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
Just noticed that the Environment sphere shader ignores opacity masks if the sphere is visible to camera. So, for things like lashes, and I suppose this goes for transmapped aweHair also, you need to render the background separately for now?
A plane with the lashes opacity mask, with and without the sphere:
Edit: I tested the env. sphere with OmUberSurface. Nope, opacity masks won't work, unless I enable Fantom.
Edit 2: Does not matter what shader is being used. Put any kind of geometry behind a transmapped object and opacity won't work. Only transparent backgrounds are accepted.
Hm, this was kinda unexpected...any chance of a hotfix?
Red horisontalplane with opacity mask, green plane behind it, transparent background:
Seems just to be working fine for me. Opacity mask works with the plane as expected.
Did you test it with a non transparent background?
I see. So. something like a what you see here.
Instead of.
Something like that?
Ahh this is getting confusing
. Try loading a primitive plane with an opacity map and opacity enabled. Load the awe environment with the sphere set to visible to camera. Is the mask still working properly?
That's literally what i did. both the plane and the sphere have the same shader.
Check if you have the .sdl from the updated builds.
I will probably change the behavior to the 1st pic, which is what is used for AweAreaPT.
Enjoying this info/thread on the update. Thanks for posting your progress/experiments.
I was curious about the nature of the HDRIs you're using with AWE. We've got 'standard' 3DL and Iray HDRIs widely available as well as loads of external HDRI sources out there, etc. Add parris's (PA) "IBL master" product to the 3DL/Iray mix and the head spins a bit...
Can I ask how you would describe the most 'plug n play' HDRI format you use? Considering the collection of both IRAY and 3DL HDRI products I've got, are there any favorite conversions or setting-hacks you would recommend to get them to work with the AWE toolkit?
(If you've already covered this in one of the (your) AWE threads, please just let me know and I'll happily go find your notes. Some long forum threads I'll slog through, but I really enjoy your explorations and thoughts as you go, and always wish there were more when I get to the ends of them, with wowie and kettu's comments and insights making for fascinating - if often slightly over-my-head - reading).
tnx to all,
--ms
The plot deepens here:)) DS 4.20
Just found out that setting opacity strength to anything less than 100% (1) makes opacity maps work. Looking more and more like a DS bug to me?!
I hope these screenshots illuminate the issue better (pun intended):
opacity strength 100
opacity strength 99.999
And here's a screengrab of my current applications/DS/shaders/wowie content:
HOWEVER, what really puzzles me is this testrender of J Tomalin's the Library, plenty of opacity masks here on windows and glass surfaces, but of course I don't use "multiply opacity with specular" to mask out areas, so...
Anyway, really happy to see so little specular noise here, using about 30 small emissive cubes for lights and candles + environmental reflections from a .png.
I'd be happy to ellaborate a bit, but not totally familiar with various exposure- etc values yet.. will probably share some test results in a while. You know, there are good and bad HDRIs. I've certainly downloaded quite a bunch from places like PolyHaven and the first thing I check is the announced EV range. The ones with extreme values I tend to ignore because it can take some serious tinkering to make them work. And I don't care much for those clear sky/sunny midday HDRIs either, blue skylight and sharp sunlight is pretty easy to set up with arealight emitters, in my opinion.
The "right" workflow for using HDRI light, again in my opinion, would be to set up your characters and stuff using a lookdev setup with PT Arealights, that you know will give you consistant results, then set up the HDRI light and dome appearance according to those skinsettings etc. Which is what I tried in that spycar dude render;)
Regarding actual formats: When I say HDRI I mean .hdr, not .exr, which won't work at all atleast on Mac, as far as I'm concerned. And there seems to be atleast one general rule of thumb: If you set up your surfaces using the AreaPT light, it's likely that , when shifting to HDRI light, you need to reduce atleast diffuse exposure offset by something like 2EVs.
Hmm, might be DS related. I'm using both 4.7 and 4.20.1.43 build. No problem with those builds at least.
For HDRIs, You can use any type if you're willing to tweak though. Basically, you just crank up the exposure and tweak gamma settings. So, a low range JPG which basically goes 0 to 1, will be like 0 to 4 with exposure set to 2 (2^2 * 0 to 1 = 0 to 4). Then tweak gamma so you ramp gets skewed a bit. If the image is high res enough,, there shouldn't be too much noise in reflections.
If you're exposure is too high, you'd may get something like Sven's example with the lamp on the car.
Thanks for your thoughts. It looks like your (and wowie's) experience with HDRIs is much like mine - dependent upon each image itself.
