Jump to content
Medhue Simoni

LL Announced a Project to BAKE Textures on Meshes

Recommended Posts

2 minutes ago, Fluffy Sharkfin said:

I really can't think of any other possible reason for someone to suggest that it's the customers job to do the work of those employed by the company whom they're paying to provide a service!?

Isn't that how SL always has worked?

Seriously, I'm sure I'm not the only one who agrees with Penny that dynamically baked normal and specular maps would be absolutely wonderful. But can you imagine how much work it would take to debelop something like that? My head is spinning just thinking about it.

Share this post


Link to post
Share on other sites
1 minute ago, ChinRey said:

Isn't that how SL always has worked?

Seriously, I'm sure I'm not the only one who agrees with Penny that dynamically baked normal and specular maps would be absolutely wonderful. But can you imagine how much work it would take to debelop something like that? My head is spinning just thinking about it.

It is, but that doesn't mean we have to like or encourage it! :)

As for clothing layers and materials I agree completely, especially when it comes to normal maps since you can't just slap one normal map on top of another using the alpha channel as a mask, for a start the alpha channel of the normal map is, I believe, already in use as the specular exponent channel (I guess you could cheat and use the same alpha you're using for the diffuse though), but it would just look weird since the height differences of the clothing layers wouldn't be present automatically so they'd have to add additional height information to each layer in order to mimic the effect of layers of clothing.

The concept of unlocking the UV mapping on clothing layers so that instead of it always being set to 1:1 it could be scaled and then repeated or used as a decal may be equally complicated, but it's about the most basic addition to the new feature I could think of which would allow a greater scope of use and potentially greater impact when it comes to improving performance beyond just reducing the sheer number of polygons and textures being displayed on some avatars.

Share this post


Link to post
Share on other sites
50 minutes ago, Fluffy Sharkfin said:

Yes I caught the overall tone of your post but opted to ignore the snark since I assume it's a product of some weird sort of virtual world Stockholm Syndrome.  I really can't think of any other possible reason for someone to suggest that it's the customers job to do the work of those employed by the company whom they're paying to provide a service!?  Or are you suggesting that the continued viability and growth of SL as a platform is purely of benefit to its users and in no way is LL profiting from its existence?

With most other "companies whom you're paying to provide a service," when you ask them to make some major, impractical, ill-thought out change largely to make yourself happy you'll end up either having a pleasant conversation with someone in a cubicle in Bangalore or have your E-mail answered with the Form Letter to That Kind of Guy/Girl/Frog/Bovine.

At least with Second Life there's the possibility that you can do the work yourself.

Provided, of course, that you can do the work.

 

 

 

Share this post


Link to post
Share on other sites

 Like Fluff tried to explain to Theresa, the whole entire point of this feature is to make it so people stop making onion-skin avatars, which both add an extra layer of complication to mesh avatars, and are obscenely inefficient for rendering. You want to talk about things that hurt framerates, there's a big one right there.

39 minutes ago, ChinRey said:

Isn't that how SL always has worked?

Seriously, I'm sure I'm not the only one who agrees with Penny that dynamically baked normal and specular maps would be absolutely wonderful. But can you imagine how much work it would take to debelop something like that? My head is spinning just thinking about it.

Well, here's my view.

When LL introduced mesh import they had no intention of adding rigged mesh at all, because "can you imagine how much more work it would take to develop something like that?". But content creators insisted, and did not let up. So LL decided "ok, we'll give you rigged mesh but we really doubt anyone will actually use it, so we're not going to let rigged mesh use the full shape sliders, just the basic skeleton. Because, can you imagine how much more work it would take to develop something like that?" So we got the initial version of rigged mesh.

 A lot of us complained at how crippled rigged mesh was because of this. Some people said "Well, if you don't like it, go make your own virtual world because we should never call LL out for doing something wrong. We should just be happy with what we get." But we persisted, and thanks to that persistence, LL finally relented "Ok, we don't think anyone will actually use this, but we'll do a new version of rigged mesh called fitmesh." We said "Great, but it should also allow for facial/hand animations." To which LL replied, "No, because can you imagine how much more work it would take to develop something like that?"

 So, as time went on it became clear that fitmesh just wasn't cutting it, people insisted we get proper rigged mesh, finally. Eventually, after some years, LL relented. "Ok, we'll give you what you asked for in the first place." And so now we got Bento.

 By now LL has put a whole lot of extra work and money into developing rigged mesh, well beyond what  it would have cost them if they'd just done it right the first time.

