[LL Reaches Out on PBR]

27 minutes ago, Cynite00 said:

The ROI is simply no longer there for the amount of skill, time and dedication required to put out a good product for the platform, for pennies in return.

This has been SL since the mid 2010's at least. Unless your one of the "super-brands" if your here purely to make money, your gunna have a bad time. You could do almost anything else with your time and skills and make more money. Which is a shame.

A few people are happy to pay for quality, but we are few and far between. Logically, PBR content should be more expensive as it takes more time to make.

[LL Reaches Out on PBR]

25 minutes ago, HeathcliffMontague said:

It does for a lot of folks. Not only folks on old office Dells, as so many here like to say, but folks on medium setups with mid GPUs as well. Setups that actually, usually, handle  AAA games at least OK. It also makes stuff look better, in many cases. It's a very mixed bag. But why so many refuse to acknowledge the former is beyond me. 

While I would not say I get performance loss per se, I notice that freecam "stutters" a lot when you move it around. I get bad texture thrashing (which we established in another thread is most likey "intended behaviour") I get solid FPS but my 3090 constantly sounds like its about to levitate my PC off the desk and I dont have to put the heating on in my computer room anymore.

And my PC eats everything but the newest AAA games for breakfast. Max settings all the way, all day. The only thing that seems to stress it out is SL.

[Fake mesh people in scene creation - love 'em or hate 'em?]

Personally, while I think they can add to the vibe when you first arrive at a sim, unless they actually *do* something they go to look awkward and weird very quickly.

19 hours ago, Carolyn Zapedzki said:

Please don't disrespect my barman, he maybe old, slow and overweight, but he is honest and reliable and never answers back, kind of like my ideal man

Poor guy appears to be having an existential crisis.....

[LL Reaches Out on PBR]

20 minutes ago, Porky Gorky said:

in my reflective gimp suit

Pics or I dont believe you.

[LL Reaches Out on PBR]

43 minutes ago, Porky Gorky said:

There seems to be a mindset, at least on this forum, of being all in on PBR or all out. The best approach, at least aesthetically, is to use both BP and PBR together as each has its strength and weakness depending on the type of surface. I am not sure how marketable this approach will be for the “all in” PBR content creators though.

I think you may be misinterpreting that slightly. Its not that the people here (for the most part) think like that. We are pre-empting what we think the majority will do/think. Like the "all in" creators you mention. The forum folk here have generally been around at least for a while and have seen previous disasters so were all a bit jaded at this point I guess.

For my part at least, my recent foray back into the forums is to learn as much as I can about PBR, including when it is appropriate to use them and when not. Most people (I know) dont have much of a clue about this sort of thing at the moment. 

I guess the main concern in all this that most people have is that the majority of users are going to be extremely confused by this for quite some time. Most SL users do not use the forums and rely on info given to them when they purchase an item or in the items marketplace description.

[LL Reaches Out on PBR]

23 minutes ago, Porky Gorky said:

I would suggest that PBR usage is more widespread than you realise and not everyone is sat around waiting for content creators to build new PBR content. I sell PBR materials and take any opportunity I can get to elicit feedback from customers, with the first question I ask being, what are you using this PBR material for? A common usage is that people are just updating existing content with PBR, literally one surface at a time. They try a few materials out on a surface, see what works best and seem to be having fun updating their old houses. There is also some frustration with people replacing diffuse textures with PBR materials on cut prims without understanding the implications etc, but on the whole the feedback is positive.

I dont doubt that, that is exactly what I set out to do recently too! I gave up on the grass/terrain textures, because as it stands, I don't like anything that is for sale (that I have found) or that I can create myself, and as I mentioned elsewhere in the thread, I just don't see the need for it in that specific context. Old BP looks great and does the job nicely enough. But that does not mean I am giving up entirely and entirely happy to be wowed by a creator who proves me wrong.

But even if people are updating "one texture at a time", that process will still take a while. I look around my homestead and there must be a few hundred textures here at least.

Many no mod items.

Many items with baked AO that would look extremely goofy with a seamless PBR texture slapped on them, and dont have the UV mapping to support a seamless texture anyway.

Items that COULD be retextured if they were bundled with their AO maps, but they aren't.

There are lots of potential roadblocks to retexturing things.

