PDA

View Full Version : Anyone familiar with Ska? (Free SSS plugin)


milqman
11-28-2003, 10:39 PM
Howdy

Is anyone familiar with Ska, the free subsurface scattering plugin? It's mad slow, adds 20 minutes to my render... so its nerveracking to not know what the settings do and have it take forever.

Can someone explain how it works?

Thanks

SplineGod
11-29-2003, 03:58 AM
I think you can duplicate many of the effects of SSS in LW using gradients. :)

chikega
11-29-2003, 04:09 AM
There's a tutorial on NT's site:

http://www.newtek.com/products/lightwave/contests/dec-02/winners/tutorials/layout/Shawn_Sapp/index.html

Hope that helps - and yes render times tend to be long.:)

stompbox
11-29-2003, 04:58 AM
Originally posted by SplineGod
I think you can duplicate many of the effects of SSS in LW using gradients. :)

Could you elaborate on that at all Larry?

I am a relative newbie, slogged my way up the worst of the learning curve, but wouldn't know where to start with what you're talking about. I like the idea of SSS for skin, but also plasticene look like boring3d.com
Cheers!

milqman
11-29-2003, 05:07 AM
Heheh, thanks again Larry

I agree with the previous poster (on the elaboration)... I'm not too sharp with gradients yet.

SplineGod
11-29-2003, 05:12 AM
Basically SSS gives a kind of internal luminance from light scattered below the surface of a material. Gradients can be used in the luminosity channel usually in conjuction with light or camera incidence to give the effect. Additional maps in the luminosity channel and translucency channel can help intensify the effect.
This picture was done using only luminosity maps. No lights are used at all. Theres not even an object inside either. :)
http://www.3dtrainingonline.com/examples/embryo.jpg

milqman
11-29-2003, 05:34 AM
Could you post that scene file somewhere so i could take a look at your surface set up?

chikega
11-29-2003, 05:38 AM
Originally posted by SplineGod
Basically SSS gives a kind of internal luminance from light scattered below the surface of a material. Gradients can be used in the luminosity channel usually in conjuction with light or camera incidence to give the effect. Additional maps in the luminosity channel and translucency channel can help intensify the effect.
This picture was done using only luminosity maps. No lights are used at all. Theres not even an object inside either. :)


Very cool, Larry! I like the effect.:)

leigh
11-29-2003, 06:27 AM
I personally don't like the effect that the gradients give at all, and I don't really think that it looks much like subsurface scattering either :shrug:
It's too strong (subsurface scattering is far more subtle), and using luminosity on surfaces is not only unnatural but it also screws around with any illumination in the scene, as well as radiosity, if you are using it.
Real subsurface scattering is not an angle-based effect, it is a volumetric effect.

Personally I think that if a plugin like Ska can give good results (Werner uses it to great effect), then the extra rendering time is worth it.

milqman
11-29-2003, 06:45 AM
Does one have to turn up transparency to use SSS?

SplineGod
11-29-2003, 07:00 AM
Originally posted by Leigh
I personally don't like the effect that the gradients give at all, and I don't really think that it looks much like subsurface scattering either :shrug:
It's too strong (subsurface scattering is far more subtle), and using luminosity on surfaces is not only unnatural but it also screws around with any illumination in the scene, as well as radiosity, if you are using it.
Real subsurface scattering is not an angle-based effect, it is a volumetric effect.

Personally I think that if a plugin like Ska can give good results (Werner uses it to great effect), then the extra rendering time is worth it.
Thanks Gary,
It was a fun little experiment. :)

I agree that Real subsurface scattering isnt an angle based effect and that SKA may give better results and that gradients may or may not give the same effect.
Most people dont really even know what SSS is. Milqman said he didnt understand gradients and many times Ive found that what a lot of people are shooting for with SSS can be done using gradients, sometimes they cant. A lot of people see the result of SSS or get a result with it that can be done more easily in other ways. In the end I have no idea if gradients will give the desired effect but it certainly couldnt hurt. It never hurts to learn how to fake effects using tricks or cheats. :)

HERES (http://www.3dtrainingonline.com/examples/pod.zip) the zip file with the whole setup.

telamon
11-29-2003, 10:07 AM
thanks again splinegod... Your inner reference library is amazing :D

One question... Is it possible to use such a plugin or G2 SSS algorithm in production of long animation ? Is it as slow as radiosity (for instance).

