PBR WOW!

[arton Rotaru]

1 hour ago, Rowan Amore said:

Just as a small vocal minority got us EEP, I'm sure that same vocal.minority got us PBR.

I dunno who asked for EEP, but PBR came from Linden Lab, because they want to implement the full glTF standard into SL. Part of this glTF specification is the PBR Metallic-Roughness material.

I'm not aware of any group who asked for glTF actually. 🤷‍♂️

Other than that I'm thinking pretty much the same as you about the mobile client. 😇

[CalypsoJewel]

1 hour ago, Chic Aeon said:

BACK TO PBR

 

Just wanted folks to remember when both MESH (like a dozen years ago now) and EEP came onto the scene. It was NOT PRETTY.  Give the devs some time. Someone posted that FS was waiting for PBR to be less hassle free before releasing their PBR viewer.  I have no idea where that info came from but even it is is false it  will be quite awhile (like a year? ) before FS folks are required to switch to the PBR version and then "possibly" they could choose to use the OS viewer as an option.

 

I remember reading something about this. I believe it was on the FS viewer page under the news "Firestorm 6.6.16 Release – Say hello, wave bye bye".

[Henri Beauchamp]

7 hours ago, LipstickAndDreams said:

ALM is still slower than the forward renderer, after all these years.

Actually, it depends on your GPU... Modern GPUs can run ALM (deferred rendering) just as fast as forward rendering, and even slightly faster in some specific cases (depending on what is being rendered).

However ALM consumes more memory (at least when there are materials around), and PBR even more. So, here again, depending on the scene being rendered (e.g. large venues with lots of avatars) you may see issues with ALM that you won't see happening, or only later (in even worst scenarios, such as with more avatars coming around) with forward rendering, due to the VRAM filling up to the brim and the vertex buffer allocations starting to spill over the RAM, slowing everything down and/or causing stuttering/hiccups in frame rates.

While with a GTX 970 (and older cards) I was pretty much never using ALM (since slower), with a GTX 1070Ti I could use either, and with my latest RTX 3070 ALM is most often faster. Yet, there are cases when I still prefer using forward rendering, such as when driving/flying/sailing outdoors, since then there are almost no materials in sight (so I loose nothing with ALM off), and forward rendering still uses less VRAM (meaning I can push the draw distance to the max, which is quite useful for such activities, letting you see where you go) and offers a waaaaaay better anti-aliasing (native GPU SSAA), and this despite the fact my viewer got SMAA shaders for ALM, which are themselves way better than LL's FXAA: the scene looks even more gorgeous with just forward rendering in this case.

As for old GPUs or iGPUs/APUs, yes, ALM is definitely way slower than forward rendering, often by a factor of two !

I warned LL about what would happen if they removed forward rendering. They ignored my warning; now what I ”predicted” (or rather simply inferred) is happening... 🫣

Edited December 16, 2023 by Henri Beauchamp

[LipstickAndDreams]

2 hours ago, Lyssa Greymoon said:

Well, when I got home I decided to see what it's like on my OG M1 MacBook Pro, and it's just fine. I guess I need to find out what the hell is wrong with that Mac Studio. 

Just realized it may a difference in monitor resolution. What are the resolutions for the displays of both machines? If you decrease the display resolution on your Studio, then it will likely be faster.

Edited December 16, 2023 by LipstickAndDreams

[Zalificent Corvinus]

4 hours ago, Cinnamon Mistwood said:

Change and adapt or watch the platform you enjoy die a slow, neglected death due to an inability to take a single chance or change a single thing.

That's EXACTLY what people said when LL decided to waste 6 years and tens of millions of dollars on Sansar.

 

4 hours ago, Cinnamon Mistwood said:

I don't know what goes into decision-making at the lab, but I be willing to guess that someone there did look into some market research

If that were true, they wouldn't have claimed that 60% of all SecondLifers used whyPhones, and made their FIRST failed attempt at a mobile viewer whyPhone only, and then canceled it as soon as it was approved for the whyPhone App store.

 

The change to the places/landmarks floater were based off asking a few LL employees how they found LM's in their inventory.

Their market research had them encouraging European users to "go out and vote" in the American Presidential Election.

They have no idea who uses SL, what it's used for, how, or why.

LL's market research is, to be blunt, a bad joke.

 

[Zalificent Corvinus]

2 hours ago, Chic Aeon said:

In less than a year the devs had things working so that those unbelievably slow loading sites were on par with the ones that matched our always fast to load areas.   

You do remember that sansar failed, and they fired most of its team, right ?

