[Fisio SB addOn released] Fisio: Physics Simulator

Alberto

mikethe3dguy said:

Alberto,

For that Fisio (rigid body) setup with the shattering glasses, I got a chance to try it in my actual scene file using the full 11 glasses instead of the one I was experimenting with. So in this case I'm not using just primitives anymore but mostly regular Daz objects: Genesis 8 figure and props, etc. No Gescon-fragmented glasses involved yet. All I'm trying to do now is just simulate the glasses themselves getting knocked off the bar by the object that hits them (the butt of a Genesis 8 male). I expected this part to go smoothly but was surprised at the behavior. This is just at precision level 1 and all pane settings at default at this point.

The problems:

1. The glasses started falling when the G8M was still several inches away from the glasses, and in nearly random directions, so: well before they were struck by anything.
2. A couple glasses and one other object flew TOWARDS the G8M figure and past it, moving in the direction opposite to the direction they would have moved had they been struck by  the G8M
3. Several of the glasses sank into the bar that they were sitting on, dropping through it.

So: random weirdness. I first ran the simulation with all items in my scene visible, then after I got strange results, tried again after hiding all non-essential items (nodes), but got the same result.

What is going on?

The G8M figure is using the Convex Hull collision shape. The bar is set to Box (though it's not a simple rectangular shape) and all the glasses are Cylinder. No force nodes, no link nodes, nothing animated except the G8M (or cube).

Edit: Tried substituting the G8M figure with a primitive cube set to Box collision type, but got identical results. Then I hid the cube in the scene, so theoretically there are no object collisions occuring. I ran the simulation, and after 3-4 frames the glasses jumped up off the bar and my one non-glass object sank through the bar. All simulation objects are "assigned" and all objects assigned have material nodes.

For the figure, use Mesh as the collision shape, because the Convex hull is only an approximation.

By hiding the nodes, you mean disabling the bodies, don't you?

Did the glasses suddenly jump without a body pushing them? As the bar is not a cubic or cuboid shape, it could be that the glasses aren't exactly on top of the cuboid that internally represents the bar. Try to use a cube primitive as a proxy of the bar in the simulation, it will help you to set the glasses. You can hide or eliminate it after the simulation.

Reduce the density of the material of the bar, it is much bigger than the glasses (I think) and the extreme sizes can cause some trouble to the simulation.

Try another precision level, 1 is the simpler one.  

mikethe3dguy

Alberto said:

For the figure, use Mesh as the collision shape, because the Convex hull is only an approximation.

For right now I'm back to experimenting with a cube again, but I'll try that once I can get things to work with the cube.

By hiding the nodes, you mean disabling the bodies, don't you?

I didn't - I just hid them in the scene. But now whenever I try that I'll make sure to disable them.

Did the glasses suddenly jump without a body pushing them? As the bar is not a cubic or cuboid shape, it could be that the glasses aren't exactly on top of the cuboid that internally represents the bar. Try to use a cube primitive as a proxy of the bar in the simulation, it will help you to set the glasses. You can hide or eliminate it after the simulation.

Yes they did. There are three levels of glasses. The first level is set (using DS "Align") to stack exactly on top of the bar. The second level of glasses stack exactly on top of the first, and the third level exactly on top of the second. I'll try a cube primitive then.

Reduce the density of the material of the bar, it is much bigger than the glasses (I think) and the extreme sizes can cause some trouble to the simulation. Try another precision level, 1 is the simpler one.

Glasses are default 1000kg/m3, as was everything else. I just ran a sim after setting the bar and cube to 100kg/m3 and leaving the glasses at 1000, using precision level 3. Starting sim at frame 0, at frame 1 all the glasses and my one other object on the bar sank into the bar about a cm or less. When the cube approached but was still 2 inches away, the objects started to move away from it as if the cube was already touching them. At least they didn't fly around randomly though.

I'll try again with a cube primitive instead of the bar. I hope I don't have to substitute everything with primitives - there's a bar stool the glasses hit on the way down to the floor, and it's a complex shape. Replacing it with a cube would not work.

mikethe3dguy

Alberto said:

Reduce the density of the material of the bar, it is much bigger than the glasses (I think) and the extreme sizes can cause some trouble to the simulation.