On top of that, SL now has 3 different kinds of rigged mesh, and whenever a new SL user joins, they have to learn the difference between all three kinds of rigged mesh. A lot of people are going to join SL, see this for how stupid it is, and just leave. This is just one example of how, year after year, LL spends a lot of money and puts a lot of effort into making SL that much less attractive to new users.

  • Like 4

Share this post


Link to post
Share on other sites
27 minutes ago, Fluffy Sharkfin said:

The concept of unlocking the UV mapping on clothing layers so that instead of it always being set to 1:1 it could be scaled and then repeated or used as a decal may be equally complicated, but it's about the most basic addition to the new feature I could think of which would allow a greater scope of use and potentially greater impact when it comes to improving performance beyond just reducing the sheer number of polygons and textures being displayed on some avatars.

It would only work if the baked texture could be tiled though and I can't see how that can be possible for an avatar. Tiled seed texture will only give a marginal speed icnrease for the baking process and not affect the rendring at all.

As for the performance cost of polys vs textures, I think we should put that into perxpective.

I agree that for scene objects excessive numbers of resolutions of textures is the main cause of lag and it certainly adds a lot to avatar lag too.

But even so: A scene with about 300,000 active static polys (active as in actually rendered on the screen) is moderately heavy to render. You don't really want to go much higher than that and fortunately we hardly ever do so the lag caused by gemoetry is usually less significant than the lag caused by textures.

I use a Maitreya mesh body most of the time. It's a fairly low lag mesh body but it still has more than 300,000 polys all on its own. And since those are all dynamic polys, they are much harder for the gpu to manage than the ones wh just sit still and never move or change shape or size. Not all of those polys are active of course but enough of them are that they add significantly to the overall render cost of the scene.Three or four mesh avatars in a picture can easily add more to the gpu load than the entire scene they're in. With baked textures the polycount for mesh bodies can potentially be cut to a quarter of what it is today. Even if they retin separate layers for skin and clothes, it's still a 50% reduction of the polycount. That is a significant performance improvement.

Share this post


Link to post
Share on other sites
7 minutes ago, Penny Patton said:

By now LL has put a whole lot of extra work and money into developing rigged mesh, well beyond what  it would have cost them if they'd just done it right the first time.

Oh yes. Unfortunately, shortsighted development has always been a trademark of Linden Lab. I hope and believe they've learned the lesson now that a rushed jo always takes more work in the long run. I've sen a lot of talk about backwards compatibility on this forum. How about forwards compatibility?

  • Like 2

Share this post


Link to post
Share on other sites
12 minutes ago, Theresa Tennyson said:

With most other "companies whom you're paying to provide a service," when you ask them to make some major, impractical, ill-thought out change largely to make yourself happy you'll end up either having a pleasant conversation with someone in a cubicle in Bangalore or have your E-mail answered with the Form Letter to That Kind of Guy/Girl/Frog/Bovine.

At least with Second Life there's the possibility that you can do the work yourself.

Provided, of course, that you can do the work.

 

 

 

Yes, and if I'd come onto this forum and started a thread demanding some major, impractical, ill-thought out change purely for my own amusement/benefit then I could understand your response.  However the whole topic of this thread is the fact that LL have announced this new feature, and I'm simply trying to suggest ways in which it could be expanded on so that it may be of maximum benefit to everyone.

I really don't think giving people a link to the source code and telling them "go do it yourself" is adding anything constructive to the conversation.

I can understand how you may get tired of people who continually make unreasonable demands, but since i rarely even bother to post on these forums I don't see how you can justify acting that way towards someone who's basically a complete stranger and just trying to engage in a civilized discussion!

Share this post


Link to post
Share on other sites

For most of SL's history they have had a dream of who their customer should be that was not at all who their customer was.

If you imagine who their customer "should be" as they likely saw it... a lot of this doesn't look short sighted, rather it looks like they keep getting sidetracked down the path of unnecessary stuff their ideal customer had no interest in.

The problem is that that ideal customer is a very rare creature... and has largely moved on to investing in or watching demos of people who like to wear an octopus on their head... which is another market in search of a customer...

 

At this point in time Linden Lab has largely been on the path of making the actual product its real customers want, for the last few years. The question remains though; if they are even aware of this.

This project hints that they are... hopefully...

