Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
They're RiCurves. Way better than fibermesh.
Here's a quick messy sim of MRL long & wavy hair(*); it takes a bit of fiddling and some deformer magic to get it into a simmable position when using on Genesis1 and when you want a lower hairline (not that you can see it right now, but it's important for me LOL).
I turned the skull cap off and did not do anything to the root thickness of the curves, so there's a little scalp showing.
Three minutes and a half to render. For hair this long. With all the volumetric effects you'd ever want. No opacity adjustments.
The only thing that truly needs tweaking is highlight roughness.
* This one actually: https://www.daz3d.com/mrl-dforce-long-layered-hair-for-genesis-8-female-with-colour-mixing - "colour mixing" works, as you can see
Tks for sharing! Wow that's impressive speed. With volumetric effects are you saying you used SS scattering? Translucency? Anywhichway that was fast:)) Hmm... tks for bringing this up, maybe I should get myself one of those and play around a bit
...and as a fact I've used dForce mostly for things it was not made to do, more like dStroy. Or like here for making an asymmetrical morph;) Oh btw this one did not render in 3m 30sec
What I meant is a combo of the actual 3Delight hair BRDF and the fact that this is the deal deal, loads of curves in a 3D distribution over a surface, not an approximation by flat ribbons. Think of fog/rain planes vs full-on volumetric fog or 3D rain droplets instanced in three dimensions over the scene.
...even that scalp showing, it does it in a realistic way.
And since the hair BRDF was written to work with RiCurves, which are, again, a dedicated geometric primitive developed specifically for the purpose of simulating hair and fur efficiently... a marriage made on whichever astral plane you prefer.
The hair BRDF doesn't compute "conventional" SSS but a dedicated version of scatter + translucency + highlight; Wowie used it for his hair shader but he was gearing his one towards transmapped models, so he added loads of stuff to make it work basically outside the limits intended by the devs. Me, I'm a simple fox, I take the building blocks and put them together without much extra tinkering.
Hehe, the "polygon" dForce does lend itself to weird uses. I'm thinking about whether it might ultimately be useful to solve (or at least alleviate) rigging issues with old figures, like pre-V4, think those ugly shoulder bends etc. A smoothing modifier works okay in a lot of cases... but it doesn't seem to be mappable, unlike dForce, so it smoothes everything over, even parts you don't want to get smoothed.
disclaimer: yes V4 and her friends actually do have ugly bends by default as well, but there are fixes for those. not the case for older figures, especially non-DAZ.
Ok I see what you mean;) But I'm actually not sure if this is the new strand based hair or something else? Also have no idea if wowie's shader works with that. And will it work in DS 4.9? I kind of doubt it, so would have to update the 4.10 beta to the latest version then? Or maybe I'm completely off the rails here...and also very paranoid about losing the 4.9 functionality, yes I have backups on two computers and a number of external disks so should be good, but still...every time I open DIM to install something I get the chills
.
Yeah that BRDF is really cool. Wowie's shader works like a charm on transmapped hair, no doubt, and is so easy to use:) I use it on GB hair also because I get much better results than with aweSurface. However you need to set up the hair a bit differently in the GB editor, very light colors and minimal saturation.
Getting the GenX stuff was the best buy EVER:) I used to render M4 a lot but now I only load the Gen4 figures to transfer the morphs. And, in fact, have bought some very nice G3 morphs on sale lately, a bunch of dirt cheap characters, and just the other day I got the G3UV's to Genesis1 script, which works very well. So now I have to testconvert a number of G3 skins, some of which look very promising;)
The hair shader will render with polygon, fibermesh and strand based (RiCurves) hair. As Mustakettu noted, you don't need to enable opacity at all with strand based hair. Translucency is built into the brdf, and my shader comes with overrides. By design, the shader will disable translucency when the hair color/hair color strength is 0 or you manually disable translucency or set translucency strength to 0.
Opacity should only be enabled if you need to use a mask (usually with polygon based hairs). Compared to AWE Surface, the AWE hair shader is faster at rendering less than 1 opacity values, so it should be used for hair. This is mostly due to the way 3delight hair BRDF works.