Try another precision level, 1 is the simpler one.  

I'm now using only cubes for the "bar" and the object striking the glasses. Things are improved, but there are still issues I hope can be resolved:

1. On the 2nd frame and after, the glasses sink into the "bar" at least a cm or two.
2. When the cube strikes them, the glasses go inside the moving cube up to half of their depth, as well as the floor, which is a primitive cube as well.

kwannie

Alberto said:

kwannie said:

I am having issues getting Fisio Physics Simulator to work, below are all the symtoms I get, What could be causing this? Thank you!

After clicking Run simulation
 2 dialogue box's appear that state

1st box:

DS DAZ Studio
Simulation Running
[progress bar that shows the percentage complete]
: however percentage never advances past 1%

2nd box:
DS Error
         The specified module could not be found
                      [ok]

After clicking ok on second box, another dialogue box appears stating

DS Error
Can't load FisioEngine.dll at E:My DAZ 3D Library/Applications/64-bit/DAZ
3D/DAZStudio4/libs/Fisio/FisioEngine.dll
                                 [ok]

After clicking ok on this box, I get a final dialogue box stating

DS Error
   Can't load Fisio engine
            [ok]

I do have a Fisio Pane that shows up in DS and I made sure that Serial number entered in
About installed Plugins

Were you trying to run the Fisio main plugin?

Check the installation folder of DAZ Studio (usually "/Program Files/DAZ 3D/DAZStudio4"). You should the subfolder "/libs/Fisio". Inside this, verify if there are the following files: FisioEngine.dll, ChronoEngine_multicore.dll, and ChronoEngine.dll. If not, try to reinstall the plugin. If you have more trouble, let me know, please.

 

Alberto

   I got a reply from another forum member that asked if I login to DS beccause that might be the reason I am having the issue. So, I told themI do not log into DAZ from within DS. I also recently moved the entire "My Library that I had on one hard drive to another hard drive by just copying the entire "My Library" folder. I do in fact get the following message

DS Configuration Error

Configuration of the system has changed. To use content installed using Daz Connect, the
 content owner must re-authenticate
                                   [ok] 

DS was still working as before so I foolishly disregarded the message. Do you know what I should do now to re-athenticate? Is DAZ Connect just logging in from within DS? 

All of the files you asked about being libs/Fisio are in fact located there. If the issue is that I copied "My Library" to another hard drive what do I need to do to re-authenticate?

Do I need to re-install anything, or do I just nee to login inside of DS? I did install Fisio Physics after I copied the "My Library"  to the new drive.

I have the Fisio pane and was able to apply all of the parameters to the sphere and box as in the tutorial. I only go the error dialogue box after clicking [Run simulation]

 

Alberto

I'll try again with a cube primitive instead of the bar. I hope I don't have to substitute everything with primitives - there's a bar stool the glasses hit on the way down to the floor, and it's a complex shape. Replacing it with a cube would not work.

Use the Mesh collision shape for the glasses and the stool.

The precision 4 many times is the better of the four levels. and it's not necessarily slow. Precision 2 is always slower than Precision 1 because they use the same method, but Precision 3 and 4 could be faster than 2 many times. 

Alberto

mikethe3dguy said:

I'm now using only cubes for the "bar" and the object striking the glasses. Things are improved, but there are still issues I hope can be resolved:

1. On the 2nd frame and after, the glasses sink into the "bar" at least a cm or two.
2. When the cube strikes them, the glasses go inside the moving cube up to half of their depth, as well as the floor, which is a primitive cube as well.

I created a simple scene with glasses and a box as a bar. It's attached to his message. Only 11 glasses in two levels. The friction of the material of the glasses is set to 0.4. They don't go inside the box. Check the scene, maybe it could help you.

I used instances of the glass, but it's the same to use regular glass nodes. I modeled the glass, so use it freely.

Edit: I updated the .duf, so the glass obj is embedded.

mikethe3dguy

Alberto said:

I'll try again with a cube primitive instead of the bar. I hope I don't have to substitute everything with primitives - there's a bar stool the glasses hit on the way down to the floor, and it's a complex shape. Replacing it with a cube would not work.

Use the Mesh collision shape for the glasses and the stool.