[Chic Aeon]

1 hour ago, Zalificent Corvinus said:

You do remember that sansar failed, and they fired most of its team, right ?

Didn't fail because of the tech. By the time they sold it was running beautifully.  I was there every day and most of the day

[Cinnamon Mistwood]

1 hour ago, Zalificent Corvinus said:

That's EXACTLY what people said when LL decided to waste 6 years and tens of millions of dollars on Sansar.

Does that mean you think the Lab should never do anything to change or grow?  Never try anything new.   Or does it mean you think you are the only one who knows how to make the platform grow and, of course, all the tech that they should be using? All hail Zalificent the All Knowing?

What would you suggest they do?  You don't like anyone's ideas in here.  You try to shut down every discussion started in good faith to come up with new ideas.  You don't like any decision the Lab makes.  Every suggestion is shot down.  Everything that has ever been done/tried is a failed attempt by incompetent, ignorant idiots.

How would you fix the problems you see in SL?  We have all read your scathing reports on how stupid everyone (but you) is.  So, let's hear it.  How would Zalificent solve SL's concurrency and tech problems?  You obviously know what they should be doing if you know that everything they do is wrong.

Should nobody ever do anything because... I don't know.  Because you say so?  In a perfect world, they hire you to solve all the issues. How would you do it?

Toss your ideas out here for the rest of us to see.

 

[Paul Hexem]

1 hour ago, Cinnamon Mistwood said:

Does that mean you think the Lab should never do anything to change or grow?  Never try anything new.   Or does it mean you think you are the only one who knows how to make the platform grow and, of course, all the tech that they should be using? All hail Zalificent the All Knowing?

This isn't an isolated case. Every time anything is done, or suggested be done, to try to get SL up to even remotely modern standards there's a significant number of forum users that say it's stupid, it's a giant waste, it's going to cut off all the average users, gamers this and that, SL doesn't need change, SL shouldn't be updated, etc.

People forget that PBR isn't some newfangled thing that the Lab cooked up, it's something that has origins as far back as 1997, but really developed in 2010. 23 years ago!

So it's a little silly to see in this thread the mentions of users that are going to struggle because their hardware supposedly can't support PBR.

If you're connecting to the Internet with a computer that can't support technology from almost a quarter century ago, you've got bigger things to worry about than if you can log into SL or not.

[LipstickAndDreams]

3 hours ago, Paul Hexem said:

People forget that PBR isn't some newfangled thing that the Lab cooked up, it's something that has origins as far back as 1997, but really developed in 2010. 23 years ago!

The issue isn't the technology itself. Old computers can easily run this stuff if the implementation is efficient. LL's implementation is not. Whether or not this is because they have reached the limit of OpenGL's performance capability given the overabundance of complex objects in SL, I don't know. What I do know is that the viewer's efficiency should have been prioritized before introducing shiny new crap.

Quote

If you're connecting to the Internet with a computer that can't support technology from almost a quarter century ago, you've got bigger things to worry about than if you can log into SL or not.

I don't even know where to begin with this lmao

Edited December 16, 2023 by LipstickAndDreams

[LipstickAndDreams]

4 hours ago, Cinnamon Mistwood said:

Does that mean you think the Lab should never do anything to change or grow?  Never try anything new.

Not at all, new things are fine as long as they aren't half-assed like PBR is. If the performance issues were fixed before releasing it, this thread wouldn't be filled with so much negativity.

And if you think the people in this thread are being negative about the subject, wait until other SL users are forced into this. Its going to be hilarious to watch lol

Edited December 16, 2023 by LipstickAndDreams

[Zalificent Corvinus]

4 hours ago, Paul Hexem said:

If you're connecting to the Internet with a computer that can't support technology from almost a quarter century ago, you've got bigger things to worry about than if you can log into SL or not.

It's not "cant support tech from a quarter century ago", as the computrs in question are a damn sight newer than that and CAN support tyhe tech.

It's "cant support the talentless mess that's been made TRYING to add that tech"

Imagine you had a 3 year old car, it supports "round wheels" just fine, but then some idiot at the manufacturer decides they will "improve the wheels" but somehow, despite rival brands wheels still being ROUND, your brand ends up with TRANGULAR wheels, and when you complain that the ride is now super bumpy, some oaf whines "but wheels are ancient, if your car cant support wheels, you need a new car"

 

Do you really need a "new car" or do you...

5 hours ago, Cinnamon Mistwood said:

Toss your ideas out here for the rest of us to see.

...Fire the guys who made the wheels triangular, and hire new guys who know about making round wheels.