Edited by Pussycat Catnap

Share this post


Link to post
Share on other sites
33 minutes ago, Theresa Tennyson said:

With most other "companies whom you're paying to provide a service," when you ask them to make some major, impractical, ill-thought out change largely to make yourself happy you'll end up either having a pleasant conversation with someone in a cubicle in Bangalore or have your E-mail answered with the Form Letter to That Kind of Guy/Girl/Frog/Bovine.

At least with Second Life there's the possibility that you can do the work yourself.

I am routinely sitting in meetings at work where the conversation is roughly like this:

"Customer X wants new feature Y, here is budget Z to go make it. How many engineers might you need to get it done? We're scheduling a call with customer X, manager A, and you next Tuesday to over this."

Share this post


Link to post
Share on other sites
13 minutes ago, ChinRey said:

It would only work if the baked texture could be tiled though and I can't see how that can be possible for an avatar. Tiled seed texture will only give a marginal speed icnrease for the baking process and not affect the rendring at all.

Ahh, no I think you misunderstand my meaning.  The tiling of seamless textures would occur pre-bake so that the resulting baked texture would still be a single non-tiled texture, but any of the layers used to bake that texture could contain low-resolution, seamlessly tiled textures (in the case of a dress it could be patterned fabric, in the case of a static object a brick texture), then on top of that you could apply a single non tiled texture with windows, shadows, etc.. or even a texture as a decal with tiling turned off and the scale and offset set so it appears in a certain place.  This would mean instead of needing a 1024x1024 texture containing the repeated seamless brick texture complete with windows, shadows, etc. you could make do with a 128x128 seamless brick texture and a single window texture tiled several times for the windows, with a non repeating decal map for the door (or alternatively think plaid patterns, buttons, stitching, etc).  Basically a way to have more control over the way the textures of each layer are handled before the final bake is created and passed to the object.

 

21 minutes ago, ChinRey said:

As for the performance cost of polys vs textures, I think we should put that into perxpective.

I agree that for scene objects excessive numbers of resolutions of textures is the main cause of lag and it certainly adds a lot to avatar lag too.

But even so: A scene with about 300,000 active static polys (active as in actually rendered on the screen) is moderately heavy to render. You don't really want to go much higher than that and fortunately we hardly ever do so the lag caused by gemoetry is usually less significant than the lag caused by textures.

I use a Maitreya mesh body most of the time. It's a fairly low lag mesh body but it still has more than 300,000 polys all on its own. And since those are all dynamic polys, they are much harder for the gpu to manage than the ones wh just sit still and never move or change shape or size. Not all of those polys are active of course but enough of them are that they add significantly to the overall render cost of the scene.Three or four mesh avatars in a picture can easily add more to the gpu load than the entire scene they're in. With baked textures the polycount for mesh bodies can potentially be cut to a quarter of what it is today. Even if they retin separate layers for skin and clothes, it's still a 50% reduction of the polycount. That is a significant performance improvement.

I'll admit I don't have enough technical knowledge to be able to predict which would cause more lag, though I suspect it depends on the type of performance issues you're talking about.  While high polygon count would probably cause more of a slowdown when rendering, wouldn't a high volume of texture data cause more of an issue with ram than number of polygons?

Share this post


Link to post
Share on other sites
10 minutes ago, Pussycat Catnap said:

I am routinely sitting in meetings at work where the conversation is roughly like this:

"Customer X wants new feature Y, here is budget Z to go make it. How many engineers might you need to get it done? We're scheduling a call with customer X, manager A, and you next Tuesday to over this."

And how much do these customers pay your company? I'm guessing it's probably enough to cover budget Z, manager A and you.

Share this post


Link to post
Share on other sites
50 minutes ago, Penny Patton said:

 Like Fluff tried to explain to Theresa, the whole entire point of this feature is to make it so people stop making onion-skin avatars, which both add an extra layer of complication to mesh avatars, and are obscenely inefficient for rendering. You want to talk about things that hurt framerates, there's a big one right there.

Well, here's my view.

When LL introduced mesh import they had no intention of adding rigged mesh at all, because "can you imagine how much more work it would take to develop something like that?". But content creators insisted, and did not let up. So LL decided "ok, we'll give you rigged mesh but we really doubt anyone will actually use it, so we're not going to let rigged mesh use the full shape sliders, just the basic skeleton. Because, can you imagine how much more work it would take to develop something like that?" So we got the initial version of rigged mesh.

