As promised yesterday, I would tell some more about "gravity controlled" displacementmaps. One of the most difficult things in 3D-graphics is how to simulate soft objects whose shape is deformed by gravity, I'm thinking particulair about jello pudding, drops of water and very on topic, human skin. I've done some playing in the material room and I found some ways to simulate gravity with displacement I like to share: The sample picture here is made with two similar spheres with the same dimensions and dropped on the floor: the red sphere has only a diffuse red color , the green one has displacement with some nodes attached.

Ok more to come, I'll hope I can finish this before I fall asleep behind my keyboard.

regards,

Bopperthijs.

-How can you improve things when you don't make mistakes?

Prologue : The setup: two spheres, same dimensions and dropped on the floor only the colors are different, the green sphere is the guineapig and the red one is for reference, I will only change the materials, not the size or the position

-How can you improve things when you don't make mistakes?

First act: The displacement. I attached a simplecolor node with color white, to the displacement channel and set the displacement to 20, nothing fancy yet. Ignore the other nodes they will be used later.

-How can you improve things when you don't make mistakes?

...and the result as could be expected.

-How can you improve things when you don't make mistakes?

second act: Adding gravity: to simulate the gravity I added a N-variable node and multiplied that with color node by a colormath node. The N-variable node defines the direction of the normal vector of a point on the sphere, I only need the Y-value which points in the opposite direction of the ground, so the x and the z value are set to zero and the y-value to -3. So if a normal vector points up  (in case of the upper side of the sphere) the displacement  will be negative and if he points downwards, the displacement will be positive.

-How can you improve things when you don't make mistakes?

...and this the result: not very spectaculair, it looks like there isn't much of a change  but the shape of the sphere has slightly changed and it looks like the sphere has dropped. displacement scale is 20

-How can you improve things when you don't make mistakes?

...displacementscale at 50, sphere is a little lower and the shape is more egglike...

-How can you improve things when you don't make mistakes?

... this the sphere with a displacement scale of 100, something is happening now: the sphere is getting bigger and it looks like the top is flattening, what is going on? take a look at the next render...

-How can you improve things when you don't make mistakes?

.. this is with a displacement scale of 200, I made the sphere transparant to show what is happening, This behaviour shows up when the displacementscale is bigger than the half of the sphere radius, from a mathematical POV this makes sense, Newton invented gravity by use of an apple, we invented the apple by use of gravity. Ok, but this not what I wanted, the solution is in the next post...

-How can you improve things when you don't make mistakes?

.. This is what I did: I added the original color to the result of the multiply with the N-node,  in this way the thickness of the layer added by the displacement will never be smaller than 0 ...

-How can you improve things when you don't make mistakes?

...and this is the result, as you see the sphere is much bigger now, but that's actually what I wanted. The only problem is that the sphere is penetrating the ground and I don't like that... Solution is following...

-How can you improve things when you don't make mistakes?

Third act: touching the ground. To make the sphere look like he's pressed on the ground I've added some more nodes: first a P-node of which I zeroed the x and the z values and set y-value to 0.1, then I added a clamp node , a sine node and finally multiplied that with the orginal displacement node. The P-node and the clamp-node defines a region in which the displacement is faded out to zero,  the sinus node make it more fluent. You can raise or lower this region by changing the y-value in the P-node...

-How can you improve things when you don't make mistakes?

,,, and the final result, looks a lot like the first picture. You can lower the blob by making value 1 of the color-math 2 a litlle darker, changing the displacement  will have some influence on the bottom shape, you can change that by changing the y -value of the P-node. Instead of the simple_color node also image_maps can be used offcourse. I haven't had the intension to make a scientific approach to  the problem, but more to make a bag of tricks to make it look like gravity is involved. Wanna see some usefull application? Check in tomorow for an explication of fat-distribution maps. But beware!: nudity will be involved

-How can you improve things when you don't make mistakes?

this is good stuff     thankyou

Very interesting and useful stuff.  Thanks for this.

Coppula eam se non posit acceptera jocularum.

My Store

My Gallery

Wow, that's fantastic!  bookmark

Hi, my namez: "NO, Bad Kitteh, NO!"  Whaz yurs?
BadKittehCo Store  BadKittehCo Freebies and product support

bookmark

PoserPro 2014, Windows 7, AMD FX-6300 6 core, 8 GB ram, Nvidia GeForce GTX 750 Ti

Cool  - bookmarked

The important thing is not to stop questioning.
- Albert Einstein

Indoda

Cool. BM.

Hello again,

Last night I promised to tell something about fat-distribution maps, and so here its is. I wanted to start a new tnread but because some of you have bookmarked this one, I'm continuing here.

Ok, what do I mean with a fat-distribution map? The human body has several places where it stores its bodyfat, this is sexe-dependent. A grown-up male mainly stores his fat in his belly, an adult female stores her mainly in the hips, the buttocks and the thighs. This a very natural  phenonemon, but nowadays most western people don't like it when those fat layers are getting too thick. 
I've done some experiments with displacement bodymaps in which I painted in those places where bodyfat appears: this is what I call fat-distributionmaps. One of the properties of bodyfat is, that it has no structure, It tends to hang like water in a balloon, so it's an excellent candidate for my "gravity" node.
I want to demonstrate this by means of that human bodypart that's obvious the most influenced by  gravity : The male beerbelly.

In the following post comes the samples.

regards,

Bopperthijs

-How can you improve things when you don't make mistakes?

