The technology feels vague and uncontrollable to the artist. There are a wide variety of tools out there, each one trying to give the artists some level of control. Some use paint tools to paint the weights. Some have falloff tools and curve editors for the falloff. Some have manual weight editing dialogs. Now that the technology is in AM it would seem to the artist that they could get the weights set correctly if they just had the tool like brand X. This pursuit of weight setting nirvana is, unfortunately, often not achievable. I am sure tools to adjust the weights will be added on as demand warrants it.

 

Here is the crux of the problem:

 

Take the simple, but extreme worst case example of a simple tube like an arm. Consider a single point that is at the outside of the elbow joint. It is equally close to both the bicep and forearm. So most would consider 50% - 50% to be a good choice of weights. Now, an elbow cannot twist, but for simplicity, suppose this is a joint that can. Now consider the worst case, a roll of 180 degrees on the forearm. The bicep hasn't rolled, so according to it, the point should be placed on the outside of the elbow where it was modeled. The forearm however thinks it should be rotated 180 degrees, which would place it at the inside of the elbow. Now comes the blending. The 50/50 that we chose would result in the point being halfway between these two places. That is right at the center of the elbow joint. Oops, this is what causes the collapse. If all of the points around the elbow were subject to the same 50/50 blending, they would all end up at the center. This results in an elbow that has pinched down to nothing. Now let us adjust the weights. Consider our point on the outside of the elbow again. Our only control is to do a linear blend of the inside and the outside result. That means no matter what we set these weights to (even if we had ultimate, perfect control) the resulting answer would have to lie on the line that connects the inside and outside of the elbow. None of these choices are the correct one. So no matter what the tool, or how fancy, or how gee-whiz it is you will not be able to get that point on the elbow to go where you really want. Frustrating.

 

Hash Solution 1 - Version 5.0:

 

Blended bones by using an additional bone and constraints on the other hand, set at 50/50 would come up with a new true transform that is only rotated 90 degrees on the roll axis. Now this point on the outside of the elbow is rotated to the front or the back of the elbow by 90 degrees. It is just as far from the center of the arm as it was modeled. Hence, no collapsing. Nice preservation of the volume of the arm. The only problem is it is rather technical, and prohibitively more and more complex to setup if the geometry becomes quite dense. Since it then requires many bones each with different weights. Fortunately our geometry is usually not that dense, so this technology got us by for a long time.

 

Hash Solution 2 - Version 7.1:

 

Smartskin. This is much more hands on and visual. The artist gets to remodel to specify exactly what he/she wants at a given angle. We can model to give the same results as solution 1 above. Or we have the flexibility to sculpt anything we want, like bulging muscles or whatever. The downfall of this solution, is that is tedious and adds to the initial model creation time. Additionally, in version 7.1 to 8.5 the artist was restricted to creating keys only along the primary three axes. Version 9.0 improves on this by allowing arbitrary rotations to have keyframes set.

 

Hash Solution 3 - Version 9.0:

 

Finally give in. Now that weighted vertices are supported directly in hardware by many video cards, it is hard to keep saying no to the gamers. So, do an initial release allowing weighted vertices, primarily to give the game developers a way to still build models in AM, and get reasonable joints fairly quickly and simply. Add tools for more control on a by demand basis in future releases. Smartskin can be used in conjunction with weighted vertices, it requires that you remodel out the natural collapsing that occurs with weighted vertices. For characters that demand perfection it may be just as easy to smartskin without using weighted vertices first.

 

I hope this give you some hints into the decision making process revolving around this technology.

 

--------------------------------------------------------------------------------------------

There you have some of the background behind CP weighting in Animation:Master. A careful read will reveal pitfalls to avoid and hint at further development.

 

As Luckbat suggested reading where Hash Inc went with CP Weights after v9.0 shows what they've accomplished with CP Weighting technology despite its limitations. The information provided includes some nice illustrations and project files. The link once again: CP Weights

[Gaijin]

We are embarking on a project here in the New Users Forum to identify, resource, fully illustrate and thoroughly reveal areas of Animation:Master that users of Animation:Master want to see fully explained.  The perspective we are seeking is that of the New User.  From opening the box on the first day to adding detail to your animated film successful we want to document it all.

To me, the largest and most daunting task facing any prospective animator is learning how to model new items properly. I know that the product is Animation Master, and an emphasis on scripting, storyboarding, and animation mechanics is all well and good, but I doubt that anyone buys into the product with the expectation that it comes with all of the objects necessary to tell their particular story idea - and modeling requires its own particular set of skills.

 

I also rather suspect that the effort required just to learn how to model objects, let alone implementing all of the objects required for a particular story, is a major cause of New Animator Attrition. Goodness knows I virtually fall into this category - I first bought AM in 2002 and have barely touched the program since. One reason was that I was waiting for all of the work on hair to settle down (upgraded to the 2004 CD), but another is a simple lack of time to prioritize all my hobbies with work and home life demands.

