Introducing Opacity Control for BOM Layers
tracked
Jaden Nova
I suggest an enhancement to the Bakes on Mesh (BOM) system by introducing an opacity option for BOM layers. As a designer, the current limitations of BOM force us to add additional mesh layers, particularly for makeup applications. This increases mesh complexity and necessitates the use of complex HUDs.
Enabling opacity control on BOM layers could eliminate the need for extra mesh layers, thus optimizing mesh complexity and enhancing overall performance. Additionally, this would allow for greater flexibility and creativity, as BOM makeup layers wouldn’t be limited to one or two mesh layers (depending on the number of mesh layers created by a designer).
With the recent introduction of 2k BOM textures, the capability to adjust blend modes directly on BOM layers would be highly beneficial for both designers and users. It would offer more customization options and improve the user experience.
It would positively impact the creative community and contribute to a more streamlined and efficient design process.
Log In
Gwyneth Llewelyn
I hope that this means the end of onion layers!
It's not because they're too complex (they are, but body mesh creators and layer-based content creators are surprisingly good at designing their HUDs to make this work): it's mostly because it's an utter waste of resources.
An "onion layer" is simply another full-body mesh on your avatar. And you get
three
of them (at least, by some popular brands). But not only that, each of those layers requires scripting (so that HUDs can 'send' textures to the onion layers), which must
be always on — plus, of course, more work done by the GPUs on calculating how light goes through several very closely set meshes and what gets reflected back depending on a multitude of parameters...It's true that
most
onion layer scripts are well-behaved and remain dormant unless 'woken up' by a HUD, but they nevertheless consume memory and other resources, putting a drain on the simulator servers. It's also true that, according to my experiments, a fully transparent item (meshed or not) attached to your avatar does not
count towards ARC (or whatever the metric for avatar complexity is called these days). The point being that if an object is 'marked' as being completely invisible, it's also going to be excluded from the rendering pipeline. But, of course, if you add
the extra layers, it's because you want to use
them; this means that even just a tiny tattoo with a few pixels will require the rendering engine to do a lot
of work. Now add user-defined transparencies selectable via HUD, and things only get worse and worse; it's actually amazing how SL can still do a pretty decent job handling all of that!Gwyneth Llewelyn
By contrast, 'going BOM' and getting rid of the onion layers would mean that you'd only need to get
one
texture (well... theoretically, up to 8 per mesh... but OK), since everything would be baked on the simulator anyway — just like it happens today if you wear legacy texture-only clothes over your avatar's mesh body. In this 'BOM only' scenario, handling transparency on the renderer is not an issue at all: it's only defined once, then baked together, and the result, more often than not, will be fully opaque (except, well, for those wanting to have a see-through skin... but the majority of avatars have an opaque skin, which is at the bottom of the queue) and much, much gentler on the rendering engine: instead of fetching several different textures, each assigned to a different onion layer, and having to calculate how light goes through and gets reflected from all of that... well, you just have one opaque texture with all the layers.It's not just the HUD creators, the content creators, and the end-users who will have a much easier way to get all the layers they want in the right place. It's also the renderer that gets a much-needed relief.
I understand that PBR materials
might
be a trifle harder to implement, but not overly so. From the definition, a 'texture map' is done pixel by pixel, each carrying materials information; the BOM engine could, theoretically, do all of that together — and consider the opacity level and the order the baking should be done — so, at the end of the day, every time you changed the opacity of each layer, the materials would need to be recalculated again, pixel by pixel... but you only have to do it once, while the user is tweaking their avatar, and then the results are stored in LL's BOM cache and retrieved only once.Gwyneth Llewelyn
In other words: overlapping 'BOM materials' are not much different than overlapping plain textures — it's just the maths that are much more complex. But it can be done! In fact, AFAIK, that's how it works under most rendering engines that support PBR materials. The result of a 'baked' multi-layer texture + materials, each set at a different level of opacity (and, why not, a different tint as well!), is just
another
new texture (which will even be opaque most of the time). The rendering engine, theoretically, does not care if the materials map is meant for just 'one' layer or not — it needs to validate each pixel's materials individually anyway, no matter what the 'messy' blended texture + materials looks like.Then, of course, we'd need a way to
script
all of the above... messing up with it completely... hehe, just joking! I can imagine that content creator would supply their texturing done at what they consider the 'optimal' opacity layer, and then it would be up to the user to adjust that, from the Appearances interface, for instance, or something completely new just for that purpose.Seriously, I believe that this feature request, even if it does require a bit of programming, will only build on top of the existing mechanism for BOM-on-demand, which, admittedly,
was
the toughest part to do. The rest is just investing in further development, which will save everybody's time — including the simulator's!Serena Snowfield
Yes, please! 🙏🏻
m
mikomor Resident
I would LOVE to have this implemented!!!
Beatrice Voxel
This would be a gamechanger! Instead of having to have multiple versions of a single tattoo, one could simply 'fade' the fresh version to age it. Solid to sheer clothing? yes, that too.
Jolie Serendipity
This would make customization so much easier and allow for more creative uses of the BOM layers. I cannot wait to see this implemented!
Ember Ember
Opacity control would also save clothing and makeup artists from needing to upload numerous extra textures in order to offer their customers a useful variety of preset opacity percentage options.
Eden Nova
This is one the of most upvoted requests on here. When is LL at least start working on this?
Amie Blossom
Yes please!
BOM opacity control could finally rid us of makeup appliers. Managing appliers is quite tricky in that it's very easy to lose track of what you currently have applied (but this could be improved with scripting, body/head HUDs should be able to at least display the name of the current appliers in use... but don't) whereas BOM items are obviously displayed in inventory/outfit as worn.
My use of BOM makeup is quite limited however due to the inability to fine tune opacity which is often necessary.
Keeley Blakewell
This opacity control of bom for makeup would be amazing !
Lucky Clover
Cannot imagine how many more tattoo layers I'd be buying if I was able to turn the opacity down a little. Or how many I'd make if I wouldn't have to go through the hassle of making 100%/75%/50%/25% if it's something that'd need it. I think about the 'tracked' text on this feedback request every day now. Please........ bom-based Materials are important too, everyone wants them, but opacity control would change so so so sooo much.
Lucky Clover
It's been a while and I am still vibrating about this. longing for this. dreaming of this. linking this feedback to people wherever possible.
but I do want to add on that upgrading 'tint' for bom layers in general would be nice, blend modes like multiply/screen/overlay/hue/etc would be an extremely welcome update. Stuff like skin tinters would work so much better with a low-opacity overlay or hue layer, stuff like freckles would blend better, markings over fur/scales could keep the underlying texture's shading without a separately tinted 'just the shading' layer, so many things.
Opacity is absolutely the top priority, but since the original post did mention blend modes without elaboration, pleaase consider some extras in the future. Particles kind of have some, even if the names for them are odd and they're mostly regarding the alphas, so seeing any type of new blend modes anywhere would be so nice.
Gwyneth Llewelyn
Lucky Clover — yes, yes, YES!
I was thinking exactly the same... I'm glad I'm not the only one.
Load More
→