 A lot of us complained at how crippled rigged mesh was because of this. Some people said "Well, if you don't like it, go make your own virtual world because we should never call LL out for doing something wrong. We should just be happy with what we get." But we persisted, and thanks to that persistence, LL finally relented "Ok, we don't think anyone will actually use this, but we'll do a new version of rigged mesh called fitmesh." We said "Great, but it should also allow for facial/hand animations." To which LL replied, "No, because can you imagine how much more work it would take to develop something like that?"

 So, as time went on it became clear that fitmesh just wasn't cutting it, people insisted we get proper rigged mesh, finally. Eventually, after some years, LL relented. "Ok, we'll give you what you asked for in the first place." And so now we got Bento.

 By now LL has put a whole lot of extra work and money into developing rigged mesh, well beyond what  it would have cost them if they'd just done it right the first time.

On top of that, SL now has 3 different kinds of rigged mesh, and whenever a new SL user joins, they have to learn the difference between all three kinds of rigged mesh. A lot of people are going to join SL, see this for how stupid it is, and just leave. This is just one example of how, year after year, LL spends a lot of money and puts a lot of effort into making SL that much less attractive to new users.

Fitted mesh and Project Bento were both largely developed by users.

Share this post


Link to post
Share on other sites
2 minutes ago, Theresa Tennyson said:

And how much do these customers pay your company? I'm guessing it's probably enough to cover budget Z, manager A and you.

Let's keep this in perspective. No one is asking for a special feature just for them.

LL announced a new feature they intend to add to SL. This feature has a specific purpose, to allow content creators to stop making laggy "onion-skin" avatars by letting clothes, tattoos, etcetera all be baked directly to the mesh body's surface as a single texture (or set of textures, as the case may be).

The problem some of us pointed out is that materials are such a standard feature it's inconceivable that anyone would be willing to forgo having them on their mesh bodies. This is going to limit the usefulness of this feature towards its intended goal . Either people will not want to use the baking feature because they want their skin and clothes to have materials, or they will have to stop using materials because your clothes are going to look really strange when your body's materials are rendered overtop of them.

If LL goes forward with this they will feel pressured by the community to fix the issue, in which case they can either do what they did with rigged mesh and spend a whole lot of extra money and effort fixing their mistake, or they can do like they did with pathfinding and let the feature die on the vine.

3 minutes ago, Theresa Tennyson said:

Fitted mesh and Project Bento were both largely developed by users.

Even if this were completely correct it misses the point entirely, or do you honestly believe SL has not benefitted at all from Fitmesh and Bento? I know it can get irritating seeing people complain all the time but some of us actually do understand how certain aspects of SL work and can see when LL is doing something detrimental to themselves and the userbase. If we stayed silent on that you wouldn't have Fitmesh, Bento, avatar masks, alpha masking, or a host of other features you directly benefit from even if you aren't aware of it. 

  • Like 1

Share this post


Link to post
Share on other sites
34 minutes ago, Theresa Tennyson said:

And how much do these customers pay your company? I'm guessing it's probably enough to cover budget Z, manager A and you.

Quite frankly every customer pays a companies budget. When you're in a smaller firm, you have marketing people who's very job is determine what customers want. In the past I've run the analytics for a company - and sat with marketers all day long going over reports of how people engaged with the product online and off - to figure out what was needed next.

I don't know where this mythical world of companies that can get by with ignoring customers is... it frankly isn't even here in SL.

 

Share this post


Link to post
Share on other sites
6 minutes ago, Pussycat Catnap said:

Quite frankly every customer pays a companies budget. When you're in a smaller firm, you have marketing people who's very job is determine what customers want. In the past I've run the analytics for a company - and sat with marketers all day long going over reports of how people engaged with the product online and off - to figure out what was needed next.

I don't know where this mythical world of companies that can get by with ignoring customers is... it frankly isn't even here in SL.

 

There was a beverage company that had slowing sales, so they developed a new formula and asked 200,000 people whether they preferred the old or the new formula. They then rolled out the new formula, because that's what the majority of the testers told them they preferred.

AAANNNNDDDDD... we got New Coke. Briefly.

(Admittedly it did create a buzz and cleverly masked the change of "classic" coke from sugar to high-fructose corn syrup, but still...)

Share this post


Link to post
Share on other sites

Don't try to extract a 'standard' from a famously bad example of how to do things... I believe the term for that is a straw man argument.

Edited by Pussycat Catnap
  • Like 2

Share this post


Link to post
Share on other sites
15 minutes ago, Penny Patton said:

Let's keep this in perspective. No one is asking for a special feature just for them.

