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 haven't read any "real" mag in years, but I clearly remember page-size ads for DS4 from a few years ago. I need to check my Zinio library to figure out the exact dates.
As for tonemapping software, well if you aren't rendering to a high dynamic range format, then you don't need it, of course =) But I have become fond of it since I got my copy of Vue 9 Esprit years ago... and then LuxRender (although it has a built-on tonemapper), and now I prefer to render "serious" images to EXR in 3Delight, too, and then tonemap artistically. There is a script for EXR export in my kit ;)
I like Picturenaut, but I'd like to have something like the "film" preset curves in Lux's tonemapper.
You can try with Gimp. You must get the G'MIC extension that has film simulation effects
No EXR support in Gimp though so you'll have to use TIF
You could also just use 3delight Standalone and apply a specific LUT
That looks cool, I'm downloading the 64bit versions of both.
CHEERS!
EDIT:
Got them and they work, what are your recommended G'MIC settings for renders?
CHEERS!
I don't really have any recommendation. That's a matter of taste in my opinion. That is where your artistic sensitivity come into play.
If you're a bit adventurous you could try a 2.9 development build from darkrefraction. It can be unstable but it allows to change the image precision up to 64 bit (not sure that is useful to us). The installation of GMIC was a bit tricky as for the 2.9 dev build the plugins must be installed in C:\Users\[YOURUSERNAME]\AppData\Roaming\GIMP\2.9\plug-ins
Mmmm, I think I'll stick with the build I've got, it's nice that it's 64bit. I got Photoshop CS2 when Adobe 'gave it away', but the DS Bridge starts at CS3. I'll keep an eye out for the next stable build.
CHEERS!
This is up to Aghiles and his dev team at DNA Research. If it's feasible at all. Right now, progressive means box 1x1.
Thank you! Downloading now... can the TIFF be 32-bit?
Found a couple of free LUTs here: http://www.groundcontrolcolor.com/free-luts.html
Technically I'm not breaking energy conservation since I am not using diffuse at all, just SSS and with very small modifications to Jensen's measured settings, so it's still physically plausible =)
It's biologically improbable for most people over preschool age because their skin will develop thicker epidermis with age (which, being a rough layer of dead cells, is what cancels the dermal scatter out the most). Let alone discolourations etc, which are also mostly epidermis-based.
Epidermis thickens due to exposure to UV rays (the sun) and harsh conditions like wind, heat etc. Its function is to protect the living layers of skin.
I'm writing a tip list for the docs which will include guidelines like this; it's really easy to add in a diffuse layer with the helper scripts, so it's a win-win situation when it comes to flexibility. And apart from my personal thing, my basic presets use "bare" SSS simply for the reason of clearly showing how it looks. I expect the user to add their own touch.
...a lot of DS users render improbably beautiful young females anyway. These "faeries" could use some fairy-tale translucent skin LOL
I hope to get a render of a "rugged" character done during the weekend.
Thank you! Downloading now... can the TIFF be 32-bit?
Found a couple of free LUTs here: http://www.groundcontrolcolor.com/free-luts.html
Yup, and that's DS output format
Until you get to the other end of the spectrum...the very elderly will be the same way, too.
And for creatures of faerie, it's perfectly acceptable to have very translucent, glowing skin...they are timeless, immortal beings, after all. (Either that, or they use some pretty damn good moisturizers...)
Yes. It's still physically plausible. :) And a lot of DAZ users really, really like those translucent skin shade.
Some test renders of fabric presets. Hair still needs some work and not having any bump maps included didn't help either.
If I'm honest, I think SSS is way overused, almost to the point of obsession. In real life and in movies and TV shows I've seen the only real SSS I see is in the ears when light hits them. Nobody glows orange like a belisha beacon. This is why a lot of CG skin looks so fake as it's way too orange. I live in the UK and people, unless they've been to a tanning salon, just don't look like that.
I will pay attention to what Kettu has done and will see what it all looks like tomorrow
CHEERS!
Actually, what you describe isn't subsurface scattering per se, but the backscatter part. Just like skin details we've covered earlier, they are always present, but not always visible. You won't notice it when the light is dim or evenly distributed.
Here's a very nice post by Neil Blevins:
http://www.neilblevins.com/cg_education/translucency/translucency.htm
Does the skin on the render above uses too much SSS? Probably, it's a new approach to doing skin I'm still working on, so it's not quite there yet. But those renders.are mainly done to see the fabric, not the skin.
Well, what do you know...looks like mucb better now with more tweaks. Didn't even have to resort to using SSS maps, putting the diffuse into SSS color slots or desaturating the diffuse. Still looks Very nice with an HDRI IBL pushed to outrageous contrast.