The precision 4 many times is the better of the four levels. and it's not necessarily slow. Precision 2 is always slower than Precision 1 because they use the same method, but Precision 3 and 4 could be faster than 2 many times. 

Okay, did that. Glasses still sink into the bar cube and many still react well before being struck. 14 frames in and the glasses are flying off into space basically like they were shot out of a cannon. i didn't change anything but what you suggested: Precision 4 and glasses and stool are now mesh.

Alberto

kwannie said:

Alberto

   I got a reply from another forum member that asked if I login to DS beccause that might be the reason I am having the issue. So, I told themI do not log into DAZ from within DS. I also recently moved the entire "My Library that I had on one hard drive to another hard drive by just copying the entire "My Library" folder. I do in fact get the following message

DS Configuration Error

Configuration of the system has changed. To use content installed using Daz Connect, the
 content owner must re-authenticate
                                   [ok] 

DS was still working as before so I foolishly disregarded the message. Do you know what I should do now to re-athenticate? Is DAZ Connect just logging in from within DS? 

All of the files you asked about being libs/Fisio are in fact located there. If the issue is that I copied "My Library" to another hard drive what do I need to do to re-authenticate?

Do I need to re-install anything, or do I just nee to login inside of DS? I did install Fisio Physics after I copied the "My Library"  to the new drive.

I have the Fisio pane and was able to apply all of the parameters to the sphere and box as in the tutorial. I only go the error dialogue box after clicking [Run simulation]

The plugin has nothing to do with the Library folder nor with DAZ Connect, so I don't think that was causing the issue.

Are you running Daz Studio from the folder you examined? The attached script will show the location of your installation. Inside this folder should be located the subfolder "libs/Fisio" and the DLL files FisioEngine.dll, etc.

If all the previous it's OK, I suggest using the CFF explorer to check if the dependencies of FisioEngine.dll are complete. Here you can check how to do it: https://www.daz3d.com/forums/discussion/comment/7873596/#Comment_7873596

mikethe3dguy

Alberto said:

I'll try again with a cube primitive instead of the bar. I hope I don't have to substitute everything with primitives - there's a bar stool the glasses hit on the way down to the floor, and it's a complex shape. Replacing it with a cube would not work.

Use the Mesh collision shape for the glasses and the stool.

The precision 4 many times is the better of the four levels. and it's not necessarily slow. Precision 2 is always slower than Precision 1 because they use the same method, but Precision 3 and 4 could be faster than 2 many times. 

Okay, I'm making progress. Two questions:

1. What could be causing nodes to sink into other nodes they're sitting on? Does a disparity in mesh density affect this? The primitives I'm using are cubes and are just one poly per side, and the wine glasses and jigger (my one object to get knocked over that is not a glass) are much higher dentity. Should I increase the polygon count of the cubes? (Edit: my subsequent experiments indicate that this is NOT the case)
2. Is there any way to reduce the energy imparted by my animated, moving cube? it's strange, but some glasses basically fall off the edge of the bar and some end up flying 6 meters away which is not the behavior I'd expect. I can't change the speed of the animation - that's baked in to my scenario, but the object hitting them is soft and spongy (a human body), so I'd like to somehow compensate for that so the simulation is closer to the reality of my scene.

kwannie

Thanks Alberto I did a reinstall of DS and resolve the issue. Thank you so much

mikethe3dguy

Alberto said:

mikethe3dguy said:

I'm now using only cubes for the "bar" and the object striking the glasses. Things are improved, but there are still issues I hope can be resolved:

1. On the 2nd frame and after, the glasses sink into the "bar" at least a cm or two.
2. When the cube strikes them, the glasses go inside the moving cube up to half of their depth, as well as the floor, which is a primitive cube as well.

I created a simple scene with glasses and a box as a bar. It's attached to his message. Only 11 glasses in two levels. The friction of the material of the glasses is set to 0.4. They don't go inside the box. Check the scene, maybe it could help you.

I used instances of the glass, but it's the same to use regular glass nodes. I modeled the glass, so use it freely.

Edit: I updated the .duf, so the glass obj is embedded.