 

4 hours ago, Paul Hexem said:

So it's a little silly to see in this thread the mentions of users that are going to struggle because their hardware supposedly can't support PBR.

Cant support the fuster cluck coded fubar mess that was made trying to implement a cut down version of basic PBR, and failing.

The "PBR" we're being offered has almost NO advantages for the vast majority of SL users, none, but comes with a massive hit in performance for a great many.

 

[Zalificent Corvinus]

5 hours ago, Cinnamon Mistwood said:

Does that mean you think the Lab should never do anything to change or grow? 

Changing and growing, usually requires doing things to improve the platform, and attract more customers.

Ideas that seem designed to un-improve the platform and drive customers away, don't help.

 

5 hours ago, Cinnamon Mistwood said:

You don't like anyone's ideas in here.  You try to shut down every discussion started in good faith to come up with new ideas.

Have you actually looked at the ideas I dislike?

Put in a new system for lighting, that NO customer asked for, that was poorly conceived from the start, poorly executed, delivered years late and still isn't working properly.

Put in a substandard implementation of a cut down version of something everyone else started using 10 years ago, and, despite it being tried and tested tech, make such a hash of it, that it's a laughable parody of the tech who's name it borrows.

Put in a system to help NPC's navigate around obstacles, that's so poorly done it causes massive region performance issues, and then fails because of the damage it does to the region.

Put in a webcam based auto-gurning system, that almost nobody wanted or asked for, then find out, after wasting a lot of time, effort, and money, that it was "too hard" and nobody wanted it, and cancel the project.

 

Maybe you're upset that I didn't like the suggested "improvement" where it was proposed that LL should outlaw skyboxes and platforms, forcibly relocate all ex skybox owners to a temporary holding camp, and then deport them to Sansar. Yes that was a real suggestion for "improving SL" by the same person who wanted the auto-gurning system.

 

Maybe you are butthurt that I didn't like the suggestion that LL should make it "illegal" to build within 32 m of your parcel boundary, making all 512, 1024, 2048, 4096, and most 8192 parcels instantly unusable and worthless.

 

The ideas I dislike are invariable the bloody stupid ones designed to drive paying customers off the platform, that offer NO benefits to the user base.

5 hours ago, Cinnamon Mistwood said:

How would Zalificent solve SL's concurrency and tech problems? 

Hmmm, well lets start with dev time.

First thing you do is delete all the "project viewers"

LL has a TINY dev team.

Real Games Dev setups, if they come up with a "cool new feature", they assign a group of coders to it, 8 hours a day, 5 days a week, and get it ready for a public beta test in 3 - 6 months.

LL assign one guy, a few hours a day, one day a week, because the rest of their time is divided between a whole bunch of "projects", and emergency hot fixes to the live system, usually caused by some other half finished project being rolled a couple of years too early.

So, take that 3 guys working full time take 3 months project, and try it the LL way.

ETA for public beta, 5-7 years, then somebody gets tired of waiting and spending, and they stop dev work 3 & 1/2 years in, and punt it out as a half finished mess, then sit on their hands for some months, then reluctantly start a project viewer to fix some of it.

And that hot fix project eats into the dev time for the next cool feature.

 

First thing you do is cancel Philip the Dreamer's "love machine" plan, where back office non tech staff, get a bonus for suggesting "cool new things" that harm the system and that users never wanted. That wastes so much of the limited dev resources.

6 hours ago, Cinnamon Mistwood said:

You don't like any decision the Lab makes

Hmmm, a company that runs a system where by their own admission in times past, only 30% of their customers are American, but which still acts like there is NO life east of Vegas or West of the Golden Gate, and encourages the 70% of their user base outside the US to vote for POTUS, and jumps on every faddish stupidity that's mentioned in the inner office.

VR is a flop, lets try VR, everyone hates NFT's, lets partner with an NFT outfit, we don't run a first person shooter, let's add gamification that wont satisfy the gamerz and will p*ss off our actual customers.

That's not good market research or customer service, or planning.

 

6 hours ago, Cinnamon Mistwood said:

How would you fix the problems you see in SL?

How about prioritising the ACTUAL problems, like TP-fails, inventory loss, cloud induced performance degradation, bugs left unfixed for over 5 years, you know, crazy wild off the wall non-Futureness stuff like that.

Sounds like a plan ?

 

[Ceka Cianci]

4 hours ago, Paul Hexem said:

This isn't an isolated case. Every time anything is done, or suggested be done, to try to get SL up to even remotely modern standards there's a significant number of forum users that say it's stupid, it's a giant waste, it's going to cut off all the average users, gamers this and that, SL doesn't need change, SL shouldn't be updated, etc.