[LL Reaches Out on PBR]

I have seen......mixed behaviour in the firestorm support and testing channels. Testing seems to be more for the grownups, but like any group of people who gather on the internet, its going to depend on who is online when you ask your question as to the quality of responses that you will get. Every mod I have dealt with seemed nice enough.

I get the feeling that for at least a year or two, PBR will mostly be the preserve of small skyboxes and a few showcase worlds. I am sure there are going to be a few people who are prepared to go all-in, but its going to take time for enough content to become available on the marketplace for it to become feasible to landscape and fill an entire sim or homestead/large parcel.

[LL Reaches Out on PBR]

Side note - can you believe LL now have us using the word "probe/probing" like its not an extremely creepy word?

"I might hire a prober to come and give my caves a proper professional probing later this week."

Ok.......Use plenty of lube and make sure there are no kids around please!

[LL Reaches Out on PBR]

13 minutes ago, Charlotte Bartlett said:

not what you post was about - but I thought it may be fun to show what can be painful with probes (no aliens involved).
Some of this is user error/learning but shows the journey.

1. Irregular shaped rooms because nobody just wants to live in a perfect rectangle without interesting architecture.  So you use a method to coat inside, or overlap (blending was broken on cubes during the first LL production viewers - I entered that bug and gave houses for testing to LL).  The reality is blending is a limitation on probes as the viewer interpolates between the probes and you can and will get artefacts at times.
2. Manipulating probes highlighted.  Firestorm did help a little as they added the reflection probe selection into the edit build menu v having it on the drop down build options but it is hard to do  inside a house in terms of selecting the right thing.  
3. Probes bug out (still) so even if you think your probes are done, you teleport away come back and see something is either conflicting, is it real or not.  Spoiler it can be both.
4. Probes don't lock in and it looks like a rectangle is missing, so you toggle back on to check (all the time whilst cammed inside the house to manipulate the probe in situ) then find out its the viewer not the probe or wait the next time it WAS the probe.
5. Probes rez with physics / prim but with phantom preselected (at least on Firestorm).  Link your probe to the house (as you would)... guess what happens.  yes now you have to go into each in the link set and turn physics to none as physics materalise.   I am doing more testing on this to refine workflow.
6. Dependent on the EEP being used to test during probes, it can have different challenges dependent on what probes and HDR settings are in place.   Customers not using new EEPs then have more issues (it's why I now do EEPs).
7. Probes can behave more wonky at height (bug I raised also now fixed to a degree) but can sometimes still have moments if on a build platform.
8. The biggest one - You can't have a mesh cube or sphere probe, it has to be a prim cube (or sphere)  you then connect to a build after the fact.  So you cannot create the cubemaps/probes aligned to the build during upload then configure those meshes already in place to be probes in world and appropriate physics.  Now you are trying to align prims into a mesh build, cammed inside a build, adding to link set, fixing the fact then the probes want to have physics as it's a prim not mesh uploaded with the physics model.  Again all the while why cammed into a build.
9. You can only grab a probe using edit object move.  If you are mass working on builds and are on stretch, or something else it doesn't let you select the probe as the selection method is a hack around.  More clicks, constant moving back and forth on select reflection probes versus not.  Turn on transparent versus just highlight to help, but the click can often fail.
10. No real materials or SL University videos (beyond one that was incorrect on how to upload PBR with the right opaque setting) to show creators what best practice is, what is optimization, what is really appropriate in usage etc. (Hint the Unity documentation is very helpful to translate into SL).
12. As the reflection probes are generated on prims with what we can call "cubemaps", seams and distortions can and will occur, especially when the probe's influence extends over large areas or when multiple probes blend.  There is only so much a creator can do on that and it's a limitation.  However,  if customers experience a seam will come back to the creator wanting it fixed.  I have tested multiple blend ratios during the blend bug I raised to get the optimal outcomes, but it's effectively a limitation at the interpolation level.
13. Once you are finished with probes set up and this is not necessarily quick for a complex piece of architecture, then test, test EEPs, rule out bugs, adjust to reduce artefacts etc.   
14. It's MORE WORK so much more work.  I am playing with optimization e.g. using probes on windows (glass) versus brick walls that slightly glisten - that nice look can "go".  ,Using mesh insets on building to push out external walls, so probes can slam with interior walls but not "extrude" externally.   Each build now mesh  with a dummy probe set up so I can use auto position in Firestorm to swap out each mesh dummy probe to a prim to align quicker - etc etc - but that's my own enhanced workflows I am adopting to help - no customer can be expected to deal with that cluster and of course there is no workflow best practices guidance available.
15. A customer unlinks something not realising a probe is there, moves the build and now it's broken on probes.
16. A customer doesn't know what probes are or why they need EEP (that's a huge education gap as again the standard person in SL doesn't want to muck around with probes or environments) - they have their favourite and they are done on lighting settings.
17. Why are there pink rectangles all over my house - why is half the region pink rectangles.  Wait now it's calming down.  