The fabrics look awesome (even though the texture artist seems not to have tiled the leather texture finely enough), as does the girl in both this and the next post. That braid looks challenging, though. Could you please post results if you do manage to make it look like hair? These 'solid' hairpieces are so tough to set up convincingly...
Cool, thanks a lot for the links! That car render looks really neat.
And for creatures of faerie, it's perfectly acceptable to have very translucent, glowing skin...they are timeless, immortal beings, after all. (Either that, or they use some pretty damn good moisturizers...)
Along with barrels of glitter? Andromeda's makeup artists did quite a lovely job on Trance Gemini (the Golden version), particularly in seasons 4 and 5. Okay she's technically not a faerie/elf, but as close as it gets in a mystical space opera =)
Promo shot: http://trance-gemini.com/wordpress/wp-content/uploads/2011/11/trance1002-copy-746x1080.jpg
Screenshots: http://trance-gemini.com/wordpress/?page_id=293
It's also one of those areas where The Hobbit movie beats the LotR one IMO: the elves look much more convincing (non-human) with all those glowing filters n´stuff.
@Wowie: You'll never get good results with Daz characters and photo backdrops, they are like chalk and cheese. Put her in a Stonemason set and she'll look better. The later renders of her do look better though. It's Kettu's shader and you should follow what Kettu does with it as he knows best seeing as he created it. It probably was backscatter that I was thinking of and nobody walks around permanently backlit. Most of the lighting I see is from one source and that is daylight. I want to keep lighting to a minimum number of lights and see what I get.
CHEERS!
Well, Rogerbee, but if I really wanted everyone to follow me, I wouldn't have exposed half the controls in the shader. I like what Wowie does, he's got a style of his own, and that's great. I don't see the point of everyone 200% copying someone else. That's... boring.
Who's the artist here? You folks are. I give you the tools, you make them your own.
PS: even though Kettu may not feel much like it... Kettu is a girl.
That car was actually a 3 yr old RIB I used to test out 11.0.164...it still needed 6 hrs to render, but that's way down from the 12+ it took 3 yrs ago. But since I had it active in idisplay, (just finished rendering) it was available to play with the LUTs on.
And yeah, I guess being the living avatar of a star means one should be all 'glowy'...
SSS backscattering:
You o esteemed alpha testers may have noticed there's a bunch of dedicated controls for extra SSS backscatter - in the sense of "extra scatter when backlit" (I´ve seen this word applied to retroreflectors as well, which is a different thing).
My take on this amplifies the SSS result (with optional tinting) according to Beer's absorption law: thinnest parts absorb less light = they get SSS amped accordingly.
My suggested workflow is this:
1. Set up your back light.
2. Choose the surface on which you need that extra backscatter. Choose wisely =) It will add some render time, even if not that much.
3. Turn off the main SSS strength temporarily; this will make previews faster. You may also wish to disable specular and your optional diffuse layers, if it all distracts you.
4. Enable the backscatter controls (flip "use extra SSS backscatter" to 1).
5. Find the "Back light preview (debug feature)" switch and turn it on (set to 1).
6. Now, if you render, you will see a white mask on the thinnest parts of the mesh. If you don´t see anything, play with the parameters or choose a better surface to test on (like ears).
7. The higher absorption strength and light decay parameters are, the more light gets absorbed. These are non-linear controls, so adjust them both until you get the distribution you like. May take some fiddling.
8. Disable the preview; re-enable your SSS, specular etc.
9. Render.
10. Finetune the back light tint to get the exact result you want.
Caveat:
You may get splashes of colour on eyelids, nostrils and the like. It's when the ray doing the Beer's law hits internal geometry. This is what the "Backscatter accounts for cavities" switch is for: flip it on and everything should fix itself. // the ray will start calculating if it actually sees the backlighting or not, when it hits that next wall) //
I am thinking about making the cavity fix turned on by default (because it makes physical sense), but it may add to render time some more.
Here are very basic example renders (M4, Beck texture, absorption and decay both with a value of 3, tint 255-183-149).
http://www.youtube.com/watch?v=13S58aQw7wA
Well, Rogerbee, but if I really wanted everyone to follow me, I wouldn't have exposed half the controls in the shader. I like what Wowie does, he's got a style of his own, and that's great. I don't see the point of everyone 200% copying someone else. That's... boring.
Who's the artist here? You folks are. I give you the tools, you make them your own.
PS: even though Kettu may not feel much like it... Kettu is a girl.