People forget that PBR isn't some newfangled thing that the Lab cooked up, it's something that has origins as far back as 1997, but really developed in 2010. 23 years ago!

So it's a little silly to see in this thread the mentions of users that are going to struggle because their hardware supposedly can't support PBR.

If you're connecting to the Internet with a computer that can't support technology from almost a quarter century ago, you've got bigger things to worry about than if you can log into SL or not.

I think a lot of computers can run what's out there now, mainly because they download and have things cached..

We always get people in the forums talking about how they can run this game or that game and the graphics are better here and there.. But then they get in here and it's like they hit a wall..

Those programs don't change constantly in real time like SL does and get updated once in awhile.. Let alone they have creators on staff making the content and balancing performance and eye candy.. Where in SL it's the users of all levels of knowledge in creating,  creating the things for the grid..

We have users that are creators that try to put out content that is balanced, but then some  just overload on the eye candy as well..Then you have the ones that either don't bother to read what would be efficient  or even care about it, or the ones the ones  that are just learning, tossing anything out there not knowing or not thinking about the weight and impact of their creations..

I think the real time of SL and the unbalanced content  is what makes SL much different than the downloaded programs sitting on someones system..

I think if a lot of those games nowadays didn't have rules of balance of what could go in them and put into real time changes like SL, they probably wouldn't run on older systems either..

I think SL puts a computer to the test really compared to a lot of things out there today..I think that's something that a lot of people do see for themselves when they hit the grid..

It's gonna be what it's gonna be and they are gonna do what they are gonna do.. Us, we'll either sink or swim I guess.. hehehe

 

[Qie Niangao]

Even after all these years, much Second Life content still uses only the diffuse texture map, so the rendering is pretty simplistic even with ALM turned on. There's no particular reason that glTF materials should use more memory than full Blinn-Phong materials (how often will the emissive mask be used, really? and for that matter, how many albedo maps should just be a 32x32 blank texture with an applied tint, if the model's surfaces are sensibly partitioned for materials?).

Anyway, once forward rendering is mostly dead, there will be more content using materials (glTF or Blinn-Phong) because, once everybody can see what's been out there, the market is going to shrink for diffuse-map-only content, so the grid itself is going to use more resources to render.

[Zalificent Corvinus]

47 minutes ago, Qie Niangao said:

There's no particular reason that glTF materials should use more memory than full Blinn-Phong materials

So true.

My living room wall has a pair of polished brass wall lamps. They were made, copy/mod, with flat matte 2003 style textures.

I added a 32x32 px blank white spec colour map, a spec colour tint in a pale brass colour, for that "metallic shader" effect, and there is no normal map.

The increase in VRAM usage is miniscule.

This is exactly the sort of thing real PBR is actually good at.

A PBR version of that brass lamp, wold use a 32x32 px ambient texture, in a brassy colour, a 32x32 px effects maps for the blank shiny "roughness", white "metalness" and a black "emmisive". The only 1024s on the thing would be the normal map for the decorative embossing, and maybe an AO channel, though that could probably be a 512 on a small item like a wall lamp.

 

The VRAM problem with PBR won't be inherent, just people not understanding how PBR is supposed to work, the ones who don't RTFM.

 

[Henri Beauchamp]

1 hour ago, Qie Niangao said:

There's no particular reason that glTF materials should use more memory than full Blinn-Phong materials (how often will the emissive mask be used, really? and for that matter, how many albedo maps should just be a 32x32 blank texture with an applied tint, if the model's surfaces are sensibly partitioned for materials?).

I'm afraid the theory won't apply in practice: in SL PBR materials can bear up to 5 textures: the diffuse texture (that should be populated/used, and allows non-PBR viewers, which will likely include the future mobile viewer, to see something else than a white blank face), plus the four ”base color”, ”normal”, ”metallic roughness” and ”emissive” maps, against only 3 textures for legacy materials (”diffuse”, ”normal” and ”specular” maps), and one (diffuse) for the forward renderer.

Since there is no limit either to the texture upload size for all those maps (which can all be 1024x1024 pixels wide), and even though the renderer can use smaller maps (via LODs) to render them (LL's PBR viewer uses this trick, and I further refined it myself), the memory usage (RAM, for sure, VRAM, most likely as well) is higher in practice (especially considering how SL creators love to use high resolution textures everywhere, even when they are just a total wastage of memory and won't change a thing on the final product rendering quality).

