PBR Materials @SL University

[EliseAnne85]

19 minutes ago, Innula Zenovka said:

In particular, it assumes that

• the devs who were tasked with introducing the shiny new feature also have the skills and background necessary to fix the performance issues, too, and LL chose to deploy them on the new feature rather than the fix

This is true, Innula, so true.  I've read on the forums, some years? (not really sure how long) ago that the code for SL is very old and most new coders don't know the code.  I am not a geek so I do not know how much of what I am reiterating is true or not.  But, it is what I've read here on the SL forums.  Also, there is a problem with Apple discontinuing OpenGL or something like that.  

The thing is, we can turn reflections off.  To me, what I'm seeing in some photos, especially in the video in the OP, is the reflections are too pronounced and it hurts my eyes a bit.  A slider so we could have reflections where it's comfortable on an individual basis would/could be helpful to consider as things progress.

However, to Arielle, it just isn't the things that bring nor keep people here.  I've suffered SL burnout several times.  What keeps one going in SL from the start is a bit of a mystery.  But, I think when we all started in Classic avatars, it was more equal that way.  We all looked a little goofy in our Classic avatar.  Plus, it was less expensive too.

[Love Zhaoying]

25 minutes ago, Innula Zenovka said:

Every time LL introduce a new feature, someone invariably says, "This shiny new feature is all very well, but why haven't they done something about fixing such-and-such?"   

To my mind, though, this objection is based on the highly questionable assumption that there was a simple choice between, on the one hand, the shiny new feature and, on the other, the performance improvements/fixes that have supposedly been neglected.   

In particular, it assumes that

• the devs who were tasked with introducing the shiny new feature also have the skills and background necessary to fix the performance issues, too, and LL chose to deploy them on the new feature rather than the fix;
• that the performance issues are, indeed, fixable without some major revision of the viewer/server/whatever;
• the performance issues are not, in fact, being worked on at this moment by a different team.

 Quite possibly it was a simple choice between introducing the new feature and fixing some outstanding issue, but I have no way of determining this (partly because I don't know what particular performance issues you have in mind, of course), and I don't think it's safe to assume that the choice exists and LL made the wrong call.

Some people can't resist whataboutism, it gives them something to complain about!

[Nalates Urriah]

On 2/12/2023 at 4:11 AM, Innula Zenovka said:

Maybe I'm missing the point, but since the simulator can't run faster than 45 fps, isn't anything greater than that wasted?

Meanwhile, the standard framerate at which movies and TV shows are made is 24 fps.

I never pay much attention to my fps unless I'm in a very laggy environment -- lots of moving avatars, and lots of textures and complex objects for the viewer to render -- but unless if falls much below 20 I very rarely have cause to notice it, since everything seems to be working to a perfectly acceptable standard.

Additionally, whatever framerate my viewer is reporting, what I actually see is surely going to be limited by the refresh rate of my monitor, which is 60 Hz -- that is, the screen refreshes itself 60 times a second, so no matter what the viewer thinks it's doing, 60 fps is all I'm going to see.  (Windows users can check theirs with Settings>Display>Advanced Display).

What am I missing by not having 200 fps in SL?

I think you got the idea that viewer FPS >60 doesn't do much visually. And viewer side >60 has only a few minor benefits. What people have left out so far (or I've missed) is viewer response to the server in SL combat games.

One finds out that whoever has a higher FPS has a shootout advantage. Having competed in combat play I found the players with 150 & > FPS were WAY harder to beat when I was playing with ~45 to 60 FPS. If we both hit SHOOT at the same RL time their viewer is much more likely to get the signal on its way to the server before my viewer does.

The server, viewer, and animations all deal with timing to make things look right. All the frames are snapshots of what the world looks like in that frame instance. The viewer, server, and animation are seldom in sync. The viewer gets what it can and does the best it can with that for however fast it is running.

The viewer is figuring out where an animation is and how much things have moved at a specific point in time by prorating/estimating where an arm is at the instance of the frame the viewer is rendering. The viewer is NOT taking a frame from the animation. The animation is designed based on frames. But the idea of a frame in an animation is not the same as a frame in a video. It is much more of a time marker.

An animation places key-frames at various points on the time line. Basically saying 'at this point in time the arm is here'. The viewer figures out where the arm is for the frame it is currently rendering, which is somewhere in the animations time line between key-frames. It prorates position between the nearest key-frames.

The viewer also considers the "clock". So an animation seems to play the same speed whether the viewer is running 10 FPS, 60, or 120. The idea is to adjust the time line of the animation to the current render speed to keep things moving in a realistic way and in sync with the rest of the SL world.

An animation FPS setting is really a setting for how many key-frames one can put in the animation not the play back speed. The key-frame positions at the 1-second mark should always show those positions at 1 second into the animation regardless of viewer FPS. So at 24FPS I can put 24 key-frames in the video's first second. Whether I put 24 key-frames in or 2 it should take the first second of animation 1 second to play.