SplineGod
11-29-2003, 10:21 AM
Thanks! :)
To be honest I havent seen anyone use G2s SSS in production or any form of it. Generally people seem to shy away from anything that takes very long to render. The exceptions would probably be on a film where you may generally have longer to render.
Ive never even really seen Radiosity used except that the enterprise team when they were at FI started to use some backdrop radiosity to sweeten the lighting a bit. Apparently this was fast enough to be acceptable.
I would guess that Worleys version of SSS is probably as fast or faster then anything else out there. Now as to whether its quick enough to be useful in production I dont know. Again people tend to cheat and fake things whenever possible. :)

tOd
11-29-2003, 02:57 PM
I used SSS from G2 on one of our recent productions. There were some close up shots of a girl's face where it worked really nice to soften the skin. It didnt lengthen the render times much at all.

:thumbsup:

milqman
11-29-2003, 03:53 PM
Sigh. I really like the effect Ska is giving me, but... I started this 1280x960 render last night, with 1x3 radiosity, ska turned on two surfaces, camera DOF, and enhanced medium AA...

its now morning and its only on the 3rd pass.

:cry:

tOd
11-29-2003, 06:00 PM
Originally posted by milqman
Sigh. I really like the effect Ska is giving me, but... I started this 1280x960 render last night, with 1x3 radiosity, ska turned on two surfaces, camera DOF, and enhanced medium AA...

its now morning and its only on the 3rd pass.

:cry:

We found better render times with G2 than Ska. Ska was really slow.

Finkster
11-29-2003, 07:12 PM
Originally posted by milqman
Sigh. I really like the effect Ska is giving me, but... I started this 1280x960 render last night, with 1x3 radiosity, ska turned on two surfaces, camera DOF, and enhanced medium AA...

its now morning and its only on the 3rd pass.

:cry:

Does ska make use of only particular lights in a scene, or does it make use of all the illumination (including radiosity) in a scene? If it is making use of the radiosity, maybe you should do seperate passes - a radiosoty pass (no SSS) and a seperate Ska pass (no radiosity).
That might speed things up alot if radiosity+Ska is so slow.

Now if you'll excuse me, all this talk of ska has put me in the mood for a dance.......
http://usuarios.lycos.es/sound_selekter/images/ska.gif

CG.p
11-29-2003, 07:15 PM
Originally posted by milqman
Sigh. I really like the effect Ska is giving me, but... I started this 1280x960 render last night, with 1x3 radiosity, ska turned on two surfaces, camera DOF, and enhanced medium AA...

its now morning and its only on the 3rd pass.


What is in the image? Radiosity might not be needed if ska isn't updated to use it. A shader needs to be radiosity aware.

Got a plain version of the image you can post?

comanche
11-29-2003, 07:30 PM
Originally posted by Leigh
... and using luminosity on surfaces is not only unnatural but it also screws around with any illumination in the scene, as well as radiosity, if you are using it.

But SSS does nothing else than adding luminosity to the surface!?

A well made fake can be a real time saver and can also look quite nice imo.

Cheers,
Andreas

CG.p
11-29-2003, 07:46 PM
Originally posted by comanche
But SSS does nothing else than adding luminosity to the surface!?

The same can be said of the diffuse channel, using the angle this comment is coming from.


A well made fake can be a real time saver and can also look quite nice imo.

Yes if can and has it's place..right until it blows up in your face during production and the brute force true SSS might need to be used. Faking can be the best way to go with time-eating calculations. But there IS a trade-off when you do somehting that doesn't hold up to every angle, every lighting sceme whatever. Look at the normal mapping stuff. The currect offerings are falling apart when someone deforms/rotates some models. Extreme cases can be understood but when a simple operation is done and it falls apart..that will put your butt on the line for wanting to cheat it.

SplineGod
11-29-2003, 08:32 PM
Originally posted by CG.p

Yes if can and has it's place..right until it blows up in your face during production and the brute force true SSS might need to be used. Faking can be the best way to go with time-eating calculations. But there IS a trade-off when you do somehting that doesn't hold up to every angle, every lighting sceme whatever. Look at the normal mapping stuff. The currect offerings are falling apart when someone deforms/rotates some models. Extreme cases can be understood but when a simple operation is done and it falls apart..that will put your butt on the line for wanting to cheat it.