LL announced a new feature they intend to add to SL. This feature has a specific purpose, to allow content creators to stop making laggy "onion-skin" avatars by letting clothes, tattoos, etcetera all be baked directly to the mesh body's surface as a single texture (or set of textures, as the case may be).

The problem some of us pointed out is that materials are such a standard feature it's inconceivable that anyone would be willing to forgo having them on their mesh bodies. This is going to limit the usefulness of this feature towards its intended goal . Either people will not want to use the baking feature because they want their skin and clothes to have materials, or they will have to stop using materials because your clothes are going to look really strange when your body's materials are rendered overtop of them.

If LL goes forward with this they will feel pressured by the community to fix the issue, in which case they can either do what they did with rigged mesh and spend a whole lot of extra money and effort fixing their mistake, or they can do like they did with pathfinding and let the feature die on the vine.

Even if this were completely correct it misses the point entirely, or do you honestly believe SL has not benefitted at all from Fitmesh and Bento? I know it can get irritating seeing people complain all the time but some of us actually do understand how certain aspects of SL work and can see when LL is doing something detrimental to themselves and the userbase. If we stayed silent on that you wouldn't have Fitmesh, Bento, avatar masks, alpha masking, or a host of other features you directly benefit from even if you aren't aware of it. 

Fitted mesh came about because a user experimented with rigging to the collision bones that were always part of the skeleton. Linden Lab's involvement was restricted to adding boob and butt volumes and give their official imprimatur. Revolutionized mesh avatars. Added a lot. Not much work for the Lab.

Much of the work for Project Bento was done by a group of residents recruited by the lab - for instance, the hand bones were done by the maker of Slink body parts and the sliders were configured by the developers of Avastar. Revolutionized mesh hands, heads and non-humanoid avatars. Added a lot. Not much work for the Lab.

Also, bear in mind that one of the reasons the Lab decided it would be feasible to add new bones to the skeleton was that texture and mesh transmission were shunted from the simulators to a new content-delivery network, dramatically reducing the base load of the simulators. That wasn't in place back when you feel they "should have done it in the first place."

Layered clothing is already the low-priced spread of the mesh body world. Even things like lingerie are going rigged mesh. I don't think many people would use baked-on clothing with their mesh avatars even the materials system was included for baking, particularly because materials don't show up in many lighting situations even if advanced lighting is on. However, I DO see a reason to use the existing baking system to send composited diffuse textures and to allow mesh avatar skins to use alpha layers instead of the "lag of a thousand cuts" techniques they use currently.  I just don't see a realistic need to make the significant changes to the baking system to provide materials support, especially because materials are already a separate thing and much of the same function could be done by generic materials appliers.

Of course, as I've been saying, I could be wrong and you're welcome to help develop a new system like users did for fitted mesh, Project Bento and materials themselves.

However...

Nagging isn't helping.

Share this post


Link to post
Share on other sites
27 minutes ago, Theresa Tennyson said:

Of course, as I've been saying, I could be wrong and you're welcome to help develop a new system like users did for fitted mesh, Project Bento and materials themselves.

However...

Nagging isn't helping.

You seem to have a narrow view of what "contributing" means. The same narrow view that led LL to decide they only needed to hire programmers for development.  Sometimes, frequently, to properly develop a piece of software you need people with skills other than coding. When Adobe, Autodesk and other developers of creative software are developing their programs they hire people like me in addition to the programmers who actually do the code. When videogame studios are creating software very similar to what LL provides, they hire people like me in addition to their programmers. That's why their products end up being a lot more polished than LL's.

I've been a designer since before Second Life was a glimmer in Philip's eye. I know how asset optimization works. I understand how content creators will use or dismiss certain features and why. I can build a sim that looks better than anything you've ever seen in SL while also giving you the highest framerates you've ever had in SL outside of an empty sandbox. I can take a tiny mainland parcel and build a fully featured RP area that is larger and more detailed than any full-sim RP area you've ever seen in SL. I know things about the LL appearance editor that I'm fairly confident no Linden realizes and my camera settings are probably the most popular custom SL setting that isn't standard in any viewer. What's more, I share all of this freely.

I can say with the utmost confidence that I have contributed more to SL than you and anyone else who has ever told people "you're not contributing unless you code your own features and do LL's job for them". On top of that, I have backed up everything I've ever said about SL with actual in-world results.