The extra stuff will only work with polygon based hair and most likely don't with Garibaldi and DAZ Strand based hair. This is because both don't do 'proper' uv export (the v value runs along from the root to the tip of the hair). In other words, the entire hair strand will have a single constant color (ie translucency/roughness/melanin).
Another 'behaviour' that's just plain dumb on both solutions is the baked color values (into Cs), but the latest build adds an override to Cs so you can actually use plausible hair colors using a combo of melanin values. Since Biscuilt/DAZ didn't bother exporting the proper values, that the workaround I've come up with. I haven't worked out a proper solution for this yet in the available build(s). For the time being, simply play around with the override color to have a mix between the embedded values and the melanin values.
If you prefer to use the original hair color/map values (on both polygon and strand based hair), just leave the override at 0.
I've already done lots of work on a new dev build, but haven't had the time to upload it.The most noticeable change is the red melanin values and less hacky stuff under the hood. Another thing to note, the hair will render slightly different between DS 4.7 and DS 4.8 onward. It won't be a perfect match, but close enough in shade
Side note:
So technically, I'm not pushing it beyond what the devs intended. I'm just providing enough pattern generation information which Biscuit/DAZ 'conveniently' baked into a Cs (Diffuse/Base) color, regardless whether the color is actually realistic or in the proper albedo range.
Plus other stuff like switching u/v values with a mask, since older props have ludicrous ideas about how to UV map a hair.
The "new" SBH is Garibaldi 2.0 (maybe even 1.x). DAZ kept the "make it dynamic" technology bit to the local PAs so that the competition would not drive people like me away from this store completely =D Other than that, SBH/"dForce hair", the original Garibaldi, LAMH - they all render as RiCurves. Each one just has its quirks.
I prefer Garibaldi/SBH from the creator standpoint - not in the least because it tesselates the curve automagically, and with LAMH you need to first think of how many segments you need. And also the brushes feel different + clamping totally rulez. ...also because LAMH crashed on me wayyy too often, before we discovered the magic one line to make Gari/SBH curve generation work with scripted rendering and I could go back to what felt better.
An advantage of LAMH is that it is 256% better for furred creatures because it can take the colours from the surface map, think tiger stripes etc.
An advantage of store "dForce hair" models is that they're pre-made (duh) and can be simmed. See, while the Gari/SBH brushes feel more responsive than LAMH's, the collision detection with the underlying mesh is still imperfect. Getting the "brushwork" just right takes time.
The updated AWE Hair build is up.
The link is the same. To make it easier, here's the link:
https://drive.google.com/file/d/1TY3k9Gn27ssEnsNqDoB2EhPViJU0jAFl/view?usp=sharing
Tks a bunch wowie:)
Just opened a recent scene, changed the camera and light angles. Quick first impression of the new AWE Hair...had to dial down translucency quite a bit. Also had to increase hair samples to 256, which obviously was not quite enough for a clean result. Need to play some more...
It's not due to sampling, so raising the samples won't help you. It's something under the hood, which I'm still troubleshooting.
Ah ok, tks for letting us know!
A quick play with my edge blending/ambient shader network while trying to figure out what to do next
Hmm no issues here, think the hair looks ok. Used the black hair preset with "override hair color" at 50%, 128 hairsamples, 50% translucency strength.
Testrender of FB's "A walk across the water"...hmm not sure if it needs some toxic waste or bikini babes. One does not necessarily exclude the other?
So I've looked into issues reported by Sven. Seems like there were some differences between 3Delight 11/DS 4.7 and 3Delight12/DS4.8 and up that I wasn't aware of. The differences means diffuse/specular and even displacement renders a bit differently, which would explain most of what Sven saw (and I didn't). After a bit of fiddling, I think I've managed to make the output similar for both AWE Hair and AWE SUrface. Unfortunately, I wasn't able to replicate the reflection with refraction issue. But maybe the changes I've made to specular/reflection solves that issue.
