The Subtle Power of PBR in Second Life is Much More Than Mirrors

Now that PBR graphics are fully launched on Second Life, we're seeing a barrage of pics and videos showing off reflective/mirror effects made possible from physically based rendering. But it's this video of a swimming pool created by SL content creator Bhad Craven which recently caught my eye -- watch. It's not so much the mirror effects per se, but how the light reflecting on linoleum tiles and metal surfaces adds a subtle sense of realism and immersion.

"PBR is a game changer," Bhad affirms, laughing. "I hope people start to like it as much as creators. It's so much faster since baking isn't a thing. It makes my items actually shine, not having to fake reflections is great."

Not that developing in PBR is easy -- here's some notes of caution/advice from Mr. Craven:

"It's fairly difficult to create a large scene like this, especially since a lot of the parts have to be baked individually still. It's a learning curve but still a huge update from what we previously had." He recommends learning how to use EEPs and reflection probes to maximize the use of PBR. "There's a blue halo around reflective objects which aren't inside an EEP. Getting the sweet-spot of making sure my textures are the right level of reflective, seems to be my biggest challenge so far."

He also advises watching this tutorial:

As for this swimming pool, it's available under his brand, Bad Unicorn (disclosure: a proud sponsoring partner to NWN!) at Kustom9 until October 10: Teleport here to check it out.

Comments

[Jayda]

That looks amazing. What kind of impact does this have on performance?

[PBRCanBeSoMuchMoreIfYouCanSeeIt]

Linden Lab's reliance on people learning about EE's and Reflection Probes, and using 'proper' PBR to it's fullest - as in being able to tell a story with it - that the bar has been raise FAR above the average user 'tinkering with making their own stuff'.
Those days are gone, and WILL affect new users and their experience going forward. This was a problem with Sansar - though pretty and very easy (and wonderful that you got a free world!!! SL Land is cost-prohibitive!) also added obscurity to the process of creation that raised it to 'game developer' level skills as well.

At least we didn't have to deal with EE and Reflection probes, that was all built in and you could simply upload and place objects in lickety split, and the PBR looked very nice - unlike SL with all it's custom requirements and 'workarounds' that really add friction to the creation process.

Funny how the majority of SL'ers rejected Sansar - and cling to the current SL which is doing everything Sansar promised - but badly on an outdated engine not fit for purpose.

I'm not impressed with PBR on SL quite yet. As an experienced PBR'er in this department, it's not quite there, but Sansar was easy to use and the results were represented 99% as intended. Not perfect, but it was there.
I wish more SL'ers actually had tried it out with an open mind - now as they struggle with new PBR viewers there, PBR ran fine and some would find Sansar would have run faster for them vs SL amazingly

Oh well, one can only just sit back and laugh I guess. For me, I'm waiting until the dust settles at SL with PBR, before I waste time changing my workflow to accommodate. And now that they have chased off all my customers with 'game developer' only level skills needed to make the most basic thing and compete... not sure if there is much future there..

[LOLMAO]

What's funny and laughable is in fact that some people still don't comprehend why Sansar was dead-on-arrival and failed to attract users, also beyond Second Life, despite numerous partnerships, promotions, and repurposing efforts.
Meanwhile, VRChat has been widely adopted, even by some Second Life users.

You, PBRCanblahblah (like LL in 2017), totally misunderstood who the actual potential Sansar users were, what they were looking for and where they are now.
Many people registered, logged in, piled up at the early starting places, curious to see and try it. And what happened? They didn't come back. Of the few who remained, many gave up after a while. What a splendid user retention!

The common critiques were (guess what?):
- The only good things about Sansar were its graphics (compared to SL, but in fact rather average) and the cheap world (well, maybe not the prices on the marketplace), while the rest was unsatisfactory, to put it mildly.
- Those long loading times in 2017, when the average network speed was a fraction of what it is today. For many users it was 15-30 minutes of waiting, before the world appeared, staring at an inert client, thinking it froze and didn't work, because initially Sansar was even without any loading bar. LL didn't listen to the testers of their so-called "closed beta" asking for a loading bar. And that was the result.
And of course launching that early alpha or tech demo as a 'beta whatever' in 2017 also didn't help.
Ah, and aiming it at non-gamers (yay, VR museums!) and requiring a gaming rig? Even though LL changed its aim, later, it was another misstep in a long series of missteps.

And the reviews on Steam after they tried to attract not-SL'ers gamers there in 2018? I just had a look after all these years, and I had another laugh.

"Sansar [in 2018] performs pretty poorly, on my RTX 2080 it lags anytime I set the graphics to "High", and in any other mode the graphics look terrible. The anti-aliasing is blurry and uncomfortable." [they fixed some issues later] --- "After following the hype train [LL was trying to involve YouTubers in 2019] over this game, I decided to check it out in spite of all the negative reviews [...] The negative reviews were exactly right, and they didn't even touch on some of the issues I had with Sansar." --- "I keep checking in on this game and I keep being disappointed." --- "Very long loading." --- "it wants to be IMVU so bad but its way WORSE. and i thought IMVU was bad." --- "This feels very much like a game made for desktop first, VR second. The VR UX experience felt very unintuitive and frustrating to adjust to [...] If you are looking for a VR game that provides more options and features for creating VR content, do not bother with this game." --- "This is a horror game. The user interface and the whole experience is so bad it's almost torture. I feel like the whole thing is a big social experiment and the research question is "who can take ♥♥♥♥ the longest before quitting?" Oh and speaking of quitting, the game doesn't provide such option. I bet the devs are counting how many people pull the plug because they don't know how to close programs in Windows and reporting it to Facebook or something. " --- "just play VRChat."

LL screwed it up, each misstep further alienating another slice of potential users.

Trying to frame it as closed-minded SL'ers and making fun of it, instead of understanding what in fact was rejected of Sansar, is really shortsighted and misleading. It is also exactly the kind of attitude that only helped to sink Sansar. Even who wanted to like Sansar and wasn't a blind fanboy-fangirl understood that the ship that could have been a wonder was instead a mess and full of holes.

Already at the initial launch in 2017, some people quickly realized that Sansar was at risk of being a dead-on-arrival without decisive changes.

Meanwhile, a fangirl on her blog was denying the reality and claiming that downloading the rooms was quick - several Gb of data, because LL didn't even had any guideline - "forgetting" to mention her internet speed. Only later someone started optimizing their assets, to improve the downloading time.
She was also telling that Sansar ran faster than SL, and good even on her laptop, of course also "forgetting" to mention her specs or the frame rate, and also "forgetting" another important detail: SL never ran optimally anywhere, but could run on most machines, even on low-end hardware. Sansar, on the other hand, ran ok on good hardware, but would struggle or won't work at all on mid/low-end hardware that was instead capable of running SL. And so on, clinging to it, misleading her readers.
And, again, the issues were many others.

Indeed, Sansar became such an epic fail, that had an average user concurrency of 2 digits (lately typically less than 20), dismissed not just by most SL'ers, but also by almost anyone else who tried it and also who tried it again. Congratulations.

And someone still clings so hard to it, that fails to get it even today.