Oh! Somehow I missed this post! Thank you, I'll give it a try! But I think I've discovered the trouble. Sometimes I have so many renders to get through in a day that I'll end up leaving my PC on overnight, even for up to 3-4 days in a row. Last night I shut it down after such a period. This morning the simulations are acting very differently, and much better! No sinking into box platforms, and no erratic flyaway behavior with the glasses! Got to at least restart it every 24 hours!

mikethe3dguy

Ugh! I was on the cusp of dialing in my Fisio simulations for an animation and I went and did a stupid thing: I updated my NVIDIA GPU driver! Now, any sim that's somewhat complex freezes my PC!

I'd like some help if anyone can check and chime in: I want to roll the driver back, but I don't know what version I HAD, or even better: the most recent version that will work stably with Fisio. The version I updated to that caused the problems is 536.67 (Studio). By "somewhat complex" I mean I have several objects in it that use the "Mesh" Collision Shape Type, including one Genesis 8 figure. Be nice to know what version you're using that works well.

Hoping someone can provide some info. Otherwise I'm just going to have to guess.

 

UPDATE: First tried version 531.41 (released 3/21/2023) and that was no better. Then dropped back to 527.56 (released 12/8/2022) and that seems to be working fine. Hopefully this helps someone else having similar problems.

Imago

Here are few examples of what FisioSB can do. They are just few, the possibilites are infinte, my word!

bashton

Alberto, I'm trying to do some very basic things to learn how FisioSB works and quite frankly I'm stumped. I have two rubber balls, one above the other, above a flattened-out cube. When I run the simulation the lower ball should bounce off of the cube and then collide with the other ball. They should bounce off of each other in some random direction. What actually happens in the simulation is the lower ball gets absorbed into the upper ball and the two balls, now merged into each other, bounce up and down in the same place.

So, two soft body objects; how do I make them collide with each other and bounce off of each other the way they would in the real world? What settings have I got wrong here?

Alberto

bashton said:

Alberto, I'm trying to do some very basic things to learn how FisioSB works and quite frankly I'm stumped. I have two rubber balls, one above the other, above a flattened-out cube. When I run the simulation the lower ball should bounce off of the cube and then collide with the other ball. They should bounce off of each other in some random direction. What actually happens in the simulation is the lower ball gets absorbed into the upper ball and the two balls, now merged into each other, bounce up and down in the same place.

So, two soft body objects; how do I make them collide with each other and bounce off of each other the way they would in the real world? What settings have I got wrong here?

Two soft bodies, when collide, tend to lose their kinetic energy and barely bounce. This behavior is more of the Rigid bodies.

If you want some bouncing, you have to use a high Young modulus, e.g.100 GPa, and high density (5000 kg/m3, for example). One of the two bodies has to have enabled the Full Mesh for contact property.

Here is a .duf of an example scene, if you want to check settings:

https://drive.google.com/file/d/1kIGU-6n92ihFfkC9W-n6jeIPktNWigmV/view?usp=sharing

bashton

Alberto said:

bashton said:

Alberto, I'm trying to do some very basic things to learn how FisioSB works and quite frankly I'm stumped. I have two rubber balls, one above the other, above a flattened-out cube. When I run the simulation the lower ball should bounce off of the cube and then collide with the other ball. They should bounce off of each other in some random direction. What actually happens in the simulation is the lower ball gets absorbed into the upper ball and the two balls, now merged into each other, bounce up and down in the same place.

So, two soft body objects; how do I make them collide with each other and bounce off of each other the way they would in the real world? What settings have I got wrong here?

Two soft bodies, when collide, tend to lose their kinetic energy and barely bounce. This behavior is more of the Rigid bodies.

If you want some bouncing, you have to use a high Young modulus, e.g.100 GPa, and high density (5000 kg/m3, for example). One of the two bodies has to have enabled the Full Mesh for contact property.

Here is a .duf of an example scene, if you want to check settings:

https://drive.google.com/file/d/1kIGU-6n92ihFfkC9W-n6jeIPktNWigmV/view?usp=sharing

Yes, that's for two hard balls, like hard rubber. That's easy, and you're correct that they lose their inertia. Kinetic energy is not the issue. I want them to be soft, more like balloons filled with water. I want them both to deform a lot. However, when I lower the Young Modulus setting to where they start to deform the way I want them to one ball gets absorbed by the other and bounces around inside of it. If I had two balloons in the real world, one above the other, and dropped them from a height of, say, 12 and 18 inches, and the top one was larger than the other, the smaller balloon would not get absorbed into the larger one. Well, that's what happens in the simulation. One ball (or balloon) on any setting that is really soft, gets absorbed by the other and just bounces around inside of it.