 

 You want to talk about things that don't help, it's someone who isn't contributing anything at all complaining about how people who know more than them just aren't quietly accepting every bad decision LL makes. No one forces you to read my posts. If you disagree with something I say and can back that up with actual points, great, we can have a discussion and maybe we'll both learn something, but going out of your way to derail a conversation with this incessant whinging "why can't everyone just be happy with what they get?" and making snarky, ill-informed comments? That really doesn't help.

 

  • Like 1

Share this post


Link to post
Share on other sites
15 minutes ago, Penny Patton said:

 You want to talk about things that don't help, it's someone who isn't contributing anything at all complaining about how people who know more than them just aren't quietly accepting every bad decision LL makes. No one forces you to read my posts. If you disagree with something I say and can back that up with actual points, great, we can have a discussion and maybe we'll both learn something, but going out of your way to derail a conversation with this incessant whinging "why can't everyone just be happy with what they get?" and making snarky, ill-informed comments? That really doesn't help.

 

I just did in the portion of my post that you didn't reply to.

Share this post


Link to post
Share on other sites
2 hours ago, Theresa Tennyson said:

I just don't see a realistic need to make the significant changes to the baking system to provide materials support, especially because materials are already a separate thing and much of the same function could be done by generic materials appliers.

While the same functionality can be achieved using clothing layers and separate applier HUDs, the effect on usability should also be considered.

Using the current system, a resident purchases an item of clothing, opens their inventory, wears the HUD, selects the item they wish to wear, and with the press of a single button all three required textures are applied to their avatar.

Using the new system with clothing layers and corresponding HUD, a resident purchases an item of clothing, opens their inventory and finds the correct version of the clothing layer for the item they want to wear (we'll just assume that creators will be helpful here and not name the different color variations with whimsically abstract names like "Springtime", "Frappuccino" and "Zowie!"), they then right click the layer and choose which of the meshes currently attached to them they wish it to be applied to, then they wear the HUD for the item and find the button that corresponds to the clothing layer they just applied and press that to apply the other textures required for materials, overwriting any other materials that are necessary for any other layers they may be wearing, since there's no texture baking for normal or specular maps so only one of each can be applied at a time.

The above is assuming that this feature will be used solely for mesh bodies, because without LL somehow adding in the ability to select which faces/parts of a linkset the layer is going to be applied to I don't see this feature getting much use for any other type of mesh items, and since we currently can't specify names for each individual material/face when uploading mesh I can't imagine how they could even begin to incorporate that functionality into clothing layers.

None of this seems like a step forward in terms of usability, and I don't see the majority of residents welcoming this added complexity in the process of dressing themselves considering one of the most common complaints about SL is its steep learning curve.  Also, I'm still unclear on how exactly people are going to specify which order the layers appear on the items, unless in order to wear a layer beneath other layers they're expected to strip down and reapply them all in the correct order.  And what about residents that wear a mesh body but then use separate hands and feet, if they try to wear a layer that should cover both will they have to apply that clothing layer to each attached object separately?

I'm not saying that the entire feature is pointless, but just because something is technically viable and would improve performance you can't just assume that people will adopt the feature regardless of how much more complicated it makes things or how much it impedes their ability to use existing features that they're already accustomed to.  Usability may not seem important in comparison to improvements to performance but the fact is that if end users (i.e. customers) don't like it, how long do you think it will be before creators (i.e. merchants) refuse to use it.  Frankly if they can't add some functionality which allows creators to somehow improve the quality of the content they produce then from an end-users perspective this feature is going to seem like a whole lot of stick with very little carrot attached!

  • Like 1

Share this post


Link to post
Share on other sites
1 hour ago, Fluffy Sharkfin said:

While the same functionality can be achieved using clothing layers and separate applier HUDs, the effect on usability should also be considered.

Using the current system, a resident purchases an item of clothing, opens their inventory, wears the HUD, selects the item they wish to wear, and with the press of a single button all three required textures are applied to their avatar.

 

Here's the thing. They could still do exactly that. I have never advocated getting rid of clothing layers. I just don't appreciate having to make the philosophical decision of whether to use the tattoo layer for body freckles or a certain patch of hair I am wont to wear from time to time because? Can't wear both. Not to mention that either may vanish under sheer clothing anyway. Or, of course, that I can't get a, well, tattoo.

Besides, this is the more likely scenario when a resident gets dressed with a mesh body:

They put on a piece of mesh clothing, then they grumble that part of their body is sticking out so they go to their big honkin' hud which realistically needs to be worn at all times. They then try to find a combination of the dozens of alpha cuts of their body to hide what they need hidden. Then, if this is to be repeatable, it needs to be saved into a scripted something.

