Adding to Cart…

Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.You currently have no notifications.
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
I've experimented with that. You KIND of can, almost, if you do a dynamic outfit that's vaguely similar and then have the conforming outfit collide with the dynamic outfit.
Mind you... I'm not sure it's worth doing (the small details you get with conforming outfits tend to get warped a lot with collision), but hey.
Fit Control is pretty cool, though.
There is another thing to consider. With Iray, you're basically stuck with what the renderer offers you. For example the SSS changes between the 4.8 and 4.9. With 3delight, technically you can have access to all three methods of doing SSS - point cloud, precompute and raytraced - using the same inputs, regardless of renderer version and they will just work. Technically, since even 3delight devs have foregone the SSS precompute method in favor of the raytraced SSS (only point cloud and raytraced are available as options). Materials don't need to be revised to work with a specific version of the renderer.
In essence, a shader approach like 3delight means it's infinitely extensible and robust, while a material definition language is primarily tied to how the renderer interprets material settings. The downside is, 3delight output is tied to the shader your using and in that respect, 3delight shaders for DAZ Studio is stuck in the 90s with its Lambert/Blinn/Phong approach. Thus the need for old school tricks and hacks.
There's HSS AND there is UberSurface one...two different things. HSS doesn't have the Global Occlusion over-ride and a couple of other items.
Yeah, I know the difference. HSS was basically the first omnifreaker shader I've worked with. Ubersurface was better (fresnel for the reflection and roughness for the diffuse). Plus the occlusion override of course. If I recall correctly, HSS was first bundled with the Elite textures. If I'm not mistaken UberSurface 1 is bundled with DS - in the DS default lights and shaders package. If you already have UberSurface, there's really no need to use HSS.
Right...
Except that for some reason, probably because it has less complex code, HSS renders faster for simple surfaces..
Hrm. So I'm trying to get a good outline for cartoon-style stuff.
PWToon, and DZToon and that free Geoshell outliner shader, run into problems whenever surfaces are incident to the angle of view, and you get big black segments. This can also make the thickness of lines rather uneven.
I've been experimenting with the Outline from Scripted 3Delight, but I run into problems where beards, bumpy surfaces, and faces become big blobs.
Any suggestions/tips?
lol. A funny thought. Iray has no velvet of any kind. Yet it is tooted as the best render engine for things like cars. How do you do floor mats, carpeted seats, and fuzzy dice for the cars without velvet. hmmm.
I was very shocked once I realized exactly how Iray displacement worked and why people have such problems with it.
That's because Game engines prefer the lesser realistic 'Normal' maps over actual Displacement. It was probably assumed that rough geometry could be done with low-geometry based displacement, and the rest filled in with "Glorified Bump maps". Unfortunately it doesn't work all that well for things with sharp edge features (they turn into mushrooms).
So, even if Iray had velvet, the fuzzy dice would still be a big problem, and anything else with normal maps viewed edge on. lol.
(EDIT)
So yea, those two things are the Achilles heal of Iray. Displacement map geometry density, and Velvet. Not that 3delight dose not have it's own difficulties in Studio.
And most of those are much easier to overcome...because 3DL actually uses shaders, which basically means if it doesn't already do it, you can extend the program by adding new code (the shader). Iray on the other hand has ALL the shading functions hard coded and it is just material definitions you get to play around with. Yeah you can do things like add noise and some math functions to change up those definitions some, but you cant' go and add 'real' displacement.
This is a short list, off the top of my head. I'm sure I could fill out a complete list of twenty four problems, tho this is a starter list. The first three is mostly because they appear to add a disproportionately longer time to renders then if it was just a diffuse surface being rendered.
Lists of unsolved problems in Render engines (or shaders).
The not exactly render engine Hilbert's problems. lol.
1. Ray trace reflection shader performance.
2. Semi opaque (opacity maps. Glass, lace, etc) surface render performance
3. Sub surface scattering and volumetric shader performances.
4. Consistent reflection angles independent of surface vertices density (not angle, the density of the mesh).
5. A Normal or Bump map method that dose not diminish at approaching glancing angles of view.
6. Velvet that is based on light intensity at the surface, without glowing in the dark.
7. A texture shader method that dose not produce Moire Interference Patterns with micro feature maps.
(EDIT)
My apologies Kettu, Wowie, and mjc1016. The list idea just came to me, as I was talking about that velvet thing in Iray. I know your working on shaders, and I have no idea what if any of this influences or is solved by your Work In Progress. It is merely a list of things the stuff included in Studio has "difficulty" with.
(EDIT 29Jan2016) To keep them somewhat together here are the others. They were brought up after the following posts, and discussed further.