So with a water balloon, we have a shell, yet still very pliable and stretchy, around a dense liquid that causes it to really and obviously deform. That's the kind of effect I'm trying to get.

Take the same file that you uploaded. (Thank you, by the way.) Make the top ball larger. Move it up just a little bit. Make them so they will very obviously deform when they collide with the cube or each other. You'll see the problem I'm having.

Or change out the balls and water balloons for something like a handful of Gummy Worms (the candy), or fake rubber worms like you'd use when fishing, dropped on a table. Definitely a soft body, rather than a rigid body simulation. So make it ten long thin cylinders. They would bend around each other and bounce around a bit and they would definitely deform a lot. But at no time would they (in the real world) pass through each other.

How would I do Something like that?

(Addition: I just tried three long thin cylinders. The same settings as used in your file except that I lowered the Young Modulus setting to .5. I even made different material nodes for each of the cylinders. What you would expect in the real world is that they would bounce around against off of the cube and each other and eventually settle on the cube. What happened in the simulation is that the three cylinders sucked into each other as soon as they collided and bounced on the cube as if one object. I also, just for the fun of it, ran another simulation with a sphere dropping on top of the cylinders. The cylinders sucked up into each other and then got sucked up into the sphere...just before the mesh exploded. Maybe this is a warning never to drop a rubber ball on top of Gummy Worms in the real world. Lol... Anyway, It's not what I expected.)

Alberto

bashton said:

Yes, that's for two hard balls, like hard rubber. That's easy, and you're correct that they lose their inertia. Kinetic energy is not the issue. I want them to be soft, more like balloons filled with water. I want them both to deform a lot. However, when I lower the Young Modulus setting to where they start to deform the way I want them to one ball gets absorbed by the other and bounces around inside of it. If I had two balloons in the real world, one above the other, and dropped them from a height of, say, 12 and 18 inches, and the top one was larger than the other, the smaller balloon would not get absorbed into the larger one. Well, that's what happens in the simulation. One ball (or balloon) on any setting that is really soft, gets absorbed by the other and just bounces around inside of it.

So with a water balloon, we have a shell, yet still very pliable and stretchy, around a dense liquid that causes it to really and obviously deform. That's the kind of effect I'm trying to get.

Take the same file that you uploaded. (Thank you, by the way.) Make the top ball larger. Move it up just a little bit. Make them so they will very obviously deform when they collide with the cube or each other. You'll see the problem I'm having.

Or change out the balls and water balloons for something like a handful of Gummy Worms (the candy), or fake rubber worms like you'd use when fishing, dropped on a table. Definitely a soft body, rather than a rigid body simulation. So make it ten long thin cylinders. They would bend around each other and bounce around a bit and they would definitely deform a lot. But at no time would they (in the real world) pass through each other.

How would I do Something like that?

(Addition: I just tried three long thin cylinders. The same settings as used in your file except that I lowered the Young Modulus setting to .5. I even made different material nodes for each of the cylinders. What you would expect in the real world is that they would bounce around against off of the cube and each other and eventually settle on the cube. What happened in the simulation is that the three cylinders sucked into each other as soon as they collided and bounced on the cube as if one object. I also, just for the fun of it, ran another simulation with a sphere dropping on top of the cylinders. The cylinders sucked up into each other and then got sucked up into the sphere...just before the mesh exploded. Maybe this is a warning never to drop a rubber ball on top of Gummy Worms in the real world. Lol... Anyway, It's not what I expected.)