There is more SO MUCH MORE, but a few of the examples of why they are difficult and painful for (me).

 

So simple! What's the problem?!?! 

EDIT - But seriously, thats a very helpful post. Thank you!

[LL Reaches Out on PBR]

4 minutes ago, KanryDrago said:

I think the chasing of blingy graphics has affected more than sl however. Games seem in general have substitued blingy graphics for engaging game play, like films seem to have sacrificed actual storytelling in favour of blingy CGI

Every game in recent memory.......except Baldur's Gate 3.  Gosh I love that game. But it really highlighted for me how most of the other "big" games I had played in the last few years were vapid, linear, boring, low effort in terms of gameplay and story......but you know, very nice graphics.

Indy games seem to be the place to go to get actually emotionally engaging content these days.

[LL Reaches Out on PBR]

2 minutes ago, Luna Bliss said:

We've yet to see, fully, what PBR will offer though.

I already see changes I like with the light, even when content isn't PBR, so I'm looking forward to even greater changes.

Sure, if most couldn't see the difference it wouldn't be worth the change, but I don't think that will be the case...eventually.

I do really love what the new lighting model brings to regular old BP textures. Materials that you could barely see previously now "pop" in a very satisfying way. I don't want to sound like some curmudgeon who only complains. Avatar shading looks way better too.

3 minutes ago, Luna Bliss said:

If I were you I'd just place a typical light in there for now until we understand all this better and better instructions are available.  No need to get all stressed out if you can avoid it 

It doesn't stress me out, sorry if my tone implies that.  I just get a bit "headless chicken" when I don't fully understand things and dislike not being able to get to a definitive answer. No stress, just my nature. I come in like a wrecking ball sometimes.

[LL Reaches Out on PBR]

4 minutes ago, AmeliaJ08 said:

Everything must stay the same! Nothing can change! I don't like learning!

I dont think anyone is saying that. People are moaning about the implementation/execution of the change, not the change per se. Or, speaking for myself at least.

[LL Reaches Out on PBR]

3 minutes ago, AmeliaJ08 said:

But you don't? you'd need some for your interior spaces, maybe a few to cover the outdoors?

I think there's maybe some confusion here, each object doesn't need its own reflection probes. Some are coming with them because the assumption is people have not already placed their own more generalized probes on their parcels.

I was assuming to equally light an entire homestead, I need to cover the entire surface (which I use for the most part) with probes. Is that not the case?

And as each probe an only be as big as a max-sized prim, and they are spherical so need to overlap......

Then I have some caves, so theres another bunch that I will need. 

Or am I grabbing the wrong part of the donkey here?

Creators attaching probes to every object sounds like a really, really bad idea.

[LL Reaches Out on PBR]

11 minutes ago, Flea Yatsenko said:

If you are making interior spaces and you do not actively think about filling the area with cube or sphere reflection probes you are going to have a bad time. It should be the creator's responsibility to set up reflection probes for customers. One of the best ways to learn is to study an example and getting a proper reflection probe build and dissecting is is a great place to start.

6 minutes ago, Charlotte Bartlett said:

House sellers yes and we do this as it's like having needles poked in your eyes working with probes in a house.  I would not wish that on any customer.

Furniture sellers (some of them and a certain large well known one) have taken to also adding probes to every piece of furniture.

Combine those two.....   

Exactly, if everyone is adding reflection probes to EVERYTHING things are going to get funky. 