8. An intensity or scale control for Normal maps in Studio for 3DL.
9. A texture shader method that dose not lose directionality to the surface details at distant render scales. (wording of is still in debate).
24. Better organization, presentation , and cleanliness of the shader interface.
TBC
1, 2 & 3...use progressive for 3Delight, as it enables the much faster raytrace hider. If they would enable raytrace caching, by default, it would be even faster (like up to 6x), than just progressive.
4. It's the shader, not the renderer.
5. Impossible. Bump and normals are cheats and just don't work that way. They can't work that way. That's what displacement is for.
6. Again...shader, not renderer. Using 'old school' shaders limits those kinds of things greatly. "Old school" pays no attention to energy conservation and that is essential to having velvet without looking like it was made in Chernobyl.
7. If it can't be don't with a camera...it takes a very good camera/lens shooting in basically HD (true HD) to avoid that in movies/TV. Very high resolution 32 bit textures (true 32 bit) and very high resolution images are the answer. And can't forget to include scale, too...that plays a role (why Studio does things in cm is beyond me). Because for certain things scale does matter.
1, Reflections, I guess that it is about four times longer for the Omni and Daz default shader with reflection on (scripted AoA, I don't think so, I'm not that patent, lol). Reflection over all, is not as bad as the other two (2 and 3), tho it dose make it painful to set up reflective surfaces.
2, I've had some lace items take days to run to completion (for three squares), when the rest of the render was complete within hours. Admittedly it was a multi layer lace spot (bow dress bow), it was drastic for the render. And like reflection, I've noticed some simple glass surfaces take considerably longer (over an hour) to render then when the surface was 100% opaque (fourteen minutes), as recently as that Bracelet I'm trying to make.
Especially 3 with the sub surface 'Pre compute delay'. I have one figure that it takes over 45 minuts to do the pre compute delay process, and then it only takes about three minutes to actually render the scene. That is unreasonable, and hints at unnecessary calculations wasting CPU cycles in the Sub surface 'Pre compute delay' process. That's fifteen times longer to pre compute, then to render the results. There is no reason for it, and it can be a lot better.
4. shader or not, it is a big issue in Studio.
5. until Iray dose this, some things will simply be impossible to do. 3delight can do displacement independent of geometry density, however it would be nice to have with stuff that only came with Iray Normal maps.
6. out of the four shaders included with studio. both the Iray one and the Daz Default shader have no Velvet at all, and the other two (Omni and AoA) both suffer glow in the dark side effects (especially on some cloths). It is a problem, or as I prefer, there are no problems only opportunities.
7. because the one that makes the shader has no control over the 'scale' of the users render, backing out will produce this, and zooming in on a lower detail map will look ugly. The human eye and 35mm film dose not produce Moire Interference Patterns when looking at cloths (or other surfaces with micro features that produce the texture). Simply blurring the image is not a reasonable solution either, there are other better ways that shaders can do this.
And No 5, brings up Number 8. An intensity or scale control for Normal maps in Studio for 3DL. It is lacking in the shaders I bothered looking at the 'Normal map' settings on.
Bump and normal maps are not capable, at all, and won't ever be able to simulate anything than 'straight on' views.
As to 35mm film...try it with things at 1/10th scale, with patterns designed for full scale. And I can see MIP on very contrasty, very fine details (like black and white very fine herringbone). The only way to totally eliminate it is to implement 'real' cloth...sort of like curve/particle hair where the actual weave is defined mathmatically and constructed procedurally at render time.
To eliminate the 'glow' on the velvet in the AoA or omni shaders, the velvet strength and the diffuse strength MUST when added together be no higher than 100%. Ambient MUST be off. And translucence/SSS must also fit in to that 100%...so if you want it all, you must balance them (translucence and SSS should never be used together...that's another source of the Three Mile Island effect). And finally...Fresnel needs to be enabled.
If the normal maps are properly made, by definition, they are 'scale less'...the problem is, most normal maps are not properly made. Compositing them from three different lights and polariztions or baking them from actual geometry are the only ways to properly make them...running a diffuse map through a filter in Photoshop IS NOT making a normal map.
This was thrown together in about 5 mins and renders in 43 seconds on 'draft' and and 53 sec on 'medium' render settings.
MJC1016, not arguing, just 'debating'. as for that cloth mad in joke, lol.
Aiyana - "Is it supposed to be glowing like that!? Mom! I wanted a Play Station, not leftovers from Chernobyl."
Mother - "It's not from Chernobyl, it came from Japan".