 

The task of documenting how to model objects is almost more daunting than actually learning how to model because of the vast differences in requirements each different person brings with their story ideas. For instance, my interests are strictly in the anime field (Japanese style animation) which has a very distinctive style. While creating a base mannequin of an anime character uses most of the same techniques as a photo-realistic one does, implementing the finishing details (all the hard stuff! ) is quite different. Scenery and prop modeling is another thing entirely, and all of these major categories really need up-to-date tutorials with working project files (a large majority of the current modeling tutorials available are using old versions of AM).

 

For character models, tutorials are needed for creating the base mannequin. Tutorials, plural, because not only do you need separate tutorials for each basic part of the mannequin (eyes, face, hair, torso, limbs, joints, rigging bones, etc), but you also need separate variations for each major style of animation (photo-realistic, cartoon, anime, fantasy, heroic comics, yada yada yada). Next are tutorials on how to modify the base mannequin to add clothes, shoes, equipment, etc.

 

Duplicate the above for the various major types of animals, plants, insects, etc., and don't forget the major stylistic differences while you're at it.

 

Next, indoor and outdoor scene modeling tutorials. For indoor scenes, making walls, doors, windows, glass, mirrors, tables, chairs, light fixtures, wheeeee the list is endless but can be pared down to several major category types. Outdoor scenes requires modeling terrain, plants, rocks, building exteriors, vehicles, all sorts of fun stuff. Again, these can be broken down into several major categories.

 

I'm sure you can see why a beginner can get extremely discouraged in a real hurry. Having a good library of stock objects can go quite a ways to alleviate this problem, but it doesn't cure it completely. Note that the Hash free object library - as nice as it is - is currently nowhere near the size to be considered a "good library of stock objects".

 

There are other ways around the modeling problem (and the need for so many tutorials), of course, but I have no idea if any are even feasable. For instance, a program that worked in a manner similar to Poser5 which could create basic rigged AM character models by modifying and morphing a set of base mannequins using a slider-control-like user interface would be a godsend. Similarly, being able to import/convert the output of architectural design programs like Punch!'s Home Design Architectural Series 3000/4000 would solve almost all interior set design modeling requirements.

 

I can always dream, right?

[Pengy]

Gaijin: Well spoken, all though I tend not to agree with you. I think like anything else a person would want to conquer it takes practice. While a large library of models would be great for those that only wanted to animate. I think when the time comes to tell your own story being able to create your own characters will go along way to get you there. How much time have you really given to learning how to do it? All the tutorials you'll need to get you going are out there. Once you've mastered these then you adapt,change, over come any obstacles to get your story told. In my opinion alot of people leave 3d because there is no "make dragon" button or "make animation" button. I'm by no means an expert but let me tell you I sure am enjoying the learning process. From concept to modeling,texturing and animating the characters will be from my idea and by george I'll get there some day. Just my 2 cents do with them what you will.

[Rodney]

Thanks for the input everyone.

 

Most of what will be focused on with the A:M Step by Step How to instruction will be on how A:M works. Of course in the process of showing these steps something is usually built and there is generally an end product so by default a tutorial on How to model is always a possibility.

 

I think some of the items on Gaijin's list would be perfect for paid tutorials.

I'm not suggesting a means or distribution route but I think it would most likely be in the fashion of the SIGGRAPH tapes or the Anzovin CDs.

 

I think there is good precidence for specific information to be sold in the form of books, videos and other media.

 

I'd pay for some useful condensed information on such A:M related subjects.

 

Since we are on the subject of Anime, the books that have been making there way into the states over the past year or so on how to draw Anime/Manga illustrate what would be perfect for the delivery of the information. Cost, distribution and number of willing consumers are important factors that might keep such products from seeing the light of day.

[Rodney]

In the process of collecting information we'll be identifying some resources that already exist rather than reaccomplish the information.

 

For instance, while I don't own a copy, information on the Anzovin "Get Started" CD might cover a lot of territory that wouldn't need to be repeated by us.

 

We'll do some juggling and balancing to determine what information is readily available and what isn't. While we won't avoid topics that are covered elsewhere we will take it into consideration whether the information is available and where possible move on to other areas of learning.

[Hutch]

I have to agree with Pengy. To expect to find a tutorial on how to model any and everything is pretty unrealistic. Modeling is an art like painting or drawing. Sure, a tutorial or two on how all the tools work helps to get you started. Seeing something the Cooper tutorial really helps show you how someone puts it all together. But in the end you just have to take the bits of knowledge you aquire and start using them. Sort of like the "give a man a fish and he eats today, teach a man to fish and he eats fish whenever he wants to" or however that goes. Instead of trying to get someone to give you fish every day, you just need to go fishing.