What about landscaping and outdoor stuff? Does every object sold now need a probe attached? Wont that cause conflicts if they bump into each other?

 

1 minute ago, AmeliaJ08 said:

That is impossible.

The engine has to know what it is supposed to be reflecting. You can of course automatically place probes (LL have done so) but this will always be imperfect, if you want the engine to have as accurate information as possible it is necessary to place probes.

It's not really painful, I'm not sure what is being considered so difficult about it either.

I would not say its "difficult". More annoying, time-consuming and fiddly. And also you need 50-100 extra prims free just for probes (if you want to cover a homestead). Since they cant be any bigger than a prim.

5 minutes ago, Paul Hexem said:

The point of using the word "puzzle" is that it's not complete out of the box, that's not how they work. You can't just roll it out when it's "all done". That's not how SL works. No live service works that way after initial release.

You roll out each feature as best you can. Granted LL is worse at it than the average developer, but if that's a surprise to you with PBR then you weren't here for BOM, mesh, ALM, sculpts, mono LSL, the Havok updates... Virtually anything LL has done ever.

Actually, yep, was here for all of those. Check my forum post history, I have a long rich history of bitching at LL.

Ill concede your first point, kind of. I still think a change on this scale should have been better thought through before releasing it onto the grid. Especially an environment as dynamic as SL.

[LL Reaches Out on PBR]

22 minutes ago, Paul Hexem said:

Of course the addition of PBR looks bad when you're limited to such a shortsighted view, or like others in this thread think it's just about mirrors.

Everybody seems to be skipping over the fact that this is just the first step in a serious modernization of all the creation processes in SL. PBR, gLTF, LUA code... These are all pieces of a much larger puzzle.

My apologies, I guess Im just too dense to see "The Vision". Stupid me. But I also think your deluded if you think the average resident gives a flip about that.

Think you might also have skipped this part of what I was saying:

1 hour ago, AnnabelleApocalypse said:

The best thing they could have done is keep this in the oven a bit longer. In principle, its a nice update but in practice its a fustercluck.

Finish the puzzle before you ask people to guess what the picture is supposed to be. Keep it in Beta till its ready. Ok, maybe you cant finish the WHOLE puzzle in one go, but at lease finish the section so it makes sense.

 Or are we supposed to deal with a borked, half-finished, buggy, poorly optimized mess for......? 6 months? A year? Longer?

On the promise that were just not seeing the big picture and it will be amazing in the future? So we should just keep paying our land fees an live in brokenville till LL get their moles in order?

[LL Reaches Out on PBR]

33 minutes ago, Marianne Little said:

I do not hate PBR, but frankly, the grass, moss and other nature textures in my Studio-Skye building packs look better than their Zen deck building set in only PBR. Polished granite, lacquered wood look good in the set, but the more I look at the moss, I prefer how Blinn-Phong textures look.

Exactly my conclusion too. I tried the new Skye texture bundles and really liked all the textures where you would expect reflectivity, but ALL the natural terrain textures looked "shrink-wrapped" in plastic. it might be "correct" technically, but artistically, in the context of Second Lifes current lighting model - they look like ass.

No disrespect to Skye, long may he reign. But if one of our most talented content creators cant make it look good, were all in trouble.

I was going to just carry on using my BP textures, but what is all this talk of eventually everything appearing white/blank? What fresh hell is this?

Don't you DARE turn off BP textures LL. Even you cant be that stupid. They work, leave them alone! If they do that - and invalidate 20 years of content - I wont be hanging around.

[LL Reaches Out on PBR]

The biggest "thing" for me in this whole debacle is - Its a LOT of additional effort and messing around for.......slightly nicer shadows and reflections. The effort required seems to far, far outweigh the payoff.

 

The best thing they could have done is keep this in the oven a bit longer. In principle, its a nice update but in practice its a fustercluck.

 

Elsewhere on the forums I have been asking for help with a PBR grass texture and my + the collective conclusion seems to be "yeah PBR (the broken, not actually PBR we have in SL now) as it currently stands sucks for this application, better to keep using Blinn-Phong for that".

 

Its taken several DAYS for me to tinker, google, watch tutorials, ask questions and come to that conclusion. For a single grass texture. This is more effort than its worth and more than the average user is going to put in.