Adding to mjc1016, the way velvet is done in DS default shader is basically wrong (diffuse). Since it is diffuse, it reacts to any ambient lighting so even when there's no light shining at the object from a particular direction, it 'glows'. If you just want a velvet look on a surface, a really rough specular with fresnel is the way to do it (with very sharp falloff, but the falloff itself needs to be very low). Here's an example:
https://www.thearender.com/site/index.php/news/edition-1-5.html
If you're aiming for actual fine hair, it should be done via a mix of translucency and fresnel specular with proper geometry, be it fibermesh or RiCurves (not displacement the way some people are doing).
As for normal map intensity - there is no such thing. A normal map by its definition provides normal for each point in the surface. Layman terms - it encodes angles, not height info like a bump or displacement. Since you don't have any height info, the silhouette appears flat. We've covered this already, you need displacement or relief map for that. 3delight excels in displacement, so that's the way to go.
If you look at the Digital Emily from the WikiHuman project - http://gl.ict.usc.edu/Research/DigitalEmily2/ - , you'll see they went with both displacement and micro displacement for skin. No bumps or normal maps.
Those are very impressive examples mjc1016 and wowie. I have one example rendered so far using the Daz Default shader, and after looking threw the dozen or so pages of dials in the Omni shader, I know what the 24th problem is.
In a way, resembling the 24th Hilbert problem, the organization and presentation of the shader interface and how clean it is. The Omni surface thing is full of settings for various things just tossed all over the place, and screams rushed out the door with no care how it looks. There are some things that turn on or off stuff, that are no where close to what they control. agh (face palm), lol. It is the end of my day and I need sleep, I really don't want to fuss with that right now.
In any case, this is that ZdgCloth shader included with that "OptoDistort Chair" (OMG, yes I included a cloth shader with that for use on other stuff as you like). This one is the Daz default as set up in that shader (the test vase is included in that Stone shader thing, also at shareCG).
Yep, there is not much at all to that, as most of it is in the bump map vs the gloss setting. And below, the omni shader applied to the above shader, after I found out why the bump map wasn't doing any thing and fixed that.
It needs a lot more work, and I'm just to tired to be dealing with this interface mess right now.
I believe the way the shader is presented has a lot to do with DS itself. I really hate how shaders options are presented in DS 4. DS3 shader presentation in the surfaces tab was better. Shader sections are always ordered the way it's supposed to. Each section has a visibility toggle and can be nested.
You can just expand the sections you're working with and make use of the visible space much better. In essence, pretty much like Maya and 3DSMax.
The way DS 4 UI presents information is wasteful. In the same vein, having something like this:
A graphical representation of what the settings mean would be extremely useful.
Yes...and another thing in DS4.x it has been known to throw out any groupings and randomly throw the parameters up in no particularly logical order. some of the way 4.x does things depends on setting like whether the Show subitems box is ticked or not and what options are checked off in the Surface options;.
Reworked conductors (and rubber).
Let's see - chrome, aluminium, steel, two generic alloys, dark and light generic metal, painted metals (I only used the most reflective one here) and the typical precious metal (gold, silver and copper). I'm using the metal colors from Solid Angle - https://support.solidangle.com/display/AFMUG/Specular - which I think looks closer than Sébastien Lagarde's colors for UE.
Renderer options - ray trace depth at 12, shadow samples at 32. Gamma correction enabled at gamma 2.2 (naturally).
First cup of coffee, and random thoughts. Nice render Wowie, tho I will note the tire rim and the trash can do look a little off (it may just be a lack of texture where it is expected to be from machining). The bronze/copper lamp/lantern on the left looks like some one used red juice (kool aid) to polish it.
Gold is always a tad tricky, as there are a few different flavors of gold (karats, and other alloys).
So in respect of the example chart-thing from Wikipedia, That looks good.
That Normal, thing. As a user of stuff I can't fix and share the fix with others, Pinning the adjustment on the maker and forcing each user to fix it solo is a tad unfair to the rest of us mere mortals, lol. Iray has an adjustment there, I think LuxRender has a 'gain parameter' for them, etc. Why not for Studio 3DL.
Velvet in the Omni shader. I don't think I fully grasped what you were saying when I started to try to set up that rough specular with fresnel (ferz-mla. fumbles over the words) in the surface tab. I think you were talking about making a new DLL for that rather then some setting in the Studio included Omni shader? Not to say, cloth velvet is dependent on the directions of the thread in the weave, and the omni one is purely omni directional, so that carbon fiber example is simply impossible for a few reasons in that 'list'. lol.
Not seeing as too red here...
