#SL Mesh Deformer Project Update Wk16

Inara Pey has an article out titled: Parametric Deformer: Qarl updates (2). Qarl Fizz, the former Linden writing the Mesh Deformer code, participates in the Metareality Podcasts. In the Apirl 20 broadcast. Smoke and Mirrors, Qarl talks about progress on the Mesh Deformer, known to some as AKA Parametric Deformer.

SL Mesh Feet

Mesh Feet with Deformer 0.1

The audio recording is over an hour long – 1:23:00.

The audio starts out talking about the coming SL9B celebration. Geez, Drax and I see the reaction to the SL9B announcement similarly. This and the coming Myrylyn Shinn interview I’ll be publishing have me reconsidering my mental state. But… Yay, Tateru Nino. Whatever, if you are considering doing something for SL9B, listening to the section is worthwhile. It runs from about 01:00 to 26:00 minute marks.

Botgirl’s SL9B parody song is at 25:20.

27:20 Discussion on Max Grafs standard sizing for mesh clothes starts.

The part where Qarl talks about the Deformer is at 28:00 and runs to 28:54. It’s just a few words. Basically Qarl has been busy and hasn’t worked on it for a time. He put some time in on it 4/16 and is moving on it. He reiterates that Linden Lab is committed to getting it working. He would be seriously surprised if the Deformer was rejected. In his opinion that just is not going to happen. To get the sense of things you have to hear this section it’s less than a minute.

Standard Sizes

You may or may not be following the discussions on Standard Sizing for Mesh Clothes in the SL Forum. The discussions are about what sizes to use, how to use them, and how the Deformer will affect mesh clothes. A significant thread in the discussion is in: Mesh Clothing – Viewers w/Deformer 0.1 and 0.2 started by Imagin Illyar.

Qarl addresses the sizing issue – 30:00 – I’ll try to convey the ideas rather than paraphrase. To understand some of the problems you need to understand how the avatar works and how the Deformer works. Take a look at: Mesh Deformer 0.2 Update for details. Notice how the feet changed between Deformer 0.1 and 0.2. Distortion and crumpling is much less in 0.2.

Mesh Feet - Deformer 0.2

Also, in: #SL Bare Feet I get into the problems we can expect with standardized sizes verses what we use now. In the interview Qarl is pointing out some of those problems.  In the Mesh Feet – Deformer 0.2 image we can see the avatar shape settings affecting the feet.

Deformer Complcations

For instance: Qarl says a hoodie-sweatshirt is going to react to facial expressions. I’m not sure about that because I don’t know how we rig a mesh to anything other than the skull bone. So, I’m not sure how the jaw can affect the rigged mesh we make. But, he has the code to experiment with and watch. So, he should know.

Texture stretch is a concern too. Back in: #SL Mesh Deformer Worries written in January I discussed the problem of mesh clothes designed for a Dolly Parton avatar and the Ruth avatar. Not only are there fix issues there are texture stretch issues which Qarl is pointing out in the audio.

The take away from this is that we will likely have clothes sizes as a permanent part of Second Life® Mesh Clothes. Wearing a top designed on the base avatar with a Dolly Parton size avatar is going to stretch the texture and look awful. So, we will have different size clothes to reduce texture stretch and mesh distortion.

Gianna Borgnine, Meatareality hostess, gets into why the Deformer was not part of the original mesh project. 32:00. Mesh was considered a solution for building things not clothes.

Charlar Linden has spoken to the issue. The Mesh team saw few people in the beta testing doing clothing. When mesh was released and everyone started using it for clothes and totally surprised the Lindens. In hindsight many of us think the Lindens should have seen that one coming. I see it as another example of how users and programmers get out of touch. Gianna also points out much of the beta was a boys club. They were fascinated with the tech and Gianna was the only woman complaining about the clothes not being pretty.

In Closing

The Mesh Deformer is moving ahead. From outside it appears slow… probably because it is. Obviously I’m on the outside.

The mesh Deformer is not going to resolve all the clothing issues. There will need to be sizes for large and tiny avatars to deal with texture stretch and mesh distortion.

It also means there is very likely going to be a market for custom tailored clothing. Because we will likely be able to flag mesh not to be deformed, we can have custom fit clothes, which is possible right now. I’ve been considering making some for my avatar based on my shape.

If you think people are attached to their shape now, wait until they start investing in custom clothing designed to their shape. I am adjusting my shape now. Recent articles by Penny Paton and the Deformer issues being discussed have me writing a new article on the importance of your shape. Several things I am writing are requiring more research and thought than usual.