 

But the more I thought about it, the more I came to the conclusion (IMHO) that the average SL user just WONT do this.

 

Lets say you rent a homestead (like me) and you want to do a PBR update (also like me 2 weeks ago). BP textures already look pretty nice under the new daylight model we have, no probes needed.

 

So, am I now going to COVER my sim in reflection probes, update ALL the textures and/or replace content just so I can have slightly more accurate specularity on my trees/grass/plants? No, no I am not.

 

I think also a lot of people dont seem to realize that unless your object has metallicity, there is very little (possibly zero) benefit in updating to PBR textures vs BP. Things like grass, bark, concrete just dont really need it. Why add significant extra texture load for minimal benefit?

 

EDIT - If PBR worked as intended elsewhere - adding these additional maps WOULD be beneficial, but as the lighting model in SL stands, its not.

 

PBR terrain textures seem to be a massive red herring. Totally unnecessary as things stand. BP textures with a subtle specular map is all that's needed.

[Discrepancy between Substance Sampler/Painter and inworld PBR - Too shiny!]

I've been doing more experimenting with this and came to the conclusion that Blinn-Phong does indeed still look better for grass, plants and the like.

Being able to tint the specular channel is a huge help and (I am sure there is a reason) its very frustrating we cant do that manually inworld with Metallicity/Specular in PBR.

You just cant kill the "shiny" enough with the standard Roughness and Metallic fields provided. Crunching it down in Substance Painter with a "levels" makes it better, but it still looks weird and "plastic-like". As Remi mentioned everything looks "vaccum wrapped".

I also went ahead and purchased some of Skyes new PBR terrain textures, and found those look kinda weird and shiny too. Much better than my attempts, but not something I would....you know.....actually want to use. So if my favourite landscape creator in SL cant make something I like, I doubt I stand much of a chance. lol

I dont think BP textures are going away any time soon because if there is one thing I know about SL users - we HATE things being shiny when we dont want them to be (based of our own history and standards, not what is "correct"). And items with Diffuse/Norm/Spec actually look pretty good under the new lighting system, no need for special EEPS or lighting probes.

And as we need to use spherical reflection probes in a world of square land parcels (cube ones are buggy, apparently), I get the feeling a lot of residents are just not going to bother. I dont consider myself a particularly technical user, but I have more of an idea than many, and it all just seems like a LOT of extra faff for not much payoff.

Is the average person really gong to f-around this much just to have a working mirror and some reflections here and there? I kinda doubt it.

That's also going to be a huge headache for people making content. Do you do it "the right way" and get a load of 1 star reviews? Or how customers "expect" it to look and screw standards?

[Since PBR - Awful texture thrashing in LL viewer and FS.]

On 8/16/2024 at 1:23 AM, NiranV Dean said:

Um.

The texture thrashing (if its the same i see) is actually intended. (Yes its stupid)

Thats so stupid, I think its the most plausible explanation. lol

Shouldn't it only do that when my memory is in danger of getting full at least? I can literally stand in an empty region with nothing more than a half dozen textured objects (1024 textures) and its sill triggers like crazy. 

If they dont want to make a setting, at least tie it to may actual memory usage. If this was happening because my VRAM was capping, I could understand. But I'm nowhere even close.

[Discrepancy between Substance Sampler/Painter and inworld PBR - Too shiny!]

3 hours ago, Gabriele Graves said:

We have no idea of true numbers of people happy or unhappy.  The only conclusion we can draw is that it has made more people unhappy than some of the previous things LL have introduced but to be fair there seems always to be a fair number of unhappy people no matter what changes.

That's fair I suppose, but the same could also be said of your point. We don't actually know how many content creators want to do what with their time. I don't think any extensive survey has been done.

But my point was not really about that. I don't disagree with anything you said, but what I am saying is separating it by regions would have made the transition easier and smoother for the userbase and content creators too. Let people come at it in their own time, to a degree. Over time I imagine that legacy regions would be put to bed for good as if PBR content is so game-changing, people will want to move over.

But now we are going to have a mish-mash of both types of content for quite some time. Which is much more disruptive for everyone involved.

Anywho, I think we may be in danger of going off the rails here folks. Lets not get mauled my the Moles. No point stewing over something thats not going to change. At the end of the day......my grass is looking much nicer. That's the main thing (for me anyway).