On 2/12/2023 at 5:22 AM, Jenna Huntsman said:

That page is full of misinformation and was written by a user over 10 years ago.

I'm going to add a warning to that page saying that the information contained is incorrect, as while some of the content of the page is correct, broadly speaking it gives people the wrong idea on how things actually work - so ignore the content of that page.

If you have editing privileges on the wiki, change the bad information. I've updated a number of articles in the wiki. The Lindens review significant changes.

On 2/12/2023 at 6:26 AM, EliseAnne85 said:

Oh, I see.  

However, we could lose a lot of current great content then.

No... The Lindens are careful about avoiding breaking legacy stuff. So the 'diffuse' textures will still work. Those with baked in shadows and AO will have double shadows and AO when such a texture is used in a new PBR material . We'll see the baked in shadow/AO and the shadow/AO added by PBR rendering. So... they will look funky.

23 minutes ago, Love Zhaoying said:

Some people can't resist whataboutism, it gives them something to complain about!

What if those people didn't exist? 😆

Edited February 13, 2023 by Nalates Urriah

[Love Zhaoying]

11 minutes ago, Nalates Urriah said:

34 minutes ago, Love Zhaoying said:

Some people can't resist whataboutism, it gives them something to complain about!

What if those people didn't exist? 😆

Tease!!

[arton Rotaru]

2 hours ago, Arielle Popstar said:

When I look at the recent release notes, I see a long list of resolved issues and the list of things yet to do and think how far the viewer could have been had they focused on function rather than form. Aesthetics are all well and good but unfortunately the Lab's is too focused on the eye candy only, rather than making the viewer function better in general. This PBR and the EEP make me think they are attempting to bring Sansar to S/L and well, the lesson from that platform was that the looks were not enough to draw people in so why would it now?

The known issues on the release notes page is super outdated. A lot of these are actually resolved as well, but other issues, not listed there are still open.
The resolved issues list is also very incomplete.

We talked about the improved performance earlier in this thread already. This project is not only about PBR eye candy. It also has a completely new texture streaming pipeline, besides several other performance improvements.
It's as Innula says, they are shipping the performance improvments along with new features this time.

[arton Rotaru]

44 minutes ago, EliseAnne85 said:

This is true, Innula, so true.  I've read on the forums, some years? (not really sure how long) ago that the code for SL is very old and most new coders don't know the code.  I am not a geek so I do not know how much of what I am reiterating is true or not.  But, it is what I've read here on the SL forums.  Also, there is a problem with Apple discontinuing OpenGL or something like that.

Those who are working on that project do know the code base very well. DaveP is a Linden since 2005. If you follow them on github, you could see what a major rewrite of the renderpipe this project actually is.
Nonetheless, we all, the devs, and the residents who are alpha testing this, are trying our very best to keep the over all look, and feel, of the SL we love so much, as we know it.

[EliseAnne85]

21 minutes ago, Nalates Urriah said:

No... The Lindens are careful about avoiding breaking legacy stuff. So the 'diffuse' textures will still work. Those with backed in shadows and AO will have double shadows and AO. The ones baked in and those added by PBR rendering. So... they will look funky.

So, it won't break but it may look funky.  

I have obliterated a lot of my Classic avatar and sculpty things, but I have mesh things I don't want to lose.  

I am eager to see what some of my mesh stuff will look like.  My jewelry collection took me a long time to accumulate and I wear jewelry everyday just about in my SL.  Ah, I don't want to lose certain things...this has me a bit edgy. 

 

[arton Rotaru]

3 minutes ago, EliseAnne85 said:

So, it won't break but it may look funky.

It won't look more funky than it always would be. If you have a baked shadow prim underneath an object, and someone has realtime shadows enabled, the funk is already there. Always has been like that.

[EliseAnne85]

5 minutes ago, arton Rotaru said:

If you have a baked shadow prim underneath an object, and someone has realtime shadows enabled

If it's a shadow prim, it's most likely furniture or home decor and most of that is copy/mod.  Can't you click on EDIT LINKED in the edit menu, then click on the shadow and turn that shadow to 100% transparency so it will disappear altogether?  I wasn't sure what you meant by baked shadow prim...?  I'm assuming you mean shadow prim.  

Edited February 13, 2023 by EliseAnne85

[arton Rotaru]

1 minute ago, EliseAnne85 said:

If it's a shadow prim, it's most likely furniture or home decor and most of that is copy/mod.  Can't you click on EDIT LINKED in the edit menu, then click on the shadow and turn that shadow to 100% transparency so it will disappear altogether?  I wasn't sure what you meant by baked shadow prim...?  I'm assuming you mean shadow prim. 

