HOME | DD

#caustics #dive #diving #girl #pool #redhead #swim #underwater #reflections #reflectionwater #octanerender
Published: 2015-04-27 16:00:27 +0000 UTC; Views: 11234; Favourites: 110; Downloads: 290
Redirect to original
Description
Some kind of swimming pool fun--
Here are some tech thoughts:
this is my first try of making underwater scenes... there are some issues with water material (no scattering for fog-like rays effect, for example) and lack of underwater swirls; more attention was given to reflections in water surface and caustics effect. The lighting is made using Octane's sun with high intensity; PMC render kernel at 30k spp was used (because PT produces totally different and very strange results).
Related content
Comments: 39
MouronRouge [2023-06-09 04:24:52 +0000 UTC]
π: 1 β©: 0
pumpkinman68 [2017-09-18 19:11:03 +0000 UTC]
I'm jealous when I look at your underwater caustic Β .Β I tried an underwater scene too, but I'm not satisfied by the result. I managed to have a good caustic effect, but with a "bad geometry" (eq. the nice "bee alveole" you obtained). It seems that none of the water surfaces I tried gives the correct caustic effect. So, if it's not a secret, could you tell me what surface do you use ? I'm a bit afraid that you said something like "a octane water surface you can't have with Iray", which will drive me to futur other boring tries .
π: 0 β©: 1
pnn32 In reply to pumpkinman68 [2017-09-18 19:24:51 +0000 UTC]
There are a lot of settings of Octane render kernel for achieving such effect; moreover, this is not physically correct behavior... there is no water volume, only top plane.
The material setting for this plane is simple - specular material (IoR=1.33) with bump map simulating waves (no geometry waves, displacement or whatever) - same material was used in all my renders featuring water. The important part is light source (very intensive sunlight), view angle and render time - I wait for several minutes at least to get some faint visible effect (the first results looks like aΒ total mess).
π: 0 β©: 1
pumpkinman68 In reply to pnn32 [2017-09-19 08:34:19 +0000 UTC]
May be a little misunderstanding between us (well... it's quiet difficult to explain...). Β
In a simple sentence, according to your explainations : I need your bumpmap ! Β Β
I succeed quiet well on Iray to obtainedΒ a global underwater effect. Moreover, there's a good chance that Octane and Iray don't react the same way, so settings about "specular" or light,... are quiet useless.
We have the same logic to create underwater effect : I use a plane too and not a water volume (for now, according to my tests on Iray, volumes "kill" caustics, but I'll continue the tests); I use the map as normal map, and not as bumpmap,... Same "basic logic", not the same rendering engine... so "settings" are differents. BUT, whatever the settings of the surface of each engine, one point remains : the drawing of the map give a precise drawing of the caustics !
The fact is that all the maps I used to simulate the wave give bad caustics effect. Not a bad color, nor a bad light effect, or something else,... but a bad drawing, a bad "caustics geometry" (that's why I talk about the caustics effect which looks like "deformed hexagonal bee cell" on the skin). Have a look on this picture : Laguna: Underwater Lights for Iray . It's just an exemple, but that illustrates very well my words : That's NOT what I call realistic caustics. Unfortunatly, for now, I can't find a good map to create realistic caustic .
π: 0 β©: 1
pnn32 In reply to pumpkinman68 [2017-09-19 18:54:37 +0000 UTC]
I've got that work files from my archive, blown off the dust (oh my.. I've done it in Octane v2.17 two years ago ), loaded scene into Octane, and...
sorry, I have bad news for you
As you can see, the bump map is made as procedural texture (Turbulence), i.e. it is generated by Octane and instantly applied to material in the internals of render engine.
My other thoughts are:
1. Octane can handle some complex lighting conditions, which Iray can't...
for example, 'camera obscura' inverted image projection effect with small hole in closed room wall
2. Octane still has issues with some materials combinations, for example 'water in glass volume' combination will require some tricky modeling and result will not look good from any angle
So, you just need to do more complex experiments
π: 0 β©: 1
pumpkinman68 In reply to pnn32 [2017-09-19 19:19:58 +0000 UTC]
Thanx for the research, I suspected something like that Β (I came from Lightwave, powerful 3D software but I let it drop : no tutorials, to many time spend to understand the render engine).