(Oh yes, I do know - some clothing is scripted to automate this. Which is why my aunt has a nightie that turns off a piece of her mesh that doesn't need to be turned off and she has a hole in her chest.)

Meanwhile, this system means that their avatar needs to be made up of a collection of multiple meshes with scripts to turn on each mesh face on and off. Dozens of faces. I imagine some bodies have over a hundred faces when you take all the layers into account. That, of course, means the avatar needs to be made of x/8 meshes, with x being the number of faces.

Or... they can put the piece of clothing and an alpha layer made for that clothing and be good to go

If the avatar doesn't need to be sliced into alpha cuts the main portion of the body and each clothing layer can be reduced to two faces each.

As far as what goes where - it works the same way it does now. A script. Just a much simpler one.

Share this post


Link to post
Share on other sites
4 minutes ago, Theresa Tennyson said:

Here's the thing. They could still do exactly that. I have never advocated getting rid of clothing layers. I just don't appreciate having to make the philosophical decision of whether to use the tattoo layer for body freckles or a certain patch of hair I am wont to wear from time to time because? Can't wear both. Not to mention that either may vanish under sheer clothing anyway. Or, of course, that I can't get a, well, tattoo.

Besides, this is the more likely scenario when a resident gets dressed with a mesh body:

They put on a piece of mesh clothing, then they grumble that part of their body is sticking out so they go to their big honkin' hud which realistically needs to be worn at all times. They then try to find a combination of the dozens of alpha cuts of their body to hide what they need hidden. Then, if this is to be repeatable, it needs to be saved into a scripted something.

(Oh yes, I do know - some clothing is scripted to automate this. Which is why my aunt has a nightie that turns off a piece of her mesh that doesn't need to be turned off and she has a hole in her chest.)

Meanwhile, this system means that their avatar needs to be made up of a collection of multiple meshes with scripts to turn on each mesh face on and off. Dozens of faces. I imagine some bodies have over a hundred faces when you take all the layers into account. That, of course, means the avatar needs to be made of x/8 meshes, with x being the number of faces.

Or... they can put the piece of clothing and an alpha layer made for that clothing and be good to go

If the avatar doesn't need to be sliced into alpha cuts the main portion of the body and each clothing layer can be reduced to two faces each.

As far as what goes where - it works the same way it does now. A script. Just a much simpler one.

Okay, if you're talking about the new feature being used purely as a way to apply the equivalent of a body alpha to mesh bodies, and using it to apply simple decals like tattoos and freckles etc. thereby reducing the number of layers of polygons used and the numbers of sub-objects/faces used to accommodate alpha layers in HUDs, then yes I'd agree this feature will probably work quite well (although I still think the issue of having to choose which attached mesh its applied to and therefore having to apply it multiple times if you have separate hands and feet is going to be potentially unpopular).

If that's really all that LL are trying to accomplish by implementing this new feature then I guess it will more or less achieve their goal, however I still think it would be nice if this, at some point in the future, led to them expanding the functionality of the texture baker to include more advanced methods of texturing which would allow creators to optimize and further improve the quality of content in general.

Share this post


Link to post
Share on other sites
14 hours ago, Theresa Tennyson said:

I just did in the portion of my post that you didn't reply to.

You made a snarky comment about how freckles and tattoos wouldn't affect materials anyway and when it was pointed out that you were ignoring most use-cases for this feature you made it abundantly clear that you think the only people who should have any comment on how LL goes about developing features are people who can code the features themselves if they don't like what LL is doing.

The other arguments you made have been things like "well, no one uses layer clothes anymore" which really isn't true and then doubling down on this ridiculous "if you can't code the features yourself then shut up" attitude.

That is not backing up your arguments with actual points and you can make your arguments without that attitude.

I am telling you, as a designer and content creator, and as someone who spends most of their SL time these days customizing their own avatar, that if the baking feature only works for diffuse textures and doesn't support materials that is a problem for the following reasons:

  • If you have a body with materials, those materials will be visible over your baked on clothing.
  • Beyond clothing there are other uses for baked on textures, such as fur for non-human arms and legs, used for avatars such as catgirls and minotaurs and fauns and other fantasy creatures. These are often used to blend the body to replacement non-human bodyparts but will also be rendered useless by the lack of materials support.
  • On top of that, appliers for individual materials will be useless. Let's take the most common example, regular human skin and clothes. Let's say the skin, shirt and pants each have their own materials. The baking system not supporting materials means you can only use the materials for one of these, even if the material is added via appliers like you suggested. Even if only the skin has materials and the layer clothes do not, you won't be able to apply the clothing to your body without having the skin materials appearing over the clothes. So you're right back to an onion skin avatar.

