Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.Daz 3D is part of
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
WIP, I don't promote weapons, but she looks rather harmless
, so here goes...Vicky In No Temple With, ok forget it...
So this scene is basically just another skintest, same testlight rig slightly modified, just didn't feel like using another random backdrop;) Testing the coatlayer on the skin. Looks like it adds a dimension, but also hides the glossy fresnel effect on the actual skin layer. Will have to fiddle more with thickness and strength settings. Here is a raw render lage size converted to jpeg for forum upload, just to have a look at the details.
@wowie
So I have a bit of trouble getting rid of what looks like too strong AO on the coatlayer on the skin, or darker areas at grazing angle. Glossy fresnel coat with some roughness is enabled, as well as specular/reflection with roughness 32%. Coat thickness is a very small number like 0.x. Coat strength about 35%. Any hints? Reducing thickness even more? The effect is quite subtle at these levels, but a bit unexpected.
So working a bit more on this scenario. Had to use SS, not translucency, on the grass clumps, realised they had actual thickness. The char on the left uses the coat. Maybe it's not so bad after all, hmm...
Progressive testrender
Always open to NVITWS in or out of temple.
I was going to try out your coat transmission suggestion, but I've a dearth of other skins for Gia 7 - whatever posessed Daz to have a separate UV for all G3 daz figures.
...another one...
What's that rising up front?
Dragon egg, or artists beerbelly (a common side-effect of a lot of time in front of a computer)?
I'd say it was interrupting the girls fight, but it looks more like shopping gossip 'I got this flail, 2nd hand at guillotine joes (where prices are severed)'
Don't use thickness if you're not going to apply a colored coat. The thicker the coat layer, the more you'll see the coat transmission color.
For skin, an infinitely thin coat works best.
Coat thickness is different to thin film thickness (which is its own setting. Although, for the coat, the coat IOR layer will be used for the thin film effect as well. Coat thin film will only be rendered if you have the coat specular/reflection enabled. If you want to have a thin film, a colored coat without enabling actual coat specular/reflection, just enable coat (with thickness and your choice of coat transmission color), then enable base thin film. The base thin film has its own thickness and IOR setting, separate from the main base IOR.
Didn't I tell you? Workin on it
Hahaha! Will be back with some more girls gossip action after the break...
Tks, will test your suggestions!
A quickie
Alessandro AMs Winterland terrain and skybox
I lazily just applied Awe env dome to both and gave both about 60% ambient.
Maybe a little pale and washed out
You did notice the inside of the mantle being pitch black? Use face forward is your best friend
.
Yep, but I put it down to the usual, poor light transmission in 3DL
Hot off the render engine, trying your coat suggestion.
Syd Ravens Vanilla Skye skin in coat, Sv7 Jenny in difuse. Coat strength 34% - no extra spec or bump
Cool! Very glossy though. If you used the coat specular/reflection, try making the coat a bit rougher and dial down strength a tad?
Yeah this is worth exploring some more IMO, testing contimues...
Already did, upped roguh to 8%, strength down to 12%. Also thickness down to .3 and adderd a slight yellow tint to coat transmission colour
Very nice:) Was testing Vanilla Skye textures not long ago, see if I can dig it up...ah forgot...render going
Do you use roughness maps for the face? If not, might want to look into it. Easy to do in PS or GIMP. Just need to make one set for each UVset.
...latest Island version...progressive render
Much better, much less...volcanic (with the black sand)
Tks, still working to make the sand hold for closer shots...
In certain circumstances, they do, but the more emitter polys - the higher the rendertime. You are doing material setup (lookdev) - aka constant re-renders. It would make sense to use lighting that gives you efficiency over "prettiness".
Moreover, it would make sense to use the same lighting rig for lookdev all the time: interior, exterior, character, whatever. One big enough poly, fixed emitter properties, which you could position the same distance from the point of interest of the surface you're working on. The same environment as well.
For interiors it makes sense to hide everything but the wall/floor/ceiling/window frame under consideration, so as not to have the bounce from the complete room affect your perception.
Then you will have consistent materials. Which means: figures you could drop in at will alongside any props, and they will all work together in any lighting situation.
And then you could do the interior designer job and figure out where to place the lights so as to get the most pleasing effect.
If you don't use the same lighting rig for lookdev... see, material setup is kinda like putting makeup on IRL. Imagine you have identical twins. They have the same makeup products, but one of them does their makeup in a locked closet in candlelight. The other goes outside on a sunny day to do their makeup. And then you invite them both into the same living room. And suddenly they don't really look that similar anymore.
This is such an obvious idea, and, I suppose not-so-obvious when one has been taught (by experience or 'logic') that context lighting must be context-specific.
I think this one post changes the way I 'see' my lighting, even though it probably means I toss a good amount of 'experience' to benefit from this approach.