[Discrepancy between Substance Sampler/Painter and inworld PBR - Too shiny!]

15 hours ago, Nagachief Darkstone said:

That was indeed the reasoning to modify ACES a bit so that it's not as harsh on existing content.

I understand the desire to compromise, but it feels like we have inadvertently gotten an outcome few are actually "happy" with. I know I am a bit snarky towards the lab, but I do sympathise with LL too. Its incredibly difficult to make changes like this without annoying people no matter what you do.

I still don't really get why we could not have PBR (TRUE PBR) regions, and non-PBR "legacy" regions. LL could probably have even charged a bit more for them, which seems like an idea they would be very happy with. I cant imagine it didn't come up at the board meetings. That way PBR people can PBR without changing workflows and everyone else can just get on with their virtual lives and pick it up when they are ready to and when there is enough content to actually create a full PBR region that the user would enjoy.

I've been told its "not a popular idea" on the forums (whatever that actually means) but I cant understand why that would be the case. This way just feels rather forced on everyone and its sour grapes all around. I mean, cant really change it now but it just boggles my brain. Its like they picked the method that would maximize annoyance. 

[Discrepancy between Substance Sampler/Painter and inworld PBR - Too shiny!]

21 hours ago, Zalificent Corvinus said:

What you do is disguise it with shrubbery. Joins between mesh rocs, mesh terrain layers, etc., have always been clunky, and always will be, PBR won't cure that.

Sometimes you can minimise the effect by ignoring the UV mapping and setting it to "planar" instead of default, so both pieces are scaled, rotated and transformed contiguously., but the less FLAT the pieces are, the more likely the rather cubical "planar" mapping will mess up and create nasty interference patterns on corners.

 

I know the usual drill, I was more making the observation that its even more noticeable now than previously. 
 

21 hours ago, Zalificent Corvinus said:

Essentially, NO.

SL's "PBR" is not an actual full PBR system, its a badly broken half finished mess of PART of a PBR system.

Pretty standard LL really. Your the fool for expecting something that actually works as intended. 

[Discrepancy between Substance Sampler/Painter and inworld PBR - Too shiny!]

So I managed to find a hacky way to get what I wanted. Pretty sure now it is something I am doing with my normals in Substance Designer. Basically, I binned the normal maps from SD and took the color and height map to Materialize (its a free program for making PBR) and generated the normal map that way. By setting it to the "mids" setting, I was able to get something close to the result I was after.

Still needs some tweaking, but overall, a much nicer, softer, clumpier effect that doesnt look so aggressive up close. One thing I am noticing - putting materials on mesh landscape components really makes the overlap lines very obvious in some lighting conditions.

Yucky. So that's going to be fun to try to hide. So yeah.....seems I need to dig deeper into normal map generation in Substance Designer. Thanks for the help everybody!

 

[Discrepancy between Substance Sampler/Painter and inworld PBR - Too shiny!]

I have tried that, but what I find is the normal either looks non-existent or harsh and overblown, I cant seem to find any middle ground. Also, the map disappears at a distance regardless of its intensity.

I am guessing its something to do with how I am baking, but its difficult to know sometimes what is an error and what is "expected SL behaviour". 

I did also find this - https://blenderartists.org/t/normals-disappear/1243115/6 - that seems to describe a similar problem. Though talking about Blender rather than SP.

[Discrepancy between Substance Sampler/Painter and inworld PBR - Too shiny!]

I resolved my roughness woes, but I am still having a time with Normals. So, this is the original texture on a sphere in Substance Designer....its clumpy grass, basically. 

 

and here it is in Substance Painter with the SL HDRI....... Not the greatest grass ever made, but its mine and it looks like grass, so yay.

Far

But when I bring them inworld I get this strange effect where they are really harsh looking close-up, and fade to near nonexistence when I cam back. This is a 2048 texture applied to a 5m by 5m mesh plane under a lighting probe. See how in this pic, they become quite dark and apparent the closer to the camera they are and fade out the further away they are.

And when I cam away....they crumble into nothing very quickly.

I have double checked that I am outputting OpenGL maps. I get that the dark patches are caused/exacerbated by the "clumpyness", but why is it disappearing like that?