I dont think it really matters in many cases all that much. Every shot is a trade off. Any technique, tool or plugin can potententially blow up in ones face at any time. You use what works for that shot to get it done within the constraints imposed by the show. The artists butt is always on the line no matter what. Part of the artists job IS to come up with creative solutions. You put characters in a room with the same light kit that was being used in another shot, change the camera angle and suddenly it doesnt look as good. Lighting can change fom scene to scene even for the exact same environment. The same applies to the use of SSS vs Lighting vs Gradients ; use whatever works.

CIM
11-29-2003, 09:37 PM
Originally posted by comanche
But SSS does nothing else than adding luminosity to the surface!?


That's not even close to what subsurface scattering is -- in fact, that's totally false. Try reading some of the papers out there, such as http://graphics.stanford.edu/papers/bssrdf/.

SplineGod
11-29-2003, 09:57 PM
Of course reading those papers wont give you much of an insight into whats going on. If you want to ACCURATELY SIMULATE SSS its going to take some serious crunching time. In the end I just want a fast way of getting the same effect and with a lot of latitude in how I can achieve those effects. I dont care if its accurate or not. I want to be able to get into the ball park quickly and then have tools that will allow me to tweak the last 20%.
Its like Motion Designer...I just want a quick way of getting some cloth like movement but then let me have the ability to manipulate points to get things just the way I want AFTER it gets me 80% of the way.

On the 2nd DVD that came with Attack of the Clones George Lucas was watching the animation of Yoda doing his whirling dervish bit when fighting Christopher Lee. His cloak was wrapping around him while he fought and Lucas pointed it out to the artist who said that the SIMULATION was correct. Lucas stated that while it might be correct TECHNICALLY it wasnt correct ROMANTICALLY. In other words make the cloak billow and look cool even though it wasnt correct technically. :)

All of these tools should be more like a paint brush rather then a simulation. If we treated lights as a simulation we wouldnt have the ability to tweak shadows or exclude things from the effects of light.

Mattoo
11-29-2003, 10:48 PM
Originally posted by SplineGod
Of course reading those papers wont give you much of an insight into whats going on. If you want to ACCURATELY SIMULATE SSS its going to take some serious crunching time. In the end I just want a fast way of getting the same effect and with a lot of latitude in how I can achieve those effects. I dont care if its accurate or not. I want to be able to get into the ball park quickly and then have tools that will allow me to tweak the last 20%.

SNIP
.

You just described G2. It's not completely accurate but it's bloody fast. Infact I'd happily trade in some of that speed for more accuracy.

And as to how much radiosity and other advanced rendering techniques have been used in production - there's loads of examples ranging from Lord of the Rings(SSS), Pearl Harbour (GI) to Gladiator (GI) and no doubt many more.

CIM
11-30-2003, 02:12 AM
Of course reading those papers wont give you much of an insight into whats going on. If you want to ACCURATELY SIMULATE SSS its going to take some serious crunching time.

How do you expect to accurately fake something if you don't know anything about it (as many in this thread have demonstrated)?

And as to how much radiosity and other advanced rendering techniques have been used in production - there's loads of examples ranging from Lord of the Rings(SSS), Pearl Harbour (GI) to Gladiator (GI) and no doubt many more.

That's not true. ILM (and most large productions) use ambient occlusion techniques, because radiosity or GI is far too slow for production use.

sebek27
11-30-2003, 02:16 AM
ambient occlusion techniques - can someone explain this or show me a link to this method ? thx ~!

SplineGod
11-30-2003, 02:19 AM
Originally posted by CIM
How do you expect to accurately fake something if you don't know anything about it (as many in this thread have demonstrated)?



That's not true. ILM (and most large productions) use ambient occlusion techniques, because radiosity or GI is far too slow for production use.

You fake the appearance or the effect of something. For example on the Phantom menace DVD a whole compositing sequence is shown were salt is poured to look like a waterfall. Also, Foundation Imaging in the sequence where Voyager crashes into the snow they used an air hose and baking powder to get the effect of snow. Thats all Im interested in fast ways to get the "look" without the calculation intensive simulation.
I sometimes use luminosity maps to fake internal illumination just like we use bump maps to fake real bumps.

CG.p
11-30-2003, 02:35 AM
Originally posted by sebek27
ambient occlusion techniques - can someone explain this or show me a link to this method ? thx ~!

