What a total and utter mesh... (May 2012)

Even though I had already found out about the many issues surrounding Rigged Mesh, I was being asked on a weekly basis if I would be releasing a rigged mesh body and being a stuborn so and so, I didnt like being defeated by rigging, so I percivered and continued to investigate, slow adding more and more nails in to the coffin.
This  document was written off the back of my final Mesh Investigation Project, which although more complicated than this, effectively came to the same conclusions.
Just to be clear though, the main focus of this was attempting to make a new mesh body system for customers and didnt go outside of that.

_________________________________________ 
 
Introspective look at Mesh bodies, and their realistic application in Second Life.

Since February this year, I have been buried up to my neck in research and development on mesh. And as you will have already seen, I have made significant headways in to mesh and rigged mesh. But as I have previously said, Mesh while being powerful, is unfinished. In fact its got so many drawbacks it can get unreal at times.
The simplest way to put it is in some areas mesh is a godsend. In other areas its a utter waste of time.

For me, I have been a kid in a candy shop working on ideas, and the one that draws my attention the most is mesh bodies. Its been one of the heaviest areas of development I have worked on and perhaps the biggest disappointment of mesh. But its not because there is that much of a problem with the technology. It works fine. Its how that technology has been implemented on SL that is the problem that makes mesh bodies utterly impractical.

I have had a lot of people asking me if I will be supporting “this or that” mesh body, or if I will be doing my own. On this front I have had a lot of thinking time, and the answer I will have to give right now is no, I wont be supporting either my own, or others mesh body works on clothing or skins. Not at least until several major issues with mesh have been sorted out which knowing Linden Labs, that wont be any time soon.
But I hear you ask... why?

The question is a good one really. And I will try to explain by taking a human female body, and breaking it down.
The body is effectively a grouping of a number of parts. Head, chest, waist, pelvis, boobs, butt, arms, hands, legs and feet.
The current thinking is the head should be removed. Short version is the stock SL head offers more configuration.
Arms, Chest, legs, waist and Pelvis are pretty generic but can be improved on alot. But to be clear, an arm, is still an arm and as much detail as you can add with mesh, its still an arm.
Areas where the SL AV is let down though is the crotch, the hands and feet, and of course the butt and the boobs. But for the sake of argument, the hands are okay, the crotch is very hard to get right even with mesh, so that leaves us with three areas to improve on – the feet, the butt and the boobs.

So lets take a look at what you can do with Mesh bodies, and with the Regular SL Avatar.

Starting with the SL AV. Starting off, you get avatar physics. Nice. You also get alpha masking so you can hide parts of the body. You get 3 lower clothing layers, 3 upper clothing layers, a jacket, and a Tattoo Layer. All making the avatar very configurable. You also get shape adjustment, where you can alter parts of the avatar to your specification and those specs while limited are far more advanced than you get with mesh, with fine (if sometimes poor) adjustment for multiple areas.
Finally there is also the skin layer. And also with the new viewer tech, you can not only have 7 layers of clothes on, but add multiple same version layers (so 3 shirts for example)
Sum it up... you can wear skin and 7+ layers of clothing, you can have a jiggly butt, you can wear prim parts and hide that part of the AV, and you can fine tune your shape with great detail, although not a great as mesh. But most importantly, these are all out of the box supported elements of Second Life that require NO special treatment.

Moving to mesh. Avatar Physics... no. You cant have that. Alpha masks are gone too. If you want to hide part of the avatar you have to use Invisiprims, or create custom meshes to hide the elements. You can have one upper and one lower clothing layer. No more than that due to Alpha on Alpha clipping issues. You can, assuming you have left the stock SL head do fine adjustment to the face (although even with a 100% skin matched body, there will still be a butt ugly neck seam there), but the main mesh body will only be adjustable in rough factors such as height and width – no fine tuning. Also, if are looking at improving the breasts the shoulder width causes issues with the breasts shape (I will come back to this). Finally, while you can use an SL skin on the body, much as with the clothes you will need an applier system to apply those textures to the body.

So summing this up, you get limited shape, a poor neck seam joint, no physics, no alpha masking so you can wear other prim parts, clothes and skins are limited and have to be built through appliers.

Already, you can start to see mesh is looking like a horror story. And lets just be clear here.
This is so you can improve fractionally the design of the arms, chest, legs, waist and pelvis and also improve reasonable amount on the design of the butt and the boobs.
But honestly... considering what you are losing compared to what little you gain, is it too far?

Well, it is a matter before the public, but of the people I have talked to, it seems that many have already said no. They like the flexibility of the original avatar, and the almost gimmick nature of mesh bodies wore off quickly.

So far I have focused on the technical requirements for mesh, but believe me, this story isn’t over. It does get worse.

Creating rigged mesh is a lovely idea until you realise that one of the features that makes it have a bonus, also creates an error on heavily busty avatars. Now at first this may sound very counter intuitive, but sorry – I am a perfectionist.
Breasts are supposed to be round and bouncy and when you design them in mesh, that is what you do but you also forget that they are going to be affected by the rough shape changes you make. I wont bore you to death with the hows and why, but the short version is, you actually have to keep the rough dimensions from the original shape of the mesh model. If you decide to go mega wide, or super thin, the breasts behave as the avatar does and they go thick and wide or thin and tall, as opposed to keeping that lovely round bouncy shape. I have also found similar issues to a lesser degree with the butt.
Then there is texture distortion. Using the original skin, even if that skin is high definition, creates massive issues over the polygon arrangement. This is why implant breasts using the original skin look rubbish. The short version is, on the mesh avatar, too much skin area, and too few pixels on the textures. This also leads to that classic awful nipple area.