>> "So, you just need to do more complex experiments"
Well... I'm afraid this will be boring tests because mainly oriented in imagemap search. May be I'll try to find a 2D map generator...
Thanx again, and keep on the good work !
π: 0 β©: 0
LikaKinsky [2017-08-25 22:20:55 +0000 UTC]
Wow! love this one! I adore underwater scenes and this one is just magnificent!
π: 0 β©: 1
pnn32 In reply to LikaKinsky [2017-08-26 06:59:26 +0000 UTC]
Such scenes are also hard to setup and render.. typically it's a long way of trials & errors
π: 0 β©: 1
JoePingleton [2016-09-22 18:14:41 +0000 UTC]
excellent work, she really feels like she is swimming
π: 0 β©: 0
pnn32 In reply to Skiriki [2015-05-08 16:54:06 +0000 UTC]
It's still one of the effects hard to achieve
π: 0 β©: 0
Phyl-CGI [2015-04-30 15:21:05 +0000 UTC]
Absolutely stunning
May I ask: did you postworked the hair? it's awesome...
π: 0 β©: 1
pnn32 In reply to Phyl-CGI [2015-04-30 17:02:00 +0000 UTC]
There is no postwork at all... (because I can't work well with raster images )
The hair was the biggest problem during render (note the hair alpha map issue on the line between right eye and right breast); the amount of noise pixels was another problem.
π: 0 β©: 1
Phyl-CGI In reply to pnn32 [2015-04-30 17:15:17 +0000 UTC]
Is this one of those classical transmapped hair? If yes, have you tried using one of Aeon-Soul's hair props? The closest one is Desir Hair . But be careful as these hair is composed of tens of thin strands, making the project file to explode in size and the time to render explode in length.
Oh and speaking of postworking: I am learning, I am still novice at it. Just in case it interest you, I started to practice while reading the following picture's description, from kobaltkween
Β Β Β
(She has a few excellent tutorials on Poser too, but not of any use with Octane)
π: 0 β©: 0
ttobserve [2015-04-27 22:58:16 +0000 UTC]
Whoa! Β Β
You'reΒ givingΒ SachaKalis Β some serious competition!
π: 0 β©: 1
pnn32 In reply to ttobserve [2015-04-28 16:51:51 +0000 UTC]
Whoops Β
I must also confess, that I was looked a lot of their and similar images for getting inspiration...
Definitely fav.me/d6xbit4 and fav.me/d6y67vt ones.
π: 0 β©: 1
ttobserve In reply to pnn32 [2015-04-29 09:59:58 +0000 UTC]
Right; lovely:Β Β !
But those are actual photographs,
whereas YOU, Maestro, are getting there with pure CG! Β
π: 0 β©: 0
Kooki99 [2015-04-27 22:40:48 +0000 UTC]
fantastic scene !!
Undewater is so hard to master, this is very nice !
The light passing through the water is extremely realistic as well as hair movements
π: 0 β©: 1
pnn32 In reply to Kooki99 [2015-04-28 16:53:52 +0000 UTC]
There are a lot of things to improve.. scattering for example (light rays should be visible). But render speed will drop dramatically...
π: 0 β©: 1
Kooki99 In reply to pnn32 [2015-04-28 18:17:42 +0000 UTC]
indeed, i hope you have Quad SLI
π: 0 β©: 0
JohnKent42 [2015-04-27 20:20:57 +0000 UTC]
Really well done - the caustics are great and the hair swirls work really well too
π: 0 β©: 1
pnn32 In reply to JohnKent42 [2015-04-27 20:32:05 +0000 UTC]
Thank you, the result was worth the time for experiments with render options
π: 0 β©: 0
ThierryCravatte [2015-04-27 16:40:03 +0000 UTC]
Great result, mate ! The caustics looks nice ! Thank you very much for the tips !
π: 0 β©: 1
pnn32 In reply to ThierryCravatte [2015-04-27 17:01:51 +0000 UTC]
...and a lot of noise pixels as bonus
π: 0 β©: 1
joobiewoobie [2015-04-27 16:12:24 +0000 UTC]
This is truly amazing. I love using water in octane, it always turns out good. Great job, I love it!!
π: 0 β©: 1
pnn32 In reply to joobiewoobie [2015-04-27 16:20:17 +0000 UTC]
Here are still soooo many things to improveΒ
π: 0 β©: 0