It is what the gMil shader does. It checks to see if there is an object between the shaded spot and the background. almost like background radiosity but not quite. I forget the exact difference.

CourtJester
11-30-2003, 04:22 AM
Originally posted by CG.p
Look at the normal mapping stuff. The currect offerings are falling apart when someone deforms/rotates some models.[/B]

If by "normal mapping" you are referring to the Microwave type stuff where details in geometry is turned into bump-mapped fakery, that's not the fault of normal mapping. It's a LightWave bug affecting bump maps on deforming objects.

Mattoo
11-30-2003, 04:25 AM
Originally posted by CIM
How do you expect to accurately fake something if you don't know anything about it (as many in this thread have demonstrated)?


That's not true. ILM (and most large productions) use ambient occlusion techniques, because radiosity or GI is far too slow for production use.

Hah, you got me CIM, only a true Cinefex nerd would have spotted the Pearl Harbour exageration. Granted, ambient occlusion is not radiosity even in the loosest sense but it is a more advanced rendering technique beyond what we consider standard lighting.
Regardless, the other 2 examples are valid and as I said, if I had the inclination I could probably list many more.

CIM
11-30-2003, 04:54 AM
Originally posted by Mattoo
Hah, you got me CIM, only a true Cinefex nerd would have spotted the Pearl Harbour exageration. Granted, ambient occlusion is not radiosity even in the loosest sense but it is a more advanced rendering technique beyond what we consider standard lighting.
Regardless, the other 2 examples are valid and as I said, if I had the inclination I could probably list many more.

Actually, I read about it in a RenderMan book.

CG.p
11-30-2003, 05:01 AM
Originally posted by CourtJester
If by "normal mapping" you are referring to the Microwave type stuff where details in geometry is turned into bump-mapped fakery, that's not the fault of normal mapping. It's a LightWave bug affecting bump maps on deforming objects.

That "microwave type bump mapping" isn't bump mapping, it is normal mapping. There is a difference.

Still it is something that can fall apart whether it is the fault of the programmer or Lightwave. That was my point.

milqman
11-30-2003, 05:31 AM
argh. remember that render i was talking about in my last post? that was taking forever? well. its on its eigth pass NOW. out of 9. arrrrrrrrrrrrrrrghhhhh!

i predict it'll take a full 24 bloody hours. i'll post the result here

SplineGod
11-30-2003, 05:52 AM
Originally posted by milqman
argh. remember that render i was talking about in my last post? that was taking forever? well. its on its eigth pass NOW. out of 9. arrrrrrrrrrrrrrrghhhhh!

i predict it'll take a full 24 bloody hours. i'll post the result here

Use the CHEAT Luke! :)

vvee
11-30-2003, 06:28 AM
In defense of SplineGod, i think we all should remember that we are not in fact "creating" any actual real world, hold em in your hand, twist em around and smack em against the wall anything. We are only making computer simulations.

In my opinion, if looks like SSS and smells like SSS…

When it come to real world physics, we can't even decide what light is, not to mention the millions of other phenomenon we have no idea how or why they work (take gravity for instance).

chikega
11-30-2003, 06:32 AM
Originally posted by Leigh
It's too strong (subsurface scattering is far more subtle)
That depends on the translucency of the object. If you're talking about skin - yes it's subtle. If you're talking about something like colored glass - it's not very subtle at all.
Real subsurface scattering is not an angle-based effect, it is a volumetric effect. It is both volumetric and angle based. It's just not based on the glancing angle or incidence of the camera (like fresnel) but rather based on the angle of incidence of light - especially when concerning skin.:)

--------------------------------------------

Also there are mainly two types of SSS solutions out on the market. You have faked SSS (like G2) and volumetric-based (like OGO_Hikari, sKA). Of course, the volumetric implementation has severe rendering penalties. Most special effect houses use the faked implementation for the same reasons most don't use radiosity just yet.

G2 is extremely fast b/c it is a reasonable emulation of SSS(faked). Essentially, with G2, you can't stick an object inside of a translucent object and expect to see a hint of it within. With OGO_Hikari (http://www.asahi-net.or.jp/~pq1a-ogs/block_e.html) this is possible. It's always a trade off.

If anyone hasn't seen these, here are some of my own experiments based on G2 rev1:
http://www.3ddmd.com/test_renders.htm

milqman
11-30-2003, 06:33 AM
I'll try it next... but I'd like to wait for this render to finish.