Fair enough, but, I like to learn from people that get the results I like to see, and that, right now, is you. My bad for not realising you were a girl, you're very talented, as is Mec4D who is also a girl of course. I might do things differently with lighting, but, I'll still cull the basics of that from you.
I like style to a certain degree, but, there has to be substance as well
CHEERS!
Oh my god, Cath (Mec4D) is a real artist, what with all her texture and modeling work (let alone music =)), and I'm just a wannabe TD. Thank you for your kind words, but again... nothing I do is based on any "talent". Rather, on just being able to read XD
Yes. It's still physically plausible. :) And a lot of DAZ users really, really like those translucent skin shade.
Some test renders of fabric presets. Hair still needs some work and not having any bump maps included didn't help either.
The skin in looks amazing. It has this balance of being plausible while still looking CG-ish and not getting uncanny valley.
That's a little unfair. I guess if your talking straight renders it's kind of true.
With postwork Daz characters and models look great with photo backdrops.
That's a little unfair. I guess if your talking straight renders it's kind of true.
With postwork Daz characters and models look great with photo backdrops.
It's doable in a render, too...except it's more work to do it in a render.
Yeah, that sounds odd, at first...
But, consider this, to get great results with a photo backdrop, you need a hi-res, high detail image. Then you need to get the scaling just right...for that it would help to know some of the real world sizes of key elements in the photo, to use as a reference for scaling. Then you have to do the hard part...shadow/light matching. It helps if you are using a HDR image and backdrop pair...but there's still a bit of work involved in setting it all up.
And this isn't just a Studio thing...it's pretty much the same with any model/render.
It's doable in a render, too...except it's more work to do it in a render.
Yeah, that sounds odd, at first...
But, consider this, to get great results with a photo backdrop, you need a hi-res, high detail image. Then you need to get the scaling just right...for that it would help to know some of the real world sizes of key elements in the photo, to use as a reference for scaling. Then you have to do the hard part...shadow/light matching. It helps if you are using a HDR image and backdrop pair...but there's still a bit of work involved in setting it all up.
And this isn't just a Studio thing...it's pretty much the same with any model/render.
That doesn't sound odd at all I actually meant postwork on top of all that.
You can get it close inside a 3d program but comping it together in outside compositing software helps for lots of reasons, namely tweaking the background image to match the CG. I know most people think of trying to get the CG elements to match the real world, but it works the other way too.
A lot of times it looks seriously cool to get the CG to look close to photoreal, and match the real world elements to the CG assets.
Of course it gets easier if you are doing external tonemapping...render to something like EXR and then tonemap in something else.
Yes, the 'leather' part looks best when it's using just plain fabric preset. So I guess, that's not really leather. :) Or at least, it looks better with the diffuse roughness set to very rough.
As for the braid, don't know if I can get it to look better without a bump map. I could generate one, but generally I would've preferred vendors doing that. Same thing with general solid hair pieces - you need displacement and bump to get details and in generaly i found it is just easier to use more complex hair models than making those things look good.
There's also the skull cap, which I think with most hair pieces/prop are just plain wrong - even the fibermesh ones.
She is the only one that has really cracked Iray, but then, she has worked far more extensively with that type of render engine than anyone I know. She was also the first to do really major work with HDRI but more in Poser than DS. It was her work with textures that really inspired me and I still like to use some of her textures even now.
You have inspired me just as much though and I hope to take DS as far with you as I did with Cath in Poser
CHEERS!
Sorry, it's not Kettu's shader. :) I'm not using the Radium shaders for those shots.
And I don't think Kettu is a he. :)
Yeah, sorry, I worked both out after I posted. It had been a long day and I wasn't really reading posts that clearly. CGI in movies now seems to have the characters and background as CG elements, allowing them to match each other closely. It takes a lot of work matching the detail and lighting of a photo to the same aspects of a CG character. Keep at it if you want to, but I just feel that CG in CG looks better and takes less time and matching as you have more precise control of the environment.
CHEERS!
For adventurous people who'd like one more option to get film preset there is another option I found.
While you could get a free version of compositing tools like Nuke, Da Vinci Resolve Lite or Fusion and use a LUT file inside of these;
there is also an Open Source Software that I didn't test till now and that has a lot of camera presets called Natron
Compositig softwares may seem complicated but if it's just to apply a camera preset here are simple steps for Natron
1°/ Load an Image Read Node
2°/ Load an OCIOLookTransform node (in Color->OCIO)
3°/ Load a preset from the "Look" in the OCIOLooktransform Panel
Just be careful to the input and output color space in your Read and OCIOLookTranform nodes if you want something correct
When you're happy you just have to use a "Write" node