Yeah, semantics. It was just an example. Indeed you can remove such a shadow prim, but my point was; The owner of said thing doesn't use realtime shadows, instead he likes these shadow prims. Then a friend comes over, who happens to like realtime shadows though, and has them enabled. Get what I mean?

[Ayashe Ninetails]

3 minutes ago, EliseAnne85 said:

If it's a shadow prim, it's most likely furniture or home decor and most of that is copy/mod.  Can't you click on EDIT LINKED in the edit menu, then click on the shadow and turn that shadow to 100% transparency so it will disappear altogether?  I wasn't sure what you meant by baked shadow prim...?  I'm assuming you mean shadow prim.  

Some items (furniture and homes, in particular) have shadows baked into the texture. I run into this all the time when trying to customize my couches and homes. Move or delete a cushion and there's a big ole black shadow behind it drawn directly on the texture. Need to toss something else in place to hide it.

[Coffee Pancake]

1 hour ago, Innula Zenovka said:

Every time LL introduce a new feature, someone invariably says, "This shiny new feature is all very well, but why haven't they done something about fixing such-and-such?"   

It's simpler than that.

Such-and-such is functioning as expected, therefore, it's not broken.

[EliseAnne85]

16 minutes ago, Ayashe Ninetails said:

Some items (furniture and homes, in particular) have shadows baked into the texture. I run into this all the time when trying to customize my couches and homes. Move or delete a cushion and there's a big ole black shadow behind it drawn directly on the texture. Need to toss something else in place to hide it.

I know about that.  He said shadow "prim", the prim part confused me.  

[Ayashe Ninetails]

4 minutes ago, EliseAnne85 said:

I know about that.  He said shadow "prim", the prim part confused me.  

Yeah, I actually prefer the shadow prims, but they can cause issues, too. I don't use shadows with Firestorm much, but I do when I'm doing photography. It's all a pain and prims sometimes need to be shifted around so they don't look silly.

[Arielle Popstar]

3 hours ago, Innula Zenovka said:

Every time LL introduce a new feature, someone invariably says, "This shiny new feature is all very well, but why haven't they done something about fixing such-and-such?"   

To my mind, though, this objection is based on the highly questionable assumption that there was a simple choice between, on the one hand, the shiny new feature and, on the other, the performance improvements/fixes that have supposedly been neglected.   

Is not your own objection to those objections based on an assumption that the objectors were only assuming rather than having actually done some research into whatever enhanced function they were promoting?  Aside from the typical Linden rebuttals that something is working as designed as per Cookie, there is the even more common objection that they don't have the time for this or that because of other design projects, not because it is not possible.

The problem as I see it is that the Lab favours aesthetics over function and so invests heavily into making things look pretty while neglecting the ease and usability of our inworld interactions. This results in 

Quote

if the functionality is poor, it doesn’t matter how good something looks, engagement will drop off a cliff edge and no amount of aesthetics will be able to claw it back! https://layers.studio/aesthetics-vs-function/

There is a fair amount of literature out there that discusses the problems between form and function that plays out here in Secondlife and it is readily seen that the Lab favours one over the other to such a degree that the inability to retain user users may very well be a result of that. I have to wonder too if Sansar's demise wasn't a result of that same tunnel vision for eye candy rather than a more balanced approached to also focus in on usability.

My own focus isn't so much on performance as some other aspects that I mentioned early on in the thread:

Quote

 

Plenty of things have been mentioned as potential upgrades to benefit the regular resident like enhanced inworld building tools, a Backspace/revert function in the viewer, a better Inventory UI. An easier more streamlined demo and shopping experience, better marketplace functions, etc etc. A number of those would have a much greater effect on our overall inworld experience and even a boon to the creator then prettier lights and reflections.

 

 

[Arielle Popstar]

4 hours ago, Love Zhaoying said:

Some people can't resist whataboutism, it gives them something to complain about!

It was after all what the thread was about but in any case I'll point you to a post about about how we'd like to see these sort of discussions go and hope you can learn something from the thoughts the author mentions on how to have a worthwhile discussion:

 

[EliseAnne85]

52 minutes ago, Arielle Popstar said:

If the functionality is poor, it doesn’t matter how good something looks, engagement will drop off a cliff edge and no amount of aesthetics will be able to claw it back! https://layers.studio/aesthetics-vs-function/

I read the article.  The above paragraph is the standout.  I'd have to say I became way less engaged when EEP arrived.  EEP is too much hassle, cumbersome, too dark, too much glow (had to turn off glow), full bright works horrible, pain in the neck to where I can have great lighting in my homes (finally after months of tweeking) but teleport to someone's party and my avatar looks hideous.  I can relate to how I became disengaged with SL due to EEP.  I preferred the Windlight method of the click on menu and one slider and voila!  Windlight took minutes of one's time to achieve great lighting.  This was my experience, and EEP makes it a hassle to create because you don't know what kind of lighting anyone has because EEP is a million and one lightings.  What can one say about someone who leaves SL because of how SL changes?  It happens.  I don't know what else to say.  EEP I hate the thing.  I have cool lighting in my homes.