Each pass seems to be taking 3 hours.

So Larry, your gradient trick... is that using incidence angle... light incidence... what?

CG.p
11-30-2003, 06:53 AM
For the SSS cheat you need to use the surface thickness gradient.

comanche
11-30-2003, 07:40 AM
Originally posted by CIM
That's not even close to what subsurface scattering is -- in fact, that's totally false. Try reading some of the papers out there, such as http://graphics.stanford.edu/papers/bssrdf/.

Hi CIM,

I do know what SSS is, even if my posting sounds dumb. :)

What I meant was that in the rendered image, the SSS calculation does add luminosity (and even change color) to the pixel, so it appears lighter than without SSS (the pixel!).

If it wouldn't, you would not notice the SSS effect, right?

SplineGod
11-30-2003, 08:06 AM
Originally posted by milqman
I'll try it next... but I'd like to wait for this render to finish.

Each pass seems to be taking 3 hours.

So Larry, your gradient trick... is that using incidence angle... light incidence... what?
Keep in mind that luminosity maps are the basis of getting the effect but other channels can be used as well.
Also you dont have to strictly do it with gradients either. Image maps can be used as well. They can be tied to reference objects so the texture can track the camera if desired. You can also use gradients. Not only with incidence angle but light incidence, distance to object, distance to camera etc. Gradients, like any other texture type can be blended in various ways and used as alphas to isolate the effects in specific areas. Turn on viper and experiment. :)
I took your mushroom idea and started playing with gradients in mostly the color and luminosity channels (hope you dont mind) and a bit in the diffuse. I didnt spend a whole lot of time on this but hopefully gives the idea. :)
I used gradients in with incidence angle and I used one with distance to object (a null) to get the red glow behind the stem.
I added in some procedurals to get a look like some areas are more translucent then others. Anyways I simply leave viper open and start with the basic gradients then start adding the procedurals with different types of blending modes until Im happy with that stage and then move on. All of this including the background are gradients and/or procedurals.

http://www.3dtrainingonline.com/examples/mushroom.jpg

E_Moelzer
11-30-2003, 08:22 AM
Well from my understanding there are various ways to fake the effect and one can even combine them for pretty good results.
The ones that comes to my mind are:
1.Using the thickness- gradient together with translucency.
2.The next is an incident Gradient with luminosity (as already mentioned by Larry).
3. Light- incidence together with diffuse. One can simply make the transition from light to dark a bit softer by blowing the diffuse- level up the more the surface is facing away fom the light).
4. Combinations of the effects above.

In this context, I think it is very sad that there is no texturing for the colorfilter though (this would help a lot for glass- like effects, even if this is not directly connected to subsurface- scattering).

I think one big disadvantage of many of the shaders mentioned here (like SKA), is that one cant control the strength of the effect over a surface.
I have tried a few solutions and things like the bridge of the nose or the forehead where the skin is very tight over bones tend to be much to translucent without textures to control the intensity of the effect. You will need the effect to be much stronger for ears and the softer cheeks, etc, etc.

I dont know whether G2 allows for that control, but if it does not, it is not really all that usefull.

I would also like to point out, that some nice soft lighting can help much more than SSS and I have seen many examples that proove that.
CU
Elmar

chikega
11-30-2003, 02:43 PM
Originally posted by E_Moelzer
I have tried a few solutions and things like the bridge of the nose or the forehead where the skin is very tight over bones tend to be much to translucent without textures to control the intensity of the effect. You will need the effect to be much stronger for ears and the softer cheeks, etc, etc.


With OGO_Hikari, you can control this by placing a skull object within the head object - since it's volume-based, allow for the requisite rendering penalties.

http://www.asahi-net.or.jp/~pq1a-ogs/simple_head.jpg


I dont know whether G2 allows for that control, but if it does not, it is not really all that usefull.


G2 rev1 (not G2 1.5) allows control by using gray-scale image maps in the translucency channel - since it is a "faked" solution, there are limitations in this area as well:

http://www.3ddmd.com/G2_test2a.htm

http://www.3ddmd.com/images/G2/SSS_Vein1.jpg

:)

milqman
11-30-2003, 05:04 PM
Doesn't the Surface Thickness come only in 7.5b and c? If so I guess I'll have to upgrade

Anyway. Here's what took 26 hours to render... 1x3 radiosity, med-enhanced aa, camera DOF, and ska turned on.

thanks btw for all the replies and help.

http://members.fortunecity.com/milqman/altviewc.txt

CG.p
11-30-2003, 05:11 PM
Originally posted by chikega
With OGO_Hikari, you can control this by placing a skull object within the head object - since it's volume-based, allow for the requisite rendering penalties.

