Round and round we go... its that Rigged stuff again

I know... I am back on the box over rigging again, but like Tantalus I keep chasing water around, and trying not to die of thirst.

I have been working on the latest update to the cyber department with my new Airtek cyber boots and no matter how you cut it, you still end up chasing your tail.

The problem is, I think I may be a little too set in my ways and far too much of a detail perv. But I love detail and if things keep coming out wrong, how are you meant to get past that?

Now, this really comes back to the old “horses for courses” issue that seems to surround rigged mesh.

Give you an example. I saw a customer in store the other night. She looked pretty clued up and had a nicely assembled avatar, but she was wearing a rigged mesh t-shirt. Problem for me was it was suffering sooo badly from clip through issues. Every time she moved, the mesh warped a little and pieces of skin kept clipping through the shirt.

I am pretty sure that the designer of the shirt didn’t want that, and in fact didn’t even see that when designing the shirt. However, as designers we all love the idea of one size work, and we tend not to look too deeply in to how it works in a broad spectrum way.

Of course it is fair to say we are aware of the problem, but without testing close to 50 shape set ups for the body, we cant see what is really happening on the massive variety of shape files.

Hence, here was a case of one size tries to fit all, but never will.

Going back to the boot though.

The first problem is I am a detail perv, both in the detail and fine adjustment detail. While Blender may be a powerful tool, the fine detail isn’t as detailed as you can get it in SL.

Effectively, Blender is a powerful tool, designed for powerful uses. And while it can do that, there are discrepancies between what you see in Blender and what you see in world. You don’t know something until you use it properly, watching how it sits on your personal shape, or how it works with your AO. And in those cases, I find fine adjustment is needed.

Except, its rigged mesh. You cant do fine adjustment in world – you have to do it in blender.

That is the crux of the problems – you have to almost guess what you need to do from what you see in Second Life, go back to blender, make your guess, then spend L$ uploading those tests back to SL, and pray you get it right.

This guesswork process costs time and effort in a massive way. A tweek on none rigged mesh that would have taken 10 seconds, now takes financial cost and probably the better part of 5 or 10 minutes, if not more if you guessed wrong.

Back on the perving side of the detailing comes my secondary gripe with rigging.

If I am working through static items in world, I can position, tweek, alter and add or remove elements in a short space of time until visually I have what I want. Its a process that can be done on the fly, adjustments can be made over seconds or even days as you field test items.

When it comes to Rigged mesh, your hands are pretty much tied by what you can do.

At conception, you have the item as you think it should look, however what you see in Blender isn’t what you see in SL, even though poly to poly, its exact. The feel is different.

In addition to this is the principles of LOC (Levels of Complexity). If you design a straight boot, it may have 5 individual rigged items. The boot, a sole, heel, cuff, and a tip. Each of those five items needs to be weight painted or if you do like I do, you weight paint the boot and copy that to the other items. However, the more detail parts you add in, the more you need to copy and even adjust each piece in the weighting. Eventually it become nearly unmanageable above 30 items.

The short version is, you cant add the detail in your want without driving yourself bonkers as the more complicated it gets, the more unusual behaviour in world will start to appear, resulting in you spending most of your time chasing rampant polys around trying to get them to weight themselves properly. Of course, you could argue that you could merge the meshes together to synch the weights, but that then forces more problems with editing.

Ultimately, the more detail you add, the worse you end up chasing your tail going back to the beginning.

That begs the question though – and its a serious one really. What advantage does rigging offer?

Well, Rigged mesh has two advantages. Its self adjusting to height, and the joints bend smoothly

Its disadvantages are though, lack of detail, it doesn’t self adjust to width, and you end up with a lack lustre product because the fact it is rigged means you cant do the fine detail work without driving yourself round the twist.

Realistically then, the advantages of using rigging seem to be far outweighed by the disadvantages.

This is now to the degree that I only really see a selected few advantages, which almost could be sorted through using a mixture of static and rigged elements.

Static mesh would give you far more power to detail and fine tune, even though it is mesh.

In a final look at the Airtek Boots, I have to ask the question – what advantage is there to making these in pure rigged mesh? The answer is, none.

Overall this does bring up an interesting point though. For everything it offers, Rigged mesh slowly shows its failings to the point where its applications are very limited.

If you are dealing with a simple, easy to work with item like a mesh body, it will do everything you want, even if its utterly impractical. If you are doing simple boots, it has applications as long as you can keep the detailing under control.

But anything more detailed than basic clothing, boots, etc the rigging applications are extremely limited.

To that end, while static mesh comes with its own series of disadvantages, I will be using this more to create what I need in SL with Rigged being used only where necessary.