 Do you see now why the only solution you proposed makes no sense?

9 hours ago, Fluffy Sharkfin said:

Okay, if you're talking about the new feature being used purely as a way to apply the equivalent of a body alpha to mesh bodies, and using it to apply simple decals like tattoos and freckles etc. thereby reducing the number of layers of polygons used and the numbers of sub-objects/faces used to accommodate alpha layers in HUDs, then yes I'd agree this feature will probably work quite well (although I still think the issue of having to choose which attached mesh its applied to and therefore having to apply it multiple times if you have separate hands and feet is going to be potentially unpopular).

If that's really all that LL are trying to accomplish by implementing this new feature then I guess it will more or less achieve their goal, however I still think it would be nice if this, at some point in the future, led to them expanding the functionality of the texture baker to include more advanced methods of texturing which would allow creators to optimize and further improve the quality of content in general.

This is certainly better than nothing, it will have its uses, but it remains so limited that it won't have much of an impact on reducing the performance hit created by onion skin avatars and, if I recall correctly from the Office Hours I attended, that was the main goal.

Share this post


Link to post
Share on other sites
18 minutes ago, Penny Patton said:

If you have a body with materials, those materials will be visible over your baked on clothing.

  • Beyond clothing there are other uses for baked on textures, such as fur for non-human arms and legs, used for avatars such as catgirls and minotaurs and fauns and other fantasy creatures. These are often used to blend the body to replacement non-human bodyparts but will also be rendered useless by the lack of materials support.
  • On top of that, appliers for individual materials will be useless. Let's take the most common example, regular human skin and clothes. Let's say the skin, shirt and pants each have their own materials. The baking system not supporting materials means you can only use the materials for one of these, even if the material is added via appliers like you suggested. Even if only the skin has materials and the layer clothes do not, you won't be able to apply the clothing to your body without having the skin materials appearing over the clothes. So you're right back to an onion skin avatar.

 Do you see now why the only solution you proposed makes no sense?

 

Show me where I ever said that clothing layers wouldn't be used. I said at least twice that they still are useful. They'll be a lot less of an issue when they're two faces rather than being dozens of faces. And when layers for clothing are still used all your complaints about materials not being bakeable are invalid. I did mention that baking on clothing would be a possibility, but that would strictly be for low-impact layering purposes.

In Second Life, materials are a completely separate thing. With most animals, the texture of their fur or scales is independent of their coloration. For instance, a horse's hair will generally have the same texture whether or not it has a blaze or pinto spots. Right now the Waterhorse rideable horse needs a separate layer for spots and blazes. With the ability to use baked diffuse textures that would be unnecessary, while the separate hair specular and normal layers don't even need to know that they are there.

Share this post


Link to post
Share on other sites
1 minute ago, Theresa Tennyson said:

Show me where I ever said that clothing layers wouldn't be used

Ok.

16 hours ago, Theresa Tennyson said:

Layered clothing is already the low-priced spread of the mesh body world. Even things like lingerie are going rigged mesh. I don't think many people would use baked-on clothing with their mesh avatars even the materials system was included for baking

 

3 minutes ago, Theresa Tennyson said:

And when layers for clothing are still used all your complaints about materials not being bakeable are invalid. 

No, they're not, because I made it clear from my very first post that the biggest benefit of this feature would be in getting rid of onion skin avatars and their associated performance hit, and that benefit is what would be lost if materials were not supported. That is the point I am making, that is the point you should be discussing if you're arguing with me. If you're ignoring that point then you've either misunderstood me or you are just shadow boxing with your own straw men.

8 minutes ago, Theresa Tennyson said:

In Second Life, materials are a completely separate thing. With most animals, the texture of their fur or scales is independent of their coloration. For instance, a horse's hair will generally have the same texture whether or not it has a blaze or pinto spots. Right now the Waterhorse rideable horse needs a separate layer for spots and blazes. With the ability to use baked diffuse textures that would be unnecessary, while the separate hair specular and normal layers don't even need to know that they are there.

What does this have to do with the price of eggs? I think you misunderstood me again. Have you seriously never seen a humanoid avatar with non-human limbs using a system/onion-skin layer texture to blend the limb to their human body and skin? It's pretty common.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...