That you recommend working with traditional (per toolkit) lights is an interesting strategy, although it kind of tempers my 'silver bullet' reasoning for using the HDRIs in the first place. Probably a good illusion (strategy) to burst, tho... :)
I've seen some good conversion tools for/between the various HDRIs in my wanderings, but was/am hoping someone will hop in and say they've used (more than) one of Dimension Theory's or Colm Jackson's (or Estevez? or...) sets to good effect without too much tweaking in AWE. heh. My experience is also been one of each HDRI source having very unique characteristics and each requiring custom tuning/hacking, almost on a per-scene basis. No silver bullets yet.
I've actually had some decent context-specific results in both 3DL-Uber and DS-IRAY using narrow-range jpegs/pngs as LDRI lighting, but some scenes simply need that wider-range effect that the DT 'maui' type HDRI provides to pop. Per Sven, adding a proper 'sun' light to certain 'flat' scenes can do the trick nicely. That said, a bit of the puzzle here is to see if there's an HDRI 'profile' (range and other characteristics) that one could match while using AWE and be able to count on it being a good starting point. I assume if we found a good vendor-set, that we could count on other offerings in their product-lines might work with similar/repeatable tweaks.
Certainly an easy enough test if I can get a decent scene set up and looking decent using the traditional AWE lights.
(thanks again for the update and ongoing info, wowie),
best,
--ms
So, to prove my point (heh, just kidding) I decided to set up a character using 3 PT Area light planes, then test it under HDRI light. I set opacity strength on lashes and hair to 99.9999% (which btw shortened rendertimes considerably) and used a dark JPG with reflections/refraction enabled for the environment sphere. Neutral awe light, 1024 i-samples, 256 hairsamples, diffuse/specular overrides 2/1, non progressive raw render:
(G3F Marnie skin on Genesis1, with some G3F detail displacement maps, using both specular lobes with specular maps and the coat reflections with default BRDF, bump, coat bump and displacement, no normal maps, fibermesh brows and armpit hair using aweHair, bla bla:))
I decided NOT to touch anything but the awe light environment settings and see how far it takes me when using pure HDRI light. I chose DenkiGaka's IBL Fantasy skies- over the clouds - mid morning 5000m, which I figured would be a challenge due to high EV range and sharp sunlight. Two versions with slightly different diffuse exposure offset values (raw renders) + the awe environment settings I ended up using (poor girl):
Had to raise Irradiance samples to 2048, otherwise same overrides.
Regarding products, I have bought a couple of HDRI sets like DG's stuff, because they are 16k and apparently high quality images + they are without buildings, people and cars. But I don't feel like advertising this stuff or even recommending anything. Just saying I got what I expected;) There's mostly nothing wrong with those from PolyHaven, especially the ones by Greg Zaal, and they come in many sizes and formats. Just don't use those tonemapped JPGs without replacing the embedded color profile with a generic RGB profile (I use GIMP). Also, if you convert a .hdr to .jpg or what ever low range format, remove the color profile, or it will probably mess up gamma values and what-not real good;)
I'm sure there are better workflows out there, I just dislike the thought of having to set up skins (or whatever materials) differently depending on what light I intend to use...
Here's an indoor one from PolyHaven + env. settings, same procedures, raw render:
Env. settings:
The hair looks quite dull in HDRI light, so I would try and raise specular exposure on that, I'd most certainly tweek the skin specular settings, gamma settings etc. for a final render, however, it's not totally off, is it...and not sure I get the max out of aweHair yet.
Conclusion: Wowie has provided us with fantastic tools for relatively easily making most HDRIs work, in the awe light light-section there is scene exposure, scene specular exposure, not to mention camera based exposure, all of which I did not use here. Just wow:))
Still no changes to hair or skin*, just adjusting the environment sphere surface to try and make the hair work better...interesting;)
Minor gamma adjustments in post, Irradiance samples 4096, hair samples 2048
*) Actually, I reduced skin gamma settings from 1 to 0.8, but I think it went over the top, a bit too much contrast now, I think...
The hair now has some specular noise, so that also went a bit to far, obviously. Going to zero out the awe light and start over from the env. sphere surface this time. I'm beginning to understand it's better to tweak the actual light source ie the dome first, and do the final adjustments in the awe light...maybe...
...so back to the original skin (no gamma correction), this time I zeroed the awe light and instead adjusted the environment sphere surface, got rid of most of the noise on the hair (without any adjustments to the hair) and went back to 1024 Irradiance- and 256 hair samples. 10x10 PS non progressive render. (The HDRI is Music_Hall_2_16k (IIRC) from PolyHaven, it's a few of years old, probably.)
@wowie
The opacity issue is worrying, to be frank. Not sure it works properly even with the "less than 1" workaround. I'm considering reverting back to the previous version.