Check this scene:

https://drive.google.com/file/d/1LQxm9cyN4tqf6ciOXFBF5NpNeUnEoe68/view?usp=sharing 

The trick here was to set the Autocollision margin of one sphere to Off, and its Collision margin to 5. And for the other sphere, the Full Mesh for contact property to On.

bashton

Alberto said:

bashton said:

Yes, that's for two hard balls, like hard rubber. That's easy, and you're correct that they lose their inertia. Kinetic energy is not the issue. I want them to be soft, more like balloons filled with water. I want them both to deform a lot. However, when I lower the Young Modulus setting to where they start to deform the way I want them to one ball gets absorbed by the other and bounces around inside of it. If I had two balloons in the real world, one above the other, and dropped them from a height of, say, 12 and 18 inches, and the top one was larger than the other, the smaller balloon would not get absorbed into the larger one. Well, that's what happens in the simulation. One ball (or balloon) on any setting that is really soft, gets absorbed by the other and just bounces around inside of it.

So with a water balloon, we have a shell, yet still very pliable and stretchy, around a dense liquid that causes it to really and obviously deform. That's the kind of effect I'm trying to get.

Take the same file that you uploaded. (Thank you, by the way.) Make the top ball larger. Move it up just a little bit. Make them so they will very obviously deform when they collide with the cube or each other. You'll see the problem I'm having.

Or change out the balls and water balloons for something like a handful of Gummy Worms (the candy), or fake rubber worms like you'd use when fishing, dropped on a table. Definitely a soft body, rather than a rigid body simulation. So make it ten long thin cylinders. They would bend around each other and bounce around a bit and they would definitely deform a lot. But at no time would they (in the real world) pass through each other.

How would I do Something like that?

(Addition: I just tried three long thin cylinders. The same settings as used in your file except that I lowered the Young Modulus setting to .5. I even made different material nodes for each of the cylinders. What you would expect in the real world is that they would bounce around against off of the cube and each other and eventually settle on the cube. What happened in the simulation is that the three cylinders sucked into each other as soon as they collided and bounced on the cube as if one object. I also, just for the fun of it, ran another simulation with a sphere dropping on top of the cylinders. The cylinders sucked up into each other and then got sucked up into the sphere...just before the mesh exploded. Maybe this is a warning never to drop a rubber ball on top of Gummy Worms in the real world. Lol... Anyway, It's not what I expected.)

Check this scene:

https://drive.google.com/file/d/1LQxm9cyN4tqf6ciOXFBF5NpNeUnEoe68/view?usp=sharing 

The trick here was to set the Autocollision margin of one sphere to Off, and its Collision margin to 5. And for the other sphere, the Full Mesh for contact property to On.

Okay, thanks. I'll check it out the next time I work on this but off the top of my head, without actually running it, my next question is likely to be; "That's fine for two spheres but what would the settings be for ten or fifteen?" Or a handful of something like rubber bands? I haven't tried yet but that's the questions in my head at the moment. I think as a user a few tutorials with real-world simulations like this, or for example how to use this addon for figures like characters. At least demonstrating the basic things that users would commonly attempt.

bashton

bashton said:

Alberto said:

bashton said:

Yes, that's for two hard balls, like hard rubber. That's easy, and you're correct that they lose their inertia. Kinetic energy is not the issue. I want them to be soft, more like balloons filled with water. I want them both to deform a lot. However, when I lower the Young Modulus setting to where they start to deform the way I want them to one ball gets absorbed by the other and bounces around inside of it. If I had two balloons in the real world, one above the other, and dropped them from a height of, say, 12 and 18 inches, and the top one was larger than the other, the smaller balloon would not get absorbed into the larger one. Well, that's what happens in the simulation. One ball (or balloon) on any setting that is really soft, gets absorbed by the other and just bounces around inside of it.

So with a water balloon, we have a shell, yet still very pliable and stretchy, around a dense liquid that causes it to really and obviously deform. That's the kind of effect I'm trying to get.

Take the same file that you uploaded. (Thank you, by the way.) Make the top ball larger. Move it up just a little bit. Make them so they will very obviously deform when they collide with the cube or each other. You'll see the problem I'm having.

Or change out the balls and water balloons for something like a handful of Gummy Worms (the candy), or fake rubber worms like you'd use when fishing, dropped on a table. Definitely a soft body, rather than a rigid body simulation. So make it ten long thin cylinders. They would bend around each other and bounce around a bit and they would definitely deform a lot. But at no time would they (in the real world) pass through each other.

How would I do Something like that?