G2 rev1 (not G2 1.5) allows control by using gray-scale image maps in the translucency channel - since it is a "faked" solution, there are limitations in this area as well

OGO_Hikari allows for the same thing with the translucency map.

SplineGod
11-30-2003, 09:14 PM
Originally posted by milqman
Doesn't the Surface Thickness come only in 7.5b and c? If so I guess I'll have to upgrade

Anyway. Here's what took 26 hours to render... 1x3 radiosity, med-enhanced aa, camera DOF, and ska turned on.

thanks btw for all the replies and help.

http://members.fortunecity.com/milqman/altviewc.txt
OUCH! You could paint in the effect by hand in far less time.
I think you could get the effect you want by cheating it (like any of this isnt a cheat anyways) and in far less time. :)
BTW its coming along nicely. :)

CIM
11-30-2003, 09:53 PM
Originally posted by comanche
Hi CIM,

I do know what SSS is, even if my posting sounds dumb. :)

What I meant was that in the rendered image, the SSS calculation does add luminosity (and even change color) to the pixel, so it appears lighter than without SSS (the pixel!).

If it wouldn't, you would not notice the SSS effect, right?

Subsurface scattering doesn't ADD luminosity. When light scatters through materials/layers, it obeys fresnel laws. Extinction and various other things play a part in the look too.

KillMe
11-30-2003, 09:59 PM
think of sss as depth senstive translucency

the thicker the object the less light amkes it through and whack on a load of blurr while you at it

SplineGod
11-30-2003, 10:02 PM
Originally posted by CIM
Subsurface scattering doesn't ADD luminosity. When light scatters through materials/layers, it obeys fresnel laws. Extinction and various other things play a part in the look too.
Right. The reason that the surface might appear to be more luminious is that the object, instead of appearing dark/opaque, appears to glow or have an increased luminosity because of the internally scattered light making its way thru.
In the end anything to keep the render times down, give me the look I want and allow me to tweak setting is fine. :)

E_Moelzer
11-30-2003, 10:07 PM
CIM
I think what comanche wanted to say was, that pixels that are affected by SSS are brighter than without. Which is why we use SSS after all.
Comparing the result of SKA to Larrys fake I have to say that I like Larrys result better even though it might be less accurate.
;)
It is good to know that G2 allows to control the SS- effect with a texture- map. Makes this much more interesting for me.

Chikega and CG.p:
Does that texture in the translucency- channel have to be volumetric in order to give the desired effect with Ogo Hikari?

CU
Elmar

milqman
11-30-2003, 11:30 PM
Yeh, I hardly think Ska is worth the rendertime.... at least in this case. I'm gonna try this "Surface Thickness"

SplineGod
11-30-2003, 11:44 PM
Accuracy is often times overrated. :)
Most of the time its not what I want or a client wants.
Ive seen cloud formations that people would claim looked fake if I did them as CGI. When making movies nobody relies on realistic lighting. They struggle to control shadows and use diffusers to reduce the harshess of outdoor lighting etc.
If you have a character with blue clothing and the client wants red lighting, accuracy would mean the clothing would look purple. The clients wants red lighting and the clothing to stay blue which is not accurate. Even with SSS whos to say what looks correct or not when there are so many variables that change the look of SSS in nature based on the material, light intensity, wavelength and so on? In the end its the final output that only matters and the most efficient way to get to it. Ive never had anyone care how I achieved a look so long as it was what they wanted. :)

CG.p
12-01-2003, 02:20 AM
Originally posted by E_Moelzer
Chikega and CG.p:
Does that texture in the translucency- channel have to be volumetric in order to give the desired effect with Ogo Hikari?


Hikari: doesn't have to be, it handles volumetric textures also though.

This is using an image map. Still kinda looks like it is under the surface though.

CG.p
12-01-2003, 02:27 AM
Originally posted by SplineGod

If you have a character with blue clothing and the client wants red lighting, accuracy would mean the clothing would look purple. The clients wants red lighting and the clothing to stay blue which is not accurate.