11 thoughts on “#SL Mesh Deformer Project Update Wk16

  1. As I’ve said on the Metareality website, I think Qarl is just trying to be nice to the standard sizing people. He doesn’t want the controversy. Although, I will admit here, that not all the standard sizes will become unusable, but I haven’t played with them, so I’m talking from theory. For the most part tho, this is not how you will want to make clothing. How you weight the clothing will be much more important. Again, this is theory based on my observations of the deformer. I’m just trying to help creator save time and money. I have nothing against the standard sizing people. If the deformer were not on the table, I’d be praising them.

    As far as hoodies and such., the weights are connected to bones. So the hoodie head part would be weighted to the head. The head has morphs for facial expressions and such. The head does not have a Jaw. The head would need to have a Jaw bone for it to have a real weighted Jaw. As it is, talking is also controlled by morphs. Morphs take the weights of a bone and twist, bends, and stretches them. We do not have access to the target verticies of morphs. If we did, none of this would be an issue. Plus, even custom mesh avatars would be compatible with all expressions and lip sincing, if we could reassign the morph targets. There are also morphs on the hands, feet, and even the boobs are morphs. I think there are morphs in other spots, but they don’t equate to as big an issue as the stated 1s.

    Market for custom clothing – well, if you are making it for yourself, yeah, I can see that. Making it for others, I seriously doubt. It would be a very non productive way to work. Now that being said, I have been playing around with DazStudio which makes it very easy to retarget clothing onto different shape avatars. Currently, this program won’t export the correct DAE, and no other programs have anything close to it. It is quite likely tho, that if Blender devs are alerted to this feature, then I’m sure they could create something similar. With this, custom clothing would literally be clicks away.

    • Domino and Gaia are in the circle of Blender developers. So, it is likely whatever changes we need will be lead by those two.

      Just as custom clothes are expensive in RL so they will be in SL. Custom clothes are not for everyone. They work now. They will likely work in the future.

      The Deformer does seem to be deforming mesh based on what the avatar mesh does regardless of what the mesh clothes are weighted to. So, expressions will likely affect it.

      I have seen people talking about hacks to connect to bones and parts of the avatar not commonly known.

      • Basically, Blender would have to do almost exactly what Qarl is doing. It seems to me, that this is all that DazStudio is doing. Daz’s can do this with most of their different figures tho, and I’ve actually put Dax clothing on SL avatars inside of Daz. If Blender did this tho, It would be super easy to convert any clothing to any SL character, even full mesh avatars. Tho it will take the creator some time, and it really should cost the customer at least that time that it does take. In this case, I could really see custom clothing being a super hot item. It really could boost the clothing market. Of course, when these custom outfits are uploaded, they’d have to turn the deformer setting off. And, if the customer ever changed their shape, they’d be screwed, but maybe the creator could offer 1 refit per purchase.

        I personally don’t like hacks, actually I kinda hate them. I’ve seen enough of them that I see how it completely makes SL look like a 3rd rate platform. Why? Because the beauty of SL is that we have all these creators working together to give the customer the ultimate virtual character.

        Hacks, destroys all this. Hacks aren’t compatible with other things. If I don’t have that hack, I can’t make stuff for it. The concept turns every merchants into their own world of creation. Lots of merchants love hacks, cause they think those hacks distinguishes them from the rest of every1 else and makes their products more valuable. I don’t want to use tricks to be a successful animator. I only want to allow my customers to have fun and see the beauty of SL, and maybe get into their own creative side, even if that’s just mixing and matching avatar parts. When I see so many of my customers disapointed that everything doesn’t work together, that shows me that hacks aren’t good. If people want features, ask LL to make them. Lobby for it. Show LL how awesome it would be. Also, I’d not try to be the lonecreator of this or that thing, I’d want the be the best 1, or at least try. That I can respect.

        • crap, maybe not custom mesh avatars, cause theirs no way to get the avatar, and if you could, that would destroy the whole point of the permissions and any1 could resell any avatar.

  2. Oh, as for texture stretching, this can be addressed with more verts in the stretching section, and creative weighting. You could also design the UV map to give more resolution in those stretching areas.

  3. “In hindsight many of us think the Lindens should have seen that one coming.”