[Innula Zenovka]

49 minutes ago, Arielle Popstar said:

Is not your own objection to those objections based on an assumption that the objectors were only assuming rather than having actually done some research into whatever enhanced function they were promoting? 

Yes, it is, since I've yet to see anyone explain what exactly it is they think could have been done instead of whatever Linden Lab actually did do, and why they think that people who did whatever actually was done could just as well have done whatever it was the person thinks should have been done.

However, there's a first time for everything and I look forward to your explaining what you think could have been done instead and why you think people with the skills to introduce PBR would have been able make the improvements you would have preferred to see.

 

[Rowan Amore]

22 minutes ago, EliseAnne85 said:

EEP makes it a hassle to create because you don't know what kind of lighting anyone has because EEP is a million and one lightings.  

You've never known what kind of lighting anyone else is using.  There were literally 100s if not 1000s of WL settings you could make.   

The setting you make at your home will never be what you see elsewhere unless you APPLY TO SELF and uncheck the option to change the setting for each parcel.

 

[Arielle Popstar]

8 minutes ago, Innula Zenovka said:

Yes, it is, since I've yet to see anyone explain what exactly it is they think could have been done instead of whatever Linden Lab actually did do, and why they think that people who did whatever actually was done could just as well have done whatever it was the person thinks should have been done.

However, there's a first time for everything and I look forward to your explaining what you think could have been done instead and why you think people with the skills to introduce PBR would have been able make the improvements you would have preferred to see.

The lead developer in this case has been with S/L for 18 years with unknown amount of programming experience before that. It is highly unlikely that it would have been the only project they were capable of. Even then, the Lab has the means and monies to afford to have multiple developers on staff to be able to balance out both form and function and yet it is only the former they focus on.

What I think could have been done instead of or in addition to, is spread throughout the forum aside from what I have already mentioned in my previous post.

[Arielle Popstar]

5 minutes ago, Rowan Amore said:

You've never known what kind of lighting anyone else is using.  There were literally 100s if not 1000s of WL settings you could make.   

The setting you make at your home will never be what you see elsewhere unless you APPLY TO SELF and uncheck the option to change the setting for each parcel.

 

Have to wonder if there wouldn't be a way someone could be wearing a sort of facelight that bathes someone in the windlight/eep lighting one wished to be seen in aside from what the rest of the region is set to.

[EliseAnne85]

11 minutes ago, Rowan Amore said:

You've never known what kind of lighting anyone else is using.  There were literally 100s if not 1000s of WL settings you could make.   

Windlight was more subtle if on a sim and those were mostly windlights for viewing skin.  The other windlights were very surrealistic for photo effects and by just moving the slider a little one could get some amazing effects for photos, plus highlights and shadows in just minutes and then with a click and a slide go back to a skin lighting in seconds.  But, those 1000 of surrealistic windlights you are talking about were rarely used on a sim but you could work them in minutes with just a click thru menu and a slide.  And, ALM we turned on for photos as well rather than run it all the time which a lot of people still do today.  

Edited February 14, 2023 by EliseAnne85

[Rowan Amore]

1 minute ago, Arielle Popstar said:

Have to wonder if there wouldn't be a way someone could be wearing a sort of facelight that bathes someone in the windlight/eep lighting one wished to be seen in aside from what the rest of the region is set to.

No, you'll just blind everyone in the room but give it a try!  Assuming people have attached lights enabled, of course.

[Arielle Popstar]

Just now, Rowan Amore said:

No, you'll just blind everyone in the room but give it a try!  Assuming people have attached lights enabled, of course.

Why do you laugh at that idea? You think it is impossible even after seeing what they are doing with lighting in other ways? Facelights don't have to be blinding if their intensity is toned down.

[Rowan Amore]

Just now, EliseAnne85 said:

Windlight was more subtle if on a sim and those were mostly windlights for viewing skin.  The other windlights were very surrealistic for photo effects and by just moving the slider a little one could get some amazing effects for photos, plus highlights and shadows in just minutes and then with a click and a slide go back to a skin lighting in seconds.  But, those surrealistic windlights were rarely used on a sim.  And, ALM we turned on for photos as well rather than run it all the time which a lot of people still do today.  

It's basically the same for EEP.  I have 2 folders in my settings folder.  One for everyday running around.  Nam's optimal skin and prim, edited just a.bit that I Apply to Self.  It stays that way no matter where I go.  When taking pics, I generally use the region EEP which I'll also lighten easily from the window by using the little boxes and often move the sun or moon around with the slider, just like in WL.