No it wouldn't. It would either be red or black, it is blue because it reflects the blue parts of the, normally, white light. Try it. That would be one of those things a lot of people would pick up on in the back of their head. Just like glowing objects vs. translucent ones. Glowing may work..may not if you turn the camera.

All stuff that isn't too efficient if you need a full render just to see if the animation falls apart. :)

SplineGod
12-01-2003, 03:11 AM
Originally posted by CG.p
No it wouldn't. It would either be red or black, it is blue because it reflects the blue parts of the, normally, white light. Try it. That would be one of those things a lot of people would pick up on in the back of their head. Just like glowing objects vs. translucent ones. Glowing may work..may not if you turn the camera.

All stuff that isn't too efficient if you need a full render just to see if the animation falls apart. :)

Actually the reflected light would be the same color. The blue surface absorbs all the wavelengths and reemits the blue. Since most surfaces have some degree of reflectivity both blue and red will have some degree of mixing (aka purple). If you do try and and play with this you will see more or less purple depending on how much diffuse vs spec vs gloss is used (In LW).
In the real world a blue cloth can look purple because of a certain percentage of the red light will be reflected and a certain percentage be absorbed and remitted as blue. Again both will experience some degree of mixture resulting in some purple.
The only way this wouldnt happen is if the surface was perfectly reflective ( a mirror) , completely non reflective or the light was completly monochromatic (which typically wouldnt occur in nature).
All light contains some additional wavelengths. :)
Regardless, typically the more accurate things are the longer it takes to see the results. As I said, much of the time it requires a long period to wait to see what is most likely not going to be the desired result. Ive seen clients like the effect that SSS or GI gives but only in terms of a final look and are generally clueless or unconcerned about how that final look is achieved. :)

milqman
12-01-2003, 03:19 AM
Odd... I have my bump channel cranked to 500% and the bump ain't showing....

milqman
12-01-2003, 03:35 AM
I can't get my bump map to show up in the render.

Is there anything I could be forgetting?

SplineGod
12-01-2003, 03:35 AM
What are you applying the bumps to? Procedurals or image maps?
Do you have shadows turned on? If your diffuse or luminosity values are very high that can also effect what you see when rendering.

milqman
12-01-2003, 03:37 AM
procedurals

milqman
12-01-2003, 03:49 AM
Luminosity is at 0, Diffuse is at 50, doesnt Translucency at 30...

doesnt seem to matter what bump is set at.. it doesnt show

shadows are on

SplineGod
12-01-2003, 03:55 AM
Could you post it up somewhere?

milqman
12-01-2003, 03:55 AM
A render, or the scene files?

Edit: i pm'ed you the scene files

SplineGod
12-01-2003, 05:27 AM
Ok heres what I can see...
Your bump value in the main surface editor is set pretty high.
Set it back to 100%. Open up the bump texture panel.
All of your texture values for each procedural are set to 0%. They need to be set to something higher. Try 100% to start.
Your Layer opacity for each layer wont have any effect because of the fact that none of the procedurals have anything above 0%.
That fixed it. Also, your PM is full. :)

milqman
12-01-2003, 05:41 AM
Ahhh... stupid me

forgot to look at something so obvious

thanks larry

SplineGod
12-01-2003, 05:53 AM
Originally posted by milqman
Ahhh... stupid me

forgot to look at something so obvious

thanks larry
Everyones made that mistake. The beauty is that youll probably do it again. I STILL do. ;)

comanche
12-01-2003, 09:12 AM
Originally posted by E_Moelzer
CIM
I think what comanche wanted to say was, that pixels that are affected by SSS are brighter than without. Which is why we use SSS after all.

Exactly. Thanks Elmar!

Sorry for my poor english.

Cheers

Werner
12-02-2003, 08:54 AM
I think the "ska" plugin is awesome, and once you understand how the controls work, it can created awesome effects in expectable render times. Like other tools in any 3d application, we tend to set the quality settings very high, but are then shocked by render times. Try to keep the settings to low and work your way up from there...good render quality with acceptable rendering times.

The tutorial on the NT site will help in understanding the pluging better.

Here is something I rendered in under 12 minutes

http://www.pic-vault.com/ziem/PIC002.jpg

CGTalk Moderation
01-16-2006, 07:00 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.