Above or below ground, for a while at least...

In the past I've mentioned that nowadays with alpha-based shoes I seem to hover above the ground or floor--and people have helpfully pointed me at the "avatar offset" knob that many third-party viewers provide to correct that and other similar issues, especially for tiny/giant avatars.

Nalates Urriah writes about how SSA (server-side avatar baking) will stop that from working. The new Firestorm doesn't have that knob--because it won't work on sims with server-side baking.

If these issues concern you, please do read Ms. Urriah's blog post, follow the JIRA entries she mentions, and, if you have the time, attend the Server Beta Meeting she mentions.

Keep DRM out of HTML

W3C, the group that develops web standards, has a plan for "Encrypted Media Extensions" that I hope you will join me in opposing. Please sign the petition in opposition to EME at the Defective by Design web site.

For more information, read the Free Culture Foundation's blog post "Don't let the myths fool you: the W3C's plan for DRM in HTML5 is a betrayal to all Web users."

Dipping a toe into Cloud Party

I finally spent a few minutes in Cloud Party yesterday. They moved away from requiring a Facebook login to create a Cloud Party identity, so now, thanks to Google+'s more enlightened policy, there's Melissa Yeuxdoux on Cloud Party. (Now, if Cloud Party would stop nudging me about Facebook each time I sign on...)

I went through the tutorials--zoom back from the wall they ask you to build, to avoid hassle when you extend it one dimension or another--and there I was. (Now, how to delete the wall I built?)

I spent a little time tweaking my avatar--and have downloaded the Avatar Construction Kit. What a difference from Second Life!

Using Second Life height-related sliders is like trying to fly a helicopter. There are multiple controls that interact in perverse ways. Lengthen one part of your body and other parts shrink.

Cloud Party isn't like that. I wouldn't have thought  that the simple act of lengthening my neck and seeing that the sole result was that my neck got longer would give me such a feeling of... OK, I have to use a word that I have come to despise... empowerment, but it did. I have to suspect that one of the design goals for Second Life shape sliders was being difficult to use!

The Avatar Construction Kit requires familiarity with a 3D program like Blender, so it's going to take some time, a resource I don't seem to have much of these days, but eventually, I'll get there.

Mont St. Michel

I saw a photo of Mont St. Michel and had to go take a look for myself... and it is indeed a lovely place. It's a Second Life build based on the actual Mont St. Michel, a 100 hectare rocky island just off the northwestern coast of France (this bit of geographical knowledge provided by Google a few minutes ago). I had to take a photo...

...and link to this, as Debussy may have been inspired in part by the church on the island.

  

UPDATE: ...and to this, as the inspiration is pretty clear for this meditative piece by Mike Oldfield.

Project Materials Viewer

There is an early release Project Materials viewer (i.e. one supporting normal and specular maps) for Second Life. It's not the official release, but something people can use to try the features out.

Normal mapping involves taking information from a high polygon count version of an object and then applying it to a low polygon count version, so that you have the appearance of detail without the overhead.

Specular mapping lets one make things shiny. Without it, the best one can do is to paint on textures that show how an object reflects light for a fixed arrangement of lighting. With it, objects with a specular map reflect the light or lights actually present (if one can say "actually present" about a virtual world!). One has to admire the efforts of SL builders to prepaint highlights on their work, but having them actually generated on the fly helps enormously with verisimilitude, as you can see if you head over to New World Notes for Iris Ophelia's article on the subject with accompanying video and images that I suspect will provoke salivation.

One must note that

• Existing items won't magically become shiny or more detailed; builders will have to create new things to take advantage of the new features.
• They don't apply to avatar skins, alas. (Someday, I hope...)

That said, I am very happy to see the progress being made. Thank you to everyone working on Project Materials, and to Ms. Ophelia for writing about it!

P.S. Normal mapping will make gratuitously high-polygon mesh items even less desirable, because one can achieve the look at a much lower overhead.

P.P.S. Aliasi Stonebender points out in the comments on the NWN post that while stock avatar skins won't (at least initially) be affected by normal and specular mapping, mesh avatars with normal and specular maps probably will be. I wonder whether this will affect the adoption of mesh avatars?

Even if you don't build...

..be an informed consumer. You'll be better off for it, and so will Second Life.

