In Has it occurred to you what Project Sansar really is? I wrote about the latest Project Sansar® news. Now I see the site Road to VR has a story up about Project Sansar: First Glimpse of Linden Lab’s Next-gen Virtual World, Project Sansar. They suggest it may not be so photorealistic as a stylistic Halo type world.

Sunset Through the Arch at Voile
Second Life

The peeps at Road are speculating from the latest image releases that Sansar will be a more ‘stylized’ than ‘photorealistic’ world.  They make a mistake in saying that ‘stylish’ and ‘beautiful’ are words not used in describing Second Life™. They obviously haven’t looked at the images in: MOST EXCELLENT SecondLife (some NSFW±) or Stunning Second Life Images (safer). It can be argued that many of the images in these Flickr groups have been Photoshop’d. But, are we sure the Lab didn’t Photoshop these images for dramatic affect? I can edit video to improve appearance… Ben Lang, the author of the article, says,

While Second Life is capable of looking good in the right circumstances, the company acknowledges that it isn’t optimized for the high-performance rendering requirements of VR headsets. That’s one reason why Linen Lab is starting from scratch with Project Sansar.

I don’t see that statement as accurate. It is to my mind somewhat misleading. Sansar wasn’t started because Second Life can’t be rendered well. I think the Flickr groups show it can be. The reason given by Ebbe Altberg, CEO of Linden Lab, is that Second Life can’t render well at 90+ frames per second (FPS) as needed by VR headsets. The SL viewer has a feature to  render high resolution images, higher than normal. If you follow Strawberry Singh, you know how she uses the feature to capture and create her awesome images. Ben Lang points to his previous thoughts on stylized worlds verse photorealistic worlds (or games). He makes a good point that ‘stylized’ games have lasted longer. But, he doesn’t go into depth on what it is that has made the games last. He is only pointing out that some long lasting games with a stylized look have aged well. Some gorgeous ‘stylized’ games have not lasted, think the Myst series. Game design education and tutorials have a load of stuff about the importance and lack of importance that visual quality has on a games ability to survive and/or attract users/players. (Minecraft?) The biggest games (most users) are mostly 2D Flash based Facebook games. So, I can’t put much weight in Ben’s thinking. I’m not saying he isn’t making a valid point. I am saying the point is only a facet of a larger picture. Ben points to the lighting system that Sansar will have. He makes the point that baking in lighting* is not going to work in a world in which we are walking around. A lighting system will be needed to maintain an immersive illusion. Ben sees in the new images, “…showing simple but effective reflections, bloom, and atmospheric perspective.” That assumes the photos are raw renders from Sansar. I do expect Sansar to have a good lighting system. Ben talks about how Sansar can ‘last’ if they stay with a ‘stylized’ world. I think he is missing the point of Sansar. The Lab won’t be maintaining a ‘look’ or style. They are providing a development platform with as few limitations as possible. So, the people creating in Sansar… or with Sansar tools… wording depends on how one thinks of Sansar, will create whatever they want, just as they do with Unreal or Unity. Quotes from Ebbe do suggest that there may be some limits imposed that we haven’t heard much about. Ebbe is quoted as saying,

Our users are starting to understand that VR is a different medium and they may need to approach it differently than they have in the past. We’ll do a lot of things to help users understand how to create performant content. There’s a lot of work yet to do, but we have plans for things like automatic optimization of content, polygon reduction of content that preserves quality at the same time, including showing users that create content some sort of visual indication of how performant their content is going to be across various platforms… I’ve already see stuff in Sansar that I know you’d never see in SL.

The Quick Graphics RC Viewer (download) is an example of how the Lab is implementing showing developers and users how well content is optimized in Second Life. They haven’t stopped developers from creating un-optimized content. I think it likely they will take the same tack in Sansar. Ebbe’s words suggest it is more an educational process than imposed limits. Next page – link below