There's also other big changes under the hood, mainly the noise issue with AWE Hair and output levels with AWE Surface. I;ve implemented a new scheme that's more 'relaxed' and better adhere to albedo levels (much like in the link Mustakettu posted), Pure white surfaces shouldn't exhibit color clipping anymore.
Here are the new builds (AWE Hair and AWE Surface).
https://drive.google.com/file/d/1TY3k9Gn27ssEnsNqDoB2EhPViJU0jAFl/view?usp=sharing
https://drive.google.com/file/d/154hmnGLi6xBC1Qrx3IJVnukVnfZgiCSr/view?usp=sharing
Tks a lot wowie, off to try them out!
Ok, testrender of a scene I put together yesterday. Had to do some adjusting on the hair especially, so things have really changed, and for the better as it seems:) This is only one simple render but I already feel more at home with this version. For example, yesterday I had to set AWE Hair specular 1 to 5% to avoid overly glossy hair, here I used 40% for a similar result, which totally makes more sense to me atleast. More to come...
M4 Azrael shape and textures on Genesis 1, GB brows and fibermesh stubble using AWE Hair shader
Picked up this rather nice car on the sales...
Also, here's what I've done with the AWE AreaPT. No more need for geoshell.
Reflection, refraction and emission all enabled.
Though right now, the refraction is not actually refracting. :D Well, it works now, but I still need to tweak how the emission works with refraction.Looks promising:)
Picked up the Tropical Pond, now it looks like some of the vegetation is missing so will have to go look for it... testrender with diffuse- and reflection depth 1, only specular enabled for all the leaves, 12h. Enabling reflections seemingly almost doubles the rendertimes. I need a new rig
Testing some HDRI lighting from https://www.daz3d.com/aerial-ibl-wide-open-skies-hdri. Ended up adding an arealight plane as a fill light, and raising Irradiance samples to 2048. Had the "use diffuse textures" enabled for the metal parts and diffuse color off. Noticed that the surfaces get oversaturated unless you enable color correction and desaturate the diffusemaps. Here I used a saturation value of -0.5.
A "normal" and a postprocessed version:
@WendyLuvsCatz
Tks a lot for sharing this cool model, had to do a quick testrender right away
. Made some additional displacement bump and roughness maps;)
@wowie
I converted The Library ant did a testrender. Non progressive 8x8 PS, diffuse depth 1 and reflection depth 1. The only really bad specular noise comes from the reflections of the background floor lamps on the wooden floor. The lamp bulb uses the environmental shader with camera and reflection visibility on. The shade uses translucency and translucency shadows. The light comes from a primitive cube just above the bulb using the PT arealight shader. I've tested making the bulb invisible, turning off translucency on the shade, making the whole lamp invisible, turning off glossy fresnel and anisotropy, setting diffuse roughness to zero on the floor. Also tried disabling reflection depth override (in awe environment override settings) but the noise persists. Drawing the conclusion it's the small size of the emitter cube? If you compare the reflections of the windows on the floor they look clean enough. I placed emitters just inside every window. The environment sphere is a jpeg with default settings and that seems to work ok. Also the fire flames are the original planes with PT arealight shader applied. Works ok. Any thoughts or tips?
Untouched render:
Spotrenders, lamp hidden, only emitter:
Diffuse roughness zeroed for floor:
Glossy Fresnel disabled:
Lamp visible, Reflection depth override off:
Simple. Don't use the environment sphere shader (for lighting). Like ever
The ambient part isn't properly sampled. It's the same thing if you use the environment sphere with a HDRI. If you have to use it, be sure to add an emitter with AWE AreaPT in the same place (or enveloping it).
As per usual, not sure I follow. I use the PT arealight, that is the problem. Did you not read my post? The spotrenders are of the reflections on the floor, caused by the AWE AreaPT emitter. I also showed that hiding the lamp/lamp bulb makes no difference. My conclusion was: If I use a smallsized AWE AreaPT emitter it will cause specular noise, but a largersized works better, like the ones I used for every window.