Add to this more render buffers needed for PBR, reflection maps, etc, and the PBR renderer appears like a gluttonous pig: graphics cards under 8GB of VRAM will have trouble rendering PBR-heavy contents at 256m draw distance, when it did not cause any issue with forward or even ALM renderers.

Edited December 16, 2023 by Henri Beauchamp

[Cinnamon Mistwood]

3 hours ago, Zalificent Corvinus said:

Real Games Dev setups, if they come up with a "cool new feature", they assign a group of coders to it, 8 hours a day, 5 days a week, and get it ready for a public beta test in 3 - 6 months.

 

3 hours ago, Zalificent Corvinus said:

How about prioritising the ACTUAL problems, like TP-fails, inventory loss, cloud induced performance degradation, bugs left unfixed for over 5 years,

In that entire insult laced rant, I picked out these 2 nuggets that might be constructive. The rest of it was pointless drivel. Also, I didn't mean to put you on the defensive and rile you up. *offers you a nice lavender tea*

How big a group devs, in your opinion, is needed?  Would they need to run their ideas past you to see if you personally deem it worthy of their time? 

Could you elaborate on how you would fix each of these problems?  Just a few details beyond how they should hire group of competent people. It's just a little too vague too be helpful.  That's probably too off topic for this thread about PBR, but it gives you an out to not actually answer anything with details and an in for more insults that don't actually fix anything.

 

Edited December 16, 2023 by Cinnamon Mistwood
Spelling

[Scylla Rhiadra]

34 minutes ago, Henri Beauchamp said:

I'm afraid the theory won't apply in practice: in SL PBR materials can bear up to 5 textures: the diffuse texture (that should be populated/used, and allows non-PBR viewers, which will likely include the future mobile viewer, to see something else than a white blank face), plus the four ”base color”, ”normal”, ”metallic roughness” and ”emissive” maps, against only 3 textures for legacy materials (”diffuse”, ”normal” and ”specular” maps), and one (diffuse) for the forward renderer.

I'm still, as a non-techie, a bit confused about this.

Is not the PBR "base colour" map essentially the same as the legacy "diffuse" map? Except maybe without baked-in AO (i.e., baked-in shadow)?

[Luna Bliss]

Sprinkle some humility onto that lavender tea.

[AmeliaJ08]

22 minutes ago, Scylla Rhiadra said:

I'm still, as a non-techie, a bit confused about this.

Is not the PBR "base colour" map essentially the same as the legacy "diffuse" map? Except maybe without baked-in AO (i.e., baked-in shadow)?

Yes as far as I know "base color" is the same as diffuse

[Henri Beauchamp]

4 minutes ago, AmeliaJ08 said:

Yes as far as I know ”base color” is the same as diffuse

Not at all.

[Scylla Rhiadra]

Just now, Henri Beauchamp said:

Not at all.

Ok, sorry, can you elaborate? What is the difference?

Say I'm applying a "base colour" map to a framed picture with only one face. Does that map not include, say, the picture itself, as well as the textures of the frame, front, back, and side. Just as a diffuse map would?

[Henri Beauchamp]

15 minutes ago, Scylla Rhiadra said:

Ok, sorry, can you elaborate? What is the difference?

Color encoding and alpha differ, for a start, then the other maps mix to the base color texture, so the final aspect (including colors) will often be very different (unless your object does not need at all PBR to render in the first place). I cannot elaborate here (too complex). Have a look at the specs for full details.

Edited December 16, 2023 by Henri Beauchamp

[Scylla Rhiadra]

2 minutes ago, Henri Beauchamp said:

Color encoding and alpha differ, for a start, then the other maps add to the base color texture, so the final aspect (including colors) will be very different. I cannot elaborate here (too complex). Have a look at the specs for full details.

Fair enough: I can hardly expect a primer on this. But, speaking (as I noted) as a non-techie, this is mostly gobbledy***** to me.

What you seem to imply, among other things, though, is that I am not going to be able to create my own textures for things anymore. That the creation of "base colour" maps is a very different process than the creation of diffuse maps (which I can now produce with relative easy in Photoshop or even some other simpler graphics program)?

So, in practice, this is yet another instance of a technological leap in SL having the de facto effect of excluding non-specialists from creating a hitherto accessible form of content, much as the introduction of mesh made obsolete building with prims?

This is not a merely academic question: I texture stuff all the time in SL, and in particular import "digital art" for presentation and sale in-world. So, will the eventual obsolescence of "diffuse" maps mean that I will no longer be able to do that without more specialized knowledge and/or software?