Yea, I may be thinking of all the brass that was 'never-dull' polished all the time on the ship. "cip" of what, lol. I need to proof read that again (face palm)
like gold there are a few variations of the copper pallet. Tho I think most think of pipes or the fuse panel in there home, most of the time.
Pics from Wikipedia, for reference only.
http://en.wikipedia.org/wiki/Copper
I'm baking up a little bit of an experimental weave.
Kind of the same concept as the Carbon Fiber Stuff I keep dredging up, yet this is more of a traditional cloth of sorts.
(EDIT) It is done, the maps are over in this thread here (there CC0 by the way, have fun).
http://www.daz3d.com/forums/discussion/comment/999789/#Comment_999789
On the Daz Default shader, just dropped in on the former test vase shader.
I'm still in some disbelief over how well this came out. In the past there was some tests, and maps like this had all kinds of issues. I may have just hit that 'sweet spot' with this first test spot render. The Moire Interference Patterns may become quite prevalent at different scales. Time for some more testing.
I was backing out scale wise, when I need to zoom in, lol. Well, now I'm starting to get that nice directionality to the gloss.
And the Moire Interference Patterns. So at least the Daz Default shader has Difficulty with number 7.
7. A texture shader method that dose not produce Moire Interference Patterns with micro feature maps.
Dose this hint at a possible 9, with the loss of directionality to the surface details at distant scales, Possibly.
Because at a tiling of 10 on the test vase, the surface almost goes completely flat looking. Just looking at the carbon fiber cloth examples I've posted in this thread, this is not behaving the same. The red and blue triangles should still have directionality to the gloss at this scale.
You're probably thinking anisotropy highlights. I didn't use anisotropy, though it is available in the shader. Mostly cause I hate how it's done (both in the shader and in the shader mixer brick).
What is exactly the problem you're describing? If you are referring to the lack of details in the silhoutte of the object (again) with bump and normal maps, that's always done with displacement in any renderer. Displacement is the way to go.
I think you're confusing general fresnel and my idea of custom fresnel curve for metals. Fresnel with UberSurface only works with reflection, not specular. Only UberSurface2 have Fresnel that works with specular and reflection. Simply put - don't use velvet. In omnifreaker shaders or any of the DAZ Studio shaders. They are using diffuse, so they will never be angle dependent to incoming light and viewing angle, which is what you're seeing in real world velvet.
Here's an example:
For the top ball, I'm using velvet, while for the bottom ball, I enabled the 2nd specular, with 2% roughness (very rough) with a somewhat sharp fresnel (falloff 2 with sharpness of 70%). This is done only with a distant light.
When I added an ambient occlusion light (UE2), velvet 'breaks' since it is now receiving diffuse light from the distant light and UE2.
I don't suppose there is something like an 'Engineering handbook graph' of sorts for such 'Intelligent' vertices placement. I've noticed that there is a zone of distance around a vertices where if there is another one, they don't move as drastically. However that produces other issues with reflection angles in some shaders (At least the daz default, I have not tested other shaders yet).
I have been told that eventually with enough experience in Hexagon, I will "Just know" where to place extra vertices to keep things from moving around. It would be nice if I could just hand some one a USB stick and get a copy to install in my mind of that "Just know" stuff, lol.
No, sorry, that was in reference to a lack of intensity control for Normal maps in Studio for 3DL, and a response to the (It's a map was made wrong issue). I'm probably reading it wrong in that it is implied that it is now my responsibility to fix some one else's maps, because there to week or to strong in 3DL, and there is no intensity control for them, lol. Side chat about Proposed Problem No 8, "An intensity or scale control for Normal maps in Studio for 3DL."
I intended to respond earlier, YES! However I don't think the Daz default in 4.8 or 4.7 was bad at all. Nothing like trying to find the Opacity stuff spewed all over the Omni shader control list, lol.
Pretty graphs, I'm all for that. Tho that also opens the door for gripes about the desire to just click the curve in the graph and move it around with the mouse courser. That's a chose your battles decision I guess, how much grief do you want to get into as the programmer of that interface, lol.
Two questions about this "UberSurface2" thing, your making one hell of a sails pitch for that even tho some of that did go over my head (I don't know what half the words are and need to look them up, lol).
1, is it a compiled shader (with actual 3DL DLLs, not spaghetti land and scripts) like the UberSurface zero-dot-dirt included in Studio, or is it like the AoA thing that just takes bloody ages to do anything?
2. If I make UberSurface2 wood/stone/cloth presets for the 'Trinkets' I make and give away, what kind of major problems is that going to be? Aside form the obvious disgruntled surfer that dose not have the US2 and clicks the 'Report as inappropriate' button for spite.