Thanks (i think) for that post. My gears are going to grind on that one.
cheers,
--ms
Replaced with plane, and it's a little quicker, but not a lot quicker - There's no way to get an efficient workflow with scripted render while I need to use the standalone 3delight engine, due to the bugs in the DS 3Delight interface, and Daz bundled 3Delight on Wine is only usable with images sizewise less or around 1000.
Attached are two tests, (left has sphere lights, right, large one poly plane). I quite like how the glass looks in the left, but I sort of like the translucence in the glass of the right.
I did try a scene with just the glass walls (sort of clean-room idea), but Dome light all around was over powering and Awe spots/planes on their own was insuficient.
I don't think I'm going to get better on stained glass without building in a matching opacque lead layer.
Where did you place the planes? Try something like this.
The glass has the green/red squared, the emitter on the outside with a yellow color.
Actually, I used Awe spotlights. Aimed at the window from above - I sort of wanted at least hints of the outside HDRI visible through the glass, even if it's just a vague fuzzy outline, and emssive or light planes would obscure it.
Planes or spheres I also had in scene were up in roof of the ballroom to provide a little ambient internal light in the mostly closed ballroom interior. Initially to check the initial awe surface conversion - looking for surfaces that were too or unnaturally shiny.
I suggest using a simple emitter poly (a 1x1 m plane, no division) instead. When sized to 25% it should close enough to a spot light (sharper highlights and shadows). Place it slightly higher than the windows or where you place the spotlights and adjust the distance if necessary. You will likely have to raise the intensity scale / exposure. However, you'll have to do a re-render to see the adjustments/changes to area lights (both with the standard and scripted renderer).
One likely workflow will be to use the AWE spot/point lights first as proxy with live rendering (IPR) without the door/windows enabled. Once you've got the lighting level you want, subsitute the point/spot lights with poly emitters.
For distant light, Sven's trick should do fine. If you don't want to mess with intensity scale / exposure, you can play with the falloff instead. Zero will give no falloff similar to a directional light (or sun), while 1 will give you a linear falloff which works quite well (or close enough) to environment light.
Edit:
Here's an example setup. I disabled camera visibility for some of the walls and dial down opacity for them without enabling opacity, so we could see inside with the viewport. The glass is a single sided poly that comes with the model, so I used just a very minimal tint in the transmission color. With a bit of transmission roughness, we've got a blurry refraction of the outdoor environment sphere. As you can see in the screenshot, the emitter is scaled at 25% with exposure dial up pretty far to light the room inside. The total exposure (13 EV) actually is the same as iray's default with emission.
IPR frequently causes crashes, as does spot render (on Wine, at least).
Incidentally, Why a 1x1m sized to 25% - why not a 25cm sized to 100%?
Ah, yes too bad.
Same thing actually. I used 1 m because it's close enough to a 3 square foot softbox. So, to get to 3x4 feet, I just need to multiply either the x or z scale to 133%.
I agree, this is utterly important. I've made a number of subsets containing 3 or 4 emitters, a small jpg loaded into the env. sphere, all parented to a group for easy turn/off/on and rotation, One is as close to neutral I can get, the others have some temperature/color settings, also have one with a HDRI that is fairly neutral. If I load 3 or 4 sets into the scene, it's easy to test various light scenarios rather instantly.
I've noticed that a fallof of zero will be hard to control, a small number like 0.01 seems to work better. I use that if the "sun" emitter is close to the setting, causing over exposure on the stuff closest to the emitter. Found it easier to scale the awe environment and dome really large, place the sun just inside, and use the default falloff of 2.
Generally, while I agree that one large emitter lighting up an environment is the most efficient way, it certainly isn't the most exciting light. Using a number of smaller emitters of various shape and/or some simple primitive will produce a much more dynamic looking result.
Btw. I hesitate to use the word lookdev, makes it sound like you actually know what you are doing
A falloff of zero isn't exactly hard to control. You will be just using the intensity/strength and color. Another would be the actual emitter size.
Remember that exposure is an exponential dial. Each tick means you're doubling the intensity. WIthout a falloff, you can simply enable negative intensity scale to bring the exposure down. I've checked and you can apply up to -16 EV for AWE Area PT.
Or just use actual light bounces from diffuse planes. Which is what most photographers do when they're talking about fill light. It also has another quality that most people seem to forget - less specular highlights (along with very soft shadows).
Yup, but I've had cases of all intensity/exposure sliders being at minimum, still blowing out the entire scene. Probably also a symptom that the materials were not properly set up?
Very true!
Speaking of blowouts. I still can't come to terms with the massive difference on highlights when switching from AshikhminShirley to GGX. My opinion is still that there is some imbalance between reflection and specular when using the default BRDF with the latest release build, atleast with the PT area lightshader.