SL Viewer 3.0 and Mesh

N.B. There are lots of better places to find specific details on SL viewer and mesh. If you want more depth, follow the links in this article.

The current SL viewer is now 3.0 (V3), released just within the last few days, I believe. I haven’t tried it yet. The main difference in V3 is support of mesh technology. (More about that below.) I use several SL third-party viewers (TPVs) depending on what I need to do. A full list of certified viewer is available at:

For day to day use, I normally run Phoenix. This is by far the most used third party viewer. Pros: It’s intuitive (especially for long-time users) and has many (many) useful features for builders and just overall usability. It does support SL Viewer 2.x (V2) features such as avatar tattoo and alpha layers. Has some cool Windlight and other environmental controls that are easily accessible. Cons: It does not support certain features such as shared media (i.e., web-on-a-prim) or mesh. I also find on my computer that Phoenix tends to crash more than other viewers.

In anticipation of mesh, the developers of Phoenix have created Firestorm. This is a more V2 compatible viewer that will likely be a more robust alternative to SL’s V3. I’ve only begun to use it, expecting that Phoenix will eventually be phased out.

If I absolutely need a feature of SL V2 (basically shared media), I do use it on occasion. It was pretty much universally despised from the first day (thus the proliferation of third party alternatives), but has improved significantly since then. However it’s still awkward for builders and other “power users” and lacks many of the TPV advantages.

For taking pictures in SL, I use either Imprudence or Kirsten’s Viewer. I’ve found Imprudence to be the most stable of any viewer and it has most of the features included in Phoenix. It is my preferred viewer for taking pictures as the resolution seems to be best. (May just be an issue with anti-aliasing settings on other viewers, but I haven’t found a solution in Phoenix.) Kirsten’s was the first to implement shadow rendering. (Shadows in SL are very cool when well rendered, and other viewers have since implemented the feature, but be aware that the rendering load on your computer is massive and will tend to crash anything less than a very high end graphics card.) It has a number of other settings that are designed especially for photography and machinima (i.e., virtual cinematography), but it is less friendly for day to day use.

The major change in 3.0 is the ability to render mesh objects. As of Tuesday mesh technology (very high resolution modeling used in dedicated computer generated cinema and dedicated game consoles) is now available anywhere in SL. It has been anticipated for the last few years and has been in beta for some time. I’ve been to the beta grid and spoken to people there trying to work with them. Some of their work is stunning.

The consequences of this technology could be vast or minimal, depending on how it’s adopted. Creating meshes is quite a bit more complex than creating sculpts, which are difficult enough. Only serious creators are getting into it at this point. In the beta grid, users were able to create mesh objects in Google Sketchup and export them to a SL compatible file. There seems to be some question now whether that is possible with the full SL deployment. There are other free programs that can be used (most notably Blender), but there is no other “simple” tool for creating meshes as far as I know. That could change.

There has been a lot of excitement about the advent of mesh. Both good and bad. The technology itself renders beautiful complex objects. In fact, there have always been meshes in SL in the form of your avatar body, which is actually a relatively simple mesh. The important thing to know about meshes is that it is made up of many triangles in a kind of web fabric (hence the use of the term “mesh”).

At the risk of oversimplifying the situation, the core difference between a sculpt and a mesh is the number of triangles it uses. When you create a sculpt, it may be a 64×64 pixel sculpt map or smaller. You can apply that to a prim in SL and you can manipulate the size of that prim and nobody really cares how convoluted or big it is. But with a mesh you have a whole set of vectors for those triangles. And there could be thousands or tens of thousands of them in a mesh, and if you make it bigger it takes more triangles. The important point is that the more triangles a mesh uses, the more server resources it commands, and THE MORE IT COSTS. That’s right. SL in its infinite wisdom has determined a formula for assigning a Prim Equivalent (PE) Cost. So you may create a simple mesh that has maybe 1 PE, but if you just make it bigger it can suddenly show 100 PE. I saw meshes in the beta grid that went from maybe 24 to over 2000 PE just by scaling it. When a sim has a total limit of 15K prims, you can see how a bunch of meshes could quickly hit your limit. So mesh developers are motivated to use the smallest possible objects with the least amount of detail.

In a peculiar twist, the way PEs are calculated includes scripting. If you have a simple notecard giver script in a mesh, it will cost more. The mesh developers I’ve talked to rightfully think this is bogus. In any case, the promise of mesh is here, but it remains to be seen when or whether it will be widely adopted.

Leave a Reply

A sample text widget

Etiam pulvinar consectetur dolor sed malesuada. Ut convallis euismod dolor nec pretium. Nunc ut tristique massa.

Nam sodales mi vitae dolor ullamcorper et vulputate enim accumsan. Morbi orci magna, tincidunt vitae molestie nec, molestie at mi. Nulla nulla lorem, suscipit in posuere in, interdum non magna.