Here are my settings: I've used Michael3 as Guineapig, I applied this materialsetting to the torso, the hip and the nipples materialzones of Michael, The other zones are no affected. In the upperright corner, you see the map I made, the size is about 3000x3999 and greyscaled. I've made this map in paintshop: First I imported the seam-guide of M3 by snowsultan and scaled it up. Then I added a layer as background which I filled with the deepest black. Then I made a next layer whcih I called belly and I painted some grey and white strokes in the bellyarea, and then I blurred it with a gaussian blur with radius 50. If you don't blur it you'll get strange kinks in your displacement. I saved the map as BMP-file, you can also save it as jpg, but be sure to keep the quality at 100%. The image-map node is attached to the "gravity" nodes which I showed in the first posts. I've omitted the "touch-floor", because they won't be necessary in this case, Displacementscale is set to 10, I found out that higher values then 15 make the result ugly. There are more drawbacks but I;ll tell them in the last post. After this comes the first render without displacement; I injected a "heavy" morph to M3 and dialed it up to 0.2, I'll explain later why.

Bopper.

-How can you improve things when you don't make mistakes?

here's the first sample render without displacement.

-How can you improve things when you don't make mistakes?

And here the same render with displacement. The effect is not very strong, but enough to let it look more saggy than the other.

-How can you improve things when you don't make mistakes?

It becomes more clear when he bends a little over, Here's the without....

-How can you improve things when you don't make mistakes?

...and here with displacement.

-How can you improve things when you don't make mistakes?

Here is from another angle, without...

-How can you improve things when you don't make mistakes?

... and the same with displacement.

-How can you improve things when you don't make mistakes?

Like I said there are some disadvantages to this method:
It's hard to smooth, because the N-node take the original mesh and not the smoothed one, you get strange kinks in your surface. Keep the displacementscale small to avoid this.
When you have detailed objects in your "fat-area" like a navel or nipple, the surface gets very distorted by the displacement, you can avoid this by using a nipple-gone or navel-gone morph, but this gives strange results in the texture map. It works better in area's with little detail.
It works great on breasts but you get huge areolas and nipples
The fat-d.map is useful in combination with fatrelated FBM-morphs like heavy, voloptuous or similar, It helps to exaggarate the morph and gives it a more natural look, this is why I used a heavy morph on Michael.

Well, this is it so far. I'm well aware this can be improved, but I'm still working on it, the hard part is to make realistic mape and finding the correct settings. If anyone has any ideas for this I'm open for all  suggestions..

Have a nice evening,

Bopperthijs

-How can you improve things when you don't make mistakes?

Excellent work bopperthijs, BKMK

Getting old and still making "art" without soiling myself, now that's success.

Wow! This has all sorts of implications! And our bodies distribute fat all over, unless you are exceptionally fit- so it could be applied in lower quantities to arms, neck chin, etc...

Thanks for the walkthrough. Looking forward to seeing more as you dig deeper into this concept.

 I've been trying this with antother parts of the female body that store fat. (Am I the only one not thinking of beer-bellies but breasts ;-) ) Anyway, in general it works. I had to tweak the displacement value though. Was way too high for me.

Amazing. I never thought that the material room woulf be able to produce such results. There are a few issues though. Image quatilty, the displacement is controlled by 255 scales of grey. Sounds like a lot, but it is defenitly visible.

When animating earth's gravity isn't the only force pulling on those fatty tissues. so depending on the motion the x,y and z values for N would have to be computed for every frame.

I guess I can't show my results or link to them due to the TOS, right?

globo

I think if you put the nudity flag on, it would be OK

Amazing. I never thought that the material room woulf be able to produce such results. There are a few issues though. Image quatilty, the displacement is controlled by 255 scales of grey. Sounds like a lot, but it is defenitly visible.

Hi Globo,

That was my problem  also. I've tried to use a 16bit  greyscale tga-file, but somehow poser converts that back to a 8bit map, perhaps someone of the materialroom guru's like Bagginsbill can tell more about that or how to solve that.

The beer-bellie was a bit of a farce, of course I wanted to use it for breasts, but I didn't wanted to show the obvious, it felt a little raunchy.
I wanted  to make the ultimate  breast-shader, but till now it's still not perfect, because there are issues: 
You can't use nipple or areola morphs, because with a displacement map over it, you get very big and ugly nipples. So I made another displacement map for that too.
The displacement on the edges of the areola-group is a little crumpled and I don't know how I can fixed. I believe there are some nodes to smooth between surface edges , but I haven't figured out yet how to use them.
You can see the edges of the faces by the displacement. From a distance it's not very clear but not very suited for close-ups. I think it's the same problem as the last issue.
I see if I can dig up some renders to show my results, it's been a while since I worked on this.

Bopper.

-How can you improve things when you don't make mistakes?

You can use a 16 bits or even 24 bits image map, but it has to be a color image. For example, you can use red for the lower 8 bits of the displacement, and green for the higher 8 bits.
Two component nodes can extract the red channel and the green channel - multiply the extracted green channel by 256 and add it to the red channel, and voila - 16 bits grayscale. Add blue, and you can go to 24 bit.

The pen is mightier than the sword. But if you literally want to have some impact, use a typewriter

My gallery   My freestuff

 very interesting!!!
Love esther

MY ONLINE COMIC IS NOW LIVE

I aim to update it about once a month.  Oh, and it's free!