Read the "Topology - the often neglected entity" thread in the SL Mesh forum. It describes the process builders go through to make efficient meshes that work well with animations. Codewarrior Congrejo does an excellent job of explanation, with illustrative examples and images. Note especially this point: just having tons of polygons does not make a garment high quality--in particular, it doesn't mean that a garment will track well as your avatar bends or moves its limbs.

If you are into digital photography, you may have read about the "megapixel race" and how megapixels aren't the sole measure of quality of a camera. In the very same way, and also just as it doesn't make sense to always use high-resolution textures (you've read Penny Patton's article on how huge textures applied to tiny objects needlessly make frame rates plummet, I hope), polygon count isn't the sole measure of quality for a mesh. Over in the JIRA entry for the mesh deformer, people have expressed concern about the issue, vide this from a comment there by Jesseaitui Petion:

I'd like to respond to a comment made a few times here already, about designers making high poly content because they are "inexperienced". That's often not the case at all. Many designers are purposely manipulating their content to appeal to the "inexperienced" consumers. Consumers don't understand "high" and "low poly" in the way you and I might. Many consumers won't touch a low poly model with a 10 ft pole because it looks subpar to high poly models (I'm referring to the clothing market here). Consumers don't understand that these models may lag them. What we see as "unneccesarily high poly", they see as "this content came from a skilled and advanced designer." What we may see as "good modeling, low poly" they see as "jagged outdated content from a subpar designer." One makes you money, the other doesn't...ultimately the fashion crowd will create what sells. On the other hand, there definitely are people who've no idea what they are doing and subdivide the mesh over and over when no change in appearance even occurs. But just know that a lot of the top designers in sl are high polyin' it on purpose, knowing full well what they are doing.

Please, be an informed consumer. Ask sellers how well optimized their mesh products are. Ask reviewers to provide the information that will let you decide--including photos of the model not just in graceful runway poses, but also in poses that show how items work when the body parts they cover move or bend.

UPDATE: I asked over in the SL Mesh Forum how one can be an informed consumer, and Codewarrior Congrejo kindly responded, suggesting the following (I summarize):

• Wireframe (ctrl-alt-R toggles it) is your friend. Get demos if they're offered, wear them, and turn on wireframe. Then you can see whether an object is built efficiently.
• Performance tools show useful information.
• Develop > Show Info > Show Render Info (toggle) will show triangle count ("tris").
• Read reviews.
• Read store ads; some brag about their wares' high polygon counts.

Follow the link and read it straight from Codewarrior, who is a far better explainer than I am.

High FIdelity

....If you want to get the crowds to come around, you've gotta have glorious Technicolor, breathtaking Cinemascope and stereophonic sound. -- "Stereophonic Sound", from Silk Stockings

Hamlet has written about Philip Rosedale's High Fidelity project in NWN. It will involve, at least in part, a "next-generation virtual reality system" and may be based on voxel-based rendering (a voxel is a 3D pixel).

I hope it happens. The Second Life client has the new CHUI user interface changes (which enough people don't like that I've seen posts about turning off client upgrades so that they can avoid it), but the mesh deformer sits, completed, awaiting release. Normal/specular mapping are, Nalates Urriah reports, somehow both "arriv[ing] soon" and "still a ways off", sort of like Schrödinger's Cat. It makes me wonder how much effort is actually being put into improving Second Life (vide "Linden Lab Leaving Second Life Behind, Announces 'Patterns'" and "Linden Lab Strikes Again").

That said, any system ossifies with time. People don't like changes that break existing items that they spend L$ on, even if they are beneficial as a whole. (How many now-useless shoes with invisiprims do you have?)

I fear that serious improvements in virtual worlds will come from places other than Linden Lab, and outside of Second Life. I hope that either I'm wrong, or that High Fidelity will let us do what we do in Second Life, only more efficiently and more beautifully, so that we can continue to have, to borrow a phrase, an engine fit for our proceeding.

Big Texture in Little China

OK, actually it's a wooden cup, but it's the principle (and a chance to allude to a wonderful movie).

Penny Patton does the detective work to find out why a newly-rebuilt Milk & Cream had one spot where frame rates plummeted. Read about it, and her suggestions about how to motivate builders to avoid the problem, in her post "Why Second Life Fails".