And some feedback:
You and I have been fighting over the specular highlight behavior more or less from day 1, almost 5 years:) You've done your best to educate me in every possible way, and I'm more than grateful. But that won't change my honest opinion, the highlights are way too overexposed from the start. They are the root cause of all issues I'm beginning to believe. You know, I set up the arealights by default with a specular contribution of 50%. For surfaces I generally reduce specular strength to 50% or less, with a mid gray specular color as a starting point for anything. Furthermore I often find myself reducing specular exposure for things like glass and, especially, metals. And I STILL on occation end up with those ugly jagged highlights. And of course, using only reflection without the highlights is not a proper workflow and looks equally bad. This just can't be right?
Instead of trying to explain how to handle those things, because I obviously still don't get it, please consider adding a way of reducing overall specular exposure, maybe in EVs? I'm thinking: Reduce it by 3 or 4 EVs. done!
And, since the two specular lobes are additive, for example using reflections on lobe 2 and specular on lobe 1 can give unpredictable results, like 1% is too little but 2% is way over the top:)) So I'm asking if a blend mode selector or similar, to mix the outputs, would be feasible?
This is not meant as a rant, but the fact is I spend a considerable amount of time finding new solutions and workarounds for this issue daily, and I've given up on myself to ever grasp your workflow, so hands up, please consider adressing this thing!
Edit: Example, the sound tube/horn here is now pretty smooth with these specular settings attached. So spec1 for specular and spec 2 for reflections, if I just add a tiny bit more spec 1 I get jagged highlights both from the fire and the window light planes. The fire is set up with two planes using the fire diffuse map as color. EV 5 intensity with spec contribution 0.3. They are not visible to camera, I duplicated them and used the environment shader for visibility. Window emitters (well inside the actual windows) with EV5 intensity and spec contribution 0.2. No environment light or reflections, only camera visibility on the sphere.
OK. Can do. Just checked and the behaviour of DS 4.7 and 4.8 onwards is a bit different regarding area light. It'll be exactly half the strength it is now.
If you want to tune specular from both area light and point/spot/distant light without adjusting specular/reflection strength on each lobe, you can just use the Specular Exposure dial (in EV/intensity scale) inside the Options/Lighting section of the shader. These don't affect reflection.
I don't know if adding a mix/blend between lobes for specular is a good idea though. I like either just adding both or do an automatic blend depending on the amount of the main lobe used.
Plus, there's also bug in the code that cause specular from lights to be double what it should be.
Will definitely be a better starting point for setting up any material, I figure...great, and thanks for listening to an old hippie;)
Yea, automatic will prob be fine with the new exposure levels..looking forward!
A side project of enabling gobo/gel texture on lights., mostly cause it's easier to write code for the spot/point light shaders than the area shader one. Hopefully it will work just as well on area lights.
Decided to consolidate my point/spot/distant light into one light. The distant light is a physical sun, but no physical sky yet. I also got an idea I need to try out to avoid the pesky unwanted highlights from using ambient light with specular enabled.
Very much looking forward:)
With motion blur turned on, for some reason the Minotaur's surface becomes smoother.
In this example I set the amount of motion blur to zero to better understand the problem.
What can I do to avoid this issue?
I can spot some differences on his right forarm and hand, it looks smoother on the right, yeah? If that's what you mean?
Generally, try an amount of 10 - 20%, depending on the velocity of the motion, and go from there. A small sample amount like 2 or 3 will result in a less detailed "video" style blur, while an amount of 6 - 8 (or more) will produce a cleaner, less noisy effect. So you might want to raise the samples...
Thank you for your reply.
Excuse me. I did not write well.
When motion blur is turned on, the Minotaur 6 Body HD Details are lost.
My hope is that the Minotaur 6 Body HD Details are still valid when motion blur is turned on.
HD morph effects are lost just by turning on the motion blur check.
Image 1 : Motion blur check OFF
Image 2 : Motion blur check ON
※I have marked the obvious points.
※The amount of motion blur is set to zero to show that it is not blurred by motion blur effects.
Why is the effect of HD Morph lost just by turning on the motion blur check?
Ok I get it now, tks for clarifying:) Interesting observation, I wasn't aware of that. I'll do some testing today, to atleast try to replicate the issue.
Thank you for your reply.
I would like to test with other scenes too.
So I can confirm that I get the same result, motion blur ignores HD morphs. The only thing I can come up with is to do double renders and combine in post.
Thanks for bringing it to my attention, I guess I've always relied on micro-displacement rather than HD morphs;)
Tested without AWE Shading Kit.
When the Motion Blur check is turned on, the HD Morph effect is lost.
This is a DAZ Studio issue.
I have submitted a ticket to support.
Thank you for your test !
Support staff confirmed that the issue had been reproduced and responded that they had filed a bug report with the development team.
Tks for the reminder! I just rendered this using some ageing HD morphs and motion blur, forgetting about that bug
. Will have to find workarounds (spotrender tool) for the next couple of shots...