Finally on this front there seems to be a whole oxymoron over the Second Life idea of implants.
In Second Life, having implants isn’t just about having them. Its about the whole range of features you get from them too. Not only do you get great looks, but you get growth, clothing, nipple configurations, RLV and so much more. With rigged mesh, what you get is... well, you get seamless breasts. That’s it.

Finally, while not pertaining directly to the mesh body, those that know me, know I like to give my customers the best I can. So that would mean creating a whole range of support products for mesh bodies. Now, genuinely, if something uses a standard SL body UV Maps on the mesh, then shouldn’t it be easy?
Well... no. Most customers see my work as “skins”. But to convert my skins over to an applier format, and to move my clothing over to appliers, would mean creating over 2000 separate appliers.
And on the clothes, thats even less easy as while they may use the same textures, I use multiple clothing layers. So, instantly, those multiple layers mean I would have to merge layers, and create multiple option appliers on the clothes. And then while I could distribute the Body skins easy enough, I would need to do the face layer of other options... but I have no way of distributing the faces.

Suddenly what started out as an easy job turns in to an “oh dear” job. And as much as I have tried to work out how to get around this, I cant short of a manual intervention for each and every customer.

Logisically, it would take me around a month to polish up a range of mesh bodies based on what I have already designed. And to convert the skin and clothes over – my best guess is 2 to 3 months.

Let me just put this logically.
Three to Four months work, solid, day in day out getting appliers built and working on mesh body designs.
And what would the result be. Well, over what I already have, 10 to 15 new products...
No... thats wrong. It would be 10 to 15 fundamentally flawed new products that customers would get bored of in a few weeks.


Conclusions then?
Well, Mesh bodies really are a torment for designers like myself, as there is a massive amount of things you can do with them. And for designers personal use, they are really neat as “we” can do what we want with them. But for you as a customer, you want easy methods. Not difficult ones. And making mesh end user friendly is a whole different story.

On the Plus side, with mesh bodies, you get a nicer body shape, better butt and boobs, and proper feet. You do get the option to wear clothes too. But only on a single layer.

For the Negative Side. You loose physics, alpha mask, fine shape adjustment, you are limited to designs that include appliers for skins and clothes, you are blighted by the shape adjustment, you loose multiple clothing layers and tattoo layers, there is still a seam joint on the neck, the nipples look awful, you get texture distortions using the original textures and for the most part, they are nothing more that a short lived item.

So from a logical perspective, what you gain from using a mesh body is certainly nowhere near what you loose. And right now, I will throw an even bigger spanner in this works just to make it utterly clear just how futile I personally feel mesh bodies are.

If you want better feet, then I know for a fact given a week, I could design a perfect set of mesh feet to be used with your avatar. While replacing what you already have, they still would give you the best of both worlds and look good to boot.

If you want better breasts, they why reinvent the wheel? Breast Implants are widely available and give you massive advantages and options, and not only that they look better than Mesh breasts ever could. The only advantage Mesh bodies have is seamless breasts, but considering everything you loose with mesh bodies, is loosing that seam such a horrific prospect?

And sadly, the butt is the butt. But is it really that awful you have to sacrifice all of your options to just have a cute bum? I think personally I would say no.

And finally, as I put this down in to the ground there is a realisation of futility on this front as two things currently in the works could scupper the whole idea.

Firstly, Mesh isn’t a fixed item. Its a flexible long term project. What is applicable this week, may not be so next week. I  know of two or three projects on mesh that could fundamentally alter the whole design ideas of what we are doing with mesh at the moment. So the question is, what is round the corner and will it make mesh bodies as we know them redundant. Would you want to waste 3 months working on something, only to find that the week after you finish, all your work get made redundant?
Finally, there is another project at Linden Labs, hidden away in their projects list that could in theory make most of what we are looking to achieve with mesh bodies, useless. The so called AV2 project is a massive change to what we know at the moment with the second life avatars, both male and female. And yes, its a new, higher quality, more flexible, and better deformer base mesh avatar that is designed to “Replace” the current SL AV. I would only assume it would come with better shape overall, better muscle control, better breast shape, nicer ass, and hell, real feet. The nice thing is, it should work with all SL stock textures and be fully compatible (although I am guessing toes could be an issue).
So if AV2 is on the table, and gives you everything you could want from a mesh avatar, while giving you everything you have with the stock SL AV...



When I started working with the idea of doing a mesh body, I started at the beginning in Rome. I went through, and tried to overcome each piece that was giving me a headache.
But the more I looked, the less I became convinced until the day I ended up back in Rome seeming having gone round in a big circle in my mesh body journey ending up right back where I started, feeling a little wiser, but a lot more disappointed.

The Sentiment behind mesh bodies is solid, and as a designer, I really can make it look good.
But commercially and logistically, considering what customers will loose, the time it will take me, and the looming threat that all the work I do could be made redundant overnight I have to say, I wont go near mesh bodies from a commercial perspective, either my own design, or for others.

But I will do mesh bodies. The work wasn’t a total waste. I will be releasing, based off this work some replacement mesh avatars, but not in the way that you would think. The ones I will be doing will be unique mesh bodies in Macro and Micro designs. They also will strictly be stand alone products. Not a seamless integration with Violet Studios other products. Simply put, while stock replacement avatars are flawed, you cant do a 1 foot tall or 100 feet tall avatar based off the standard SL AV, and that is enough of a benefit to make mesh bodies practical. Although... the idea of seamless breasts is again out of the window. Any mesh avatar I will be doing will come with regular sized breasts and will be able to use vstrings as frankly they will offer a better shape.