    Yes and I wonder what they thought they were doing. Giving us and extra way to make rigid shapes for skyboxes? Of course clothes. (when Rosedale announced that mesh was going to ship at SLCC10 (29.50 into his keynote) he said “I would ship mesh if all we could do with it was attach it to avatars” and “attaching” it to an avatar certainly set up the expectation in my mind that it would be used to its full potential as rigged, deforming, conforming, dynamic, clothing …..not that it was going to be rolled out as just a 2012 version of sculpties.

    (it’s as if they roll things out half baked to guarantee themselves years of fruitful employment in the fixing… though that’s probably mistaking lethargy for strategy)

    From what I know about the standard sizing it only involves the possibility of changing 8 morphs channels to conform to 1 of 5 pre determined value sets, (XXS thru L) they don’t even need to be the exact valves just close to them, and if your avatar is close to realistic proportions it should only be needed for you to bump a channel or 3 up or down a few notches. It wont change your look radically unless you have a radically weird shape to start with. There is also no standard sizing for boys that I can find but that’s another story.

    I see the standard sizing as only a stop gap measure in any case. I don’t think you should be buying mesh clothes now and expecting them to last forever, you should be buying them as an “early adopter” and also because, if done well, they look 500% better than the old “painted on” or sculptied stuff…. but be prepared to send them to the trash as the technology improves and changes.


    Having fooled around with Poser and DazStudio myself from time to time and knowing how “fast, easy and fun”, it is to get clothes on your vicky or mike figures, I’m truly astounded at the way the mesh has be rolled out by The Lab. Of course Poser and Daz aren’t 30MB installers…. though perhaps The Lab could add 100MBs of client side features to make life easier for us.. (what’s 100mb is this day and age anyway?)

    • Personally, from the actions that LL has taken over the years, it seems to really just be extremely bad management. I’m no bigtime developer, but every single f-ing time I explain an idea to a coder and they tell me it can’t be done with our current system, I shake my head in frustration. Cause I’ve already mapped out how it can work in my head, I just don’t like to code. There has never, ever been a single time that it was impossilbe to do what I wanted to do. Usually, when the coder catches onto how I’m looking at the problem, they add in their own cool stuff. So, what I’m saying is, I think we got way too many, WE CAN’T DO THAT WITH THIS people in management at LL. Or, that’s too much work.

      To me tho, I’m not as concerned about new features, I”m more concerned that nobody in management at LL seems to care very much about fixing old bug. Old bugs that make LL look very bad. Many aren’t all that complicated either. How can I continue to make things for a platform that continues to pile bug after bug after bug, and then say “What, you mean that’s a problem?”. I like alot of what Rodvick is trying to do, but I really thought he would have some pride in the system and actually fix these bugs that should never have existed in the first place.

      • I know a lot people feel as you do. But, I see bugs and features that were filed before my joining SL being handled now. Some projects are showing more fixes coming out than new bugs and feature requests going in. Also, I never hear the front line Lindens say they can’t or won’t. But, they are given priorities.

        As to what management is doing, I see lots of people voice opinions similar to yours. But, none are involved in the Linden Research management team, which means they do know the company goals, limits, plans, agreements, legal issues, and dozens of other pressures and priorities the managers have to manage. All you can know is they aren’t doing things as you think you would.

        Nor are any of the Lindens omnipotent. I’ve known for some time that the Lindens are not SL users. I think it unreasonable to expect them to work on a project 8 hours a day and have it has their hobby. Whether it is or not, most of the Lindens don’t use SL as we do. For the community to expect them to anticipate the communities every need is unrealistic. If your not getting your issues handled, I’ll suggest you aren’t making your case and enough other people are not chiming in with you. Besides, you are not going to change the Lindens. You can only change what you are doing. You can whine, cry, and complain to all who will listen. But, that is not going to change anything. Only when you figure out how to convey the importance and value of your ideas and present them in a way that is meaningful to the Lindens will you start to see your desires met.

        • I’ll just say this. When I make a new set of character animations, I have to learn the character, front, back, and side to side. I have to learn everything I can about the character before I can really even sit down to create the animations. I don’t see the Lindens doing that. I heard Oz sent emails out asking about internal AOs. Does he wear an AO? Has he ever worn an AO. Should I expect to get a well thought out internal AO to work with, I think not.

          In my opinion, any1 who just wants a 9-5 job, should not be working at any game company. There are lots of people in the world that would bendover backwards to have a job at LL. When I look for people to partner with, the last person I want is a 9-5er. Charlar seemed like some1 that bendover backwards, and LL let him go.

  4. Pingback: Qarl's Mesh Deformer -- latest news - SLUniverse Forums

Leave a Reply

Your email address will not be published. Required fields are marked *