(Addition: I just tried three long thin cylinders. The same settings as used in your file except that I lowered the Young Modulus setting to .5. I even made different material nodes for each of the cylinders. What you would expect in the real world is that they would bounce around against off of the cube and each other and eventually settle on the cube. What happened in the simulation is that the three cylinders sucked into each other as soon as they collided and bounced on the cube as if one object. I also, just for the fun of it, ran another simulation with a sphere dropping on top of the cylinders. The cylinders sucked up into each other and then got sucked up into the sphere...just before the mesh exploded. Maybe this is a warning never to drop a rubber ball on top of Gummy Worms in the real world. Lol... Anyway, It's not what I expected.)

Check this scene:

https://drive.google.com/file/d/1LQxm9cyN4tqf6ciOXFBF5NpNeUnEoe68/view?usp=sharing 

The trick here was to set the Autocollision margin of one sphere to Off, and its Collision margin to 5. And for the other sphere, the Full Mesh for contact property to On.

Okay, thanks. I'll check it out the next time I work on this but off the top of my head, without actually running it, my next question is likely to be; "That's fine for two spheres but what would the settings be for ten or fifteen?" Or a handful of something like rubber bands? I haven't tried yet but that's the questions in my head at the moment. I think as a user a few tutorials with real-world simulations like this, or for example how to use this addon for figures like characters. At least demonstrating the basic things that users would commonly attempt.

Alright. I did try it. It did work and it is exactly the kind of effect I was looking for. As I predicted there seems to be no way to do this with more than two objects in the scene. It seems kind of strange that we would have a soft body simulator and so much difficulty bouncing soft bodies off of each other in a simulation without them passing through each other, getting absorbed into one another or exploding. It's the first thing I thought of to do with this is to bounce a bunch of soft body objects off of each other and have them deform when they collide.

mikethe3dguy

mikethe3dguy said:

Ugh! I was on the cusp of dialing in my Fisio simulations for an animation and I went and did a stupid thing: I updated my NVIDIA (RTX 3090) GPU driver! Now, any sim that's somewhat complex freezes my PC!

I'd like some help if anyone can check and chime in: I want to roll the driver back, but I don't know what version I HAD, or even better: the most recent version that will work stably with Fisio. The version I updated to that caused the problems is 536.67 (Studio). By "somewhat complex" I mean I have several objects in it that use the "Mesh" Collision Shape Type, including one Genesis 8 figure. Be nice to know what version you're using that works well.

Hoping someone can provide some info. Otherwise I'm just going to have to guess.

 

UPDATE: First tried version 531.41 (released 3/21/2023) and that was no better. Then dropped back to 527.56 (released 12/8/2022) and that seems to be working fine. Hopefully this helps someone else having similar problems.

mazinkaiserzero

Alberto said:

mazinkaiserzero said:

Hellboy said:

Hi.

I'll be that person...

Will this work to properly animate breasts or male anatomical elements?

Thanks.

Came here to ask this, so I'll second it.

I already bought it and have been trying to do a simple animation with a G8F figure moving up and down, I'm using springs with the active facets only option to select only the mesh for the breasts. Sadly, they deflate completely. I tried using the skin young presets and same results. I did notice increasing density helps a bit in keeping the shape but then gravity affects them even worse. Also, the selected mesh now doesn't seem to follow the figure at all. Turning off the gravity stops the breasts from deflating but then the breasts just stayed in the same place while the figure moved up and down. Then I tried doing a collision with a rigid body and I just couldn't get any bouncing back effect, it almost just feels like a regular dforce sim.

So I'd like to ask if you could show how to do it like in the nose press promo picture and if this can actually be used for bouncy animations with figures (and how)

Thanks! 

Use Shell instead of Spring. There are some presets for the Fisio Materials (in Smart Content - Props - Fisio SB - Materials (they're not surface material for renders, but properties preset applicable to the Fisio Materials.

If you use Spring, use the Stiffness of the Spring property to control the stiffness.

You can't animate with keyframes the Soft bodies, but you can transform the simulation into morphs and then disable the soft body (or unsign it as a soft body) to use the morphs.

The soft bodies and the rigid bodies can collide, but when using the Non Smooth contacts System type, the objects tend to stick. What are the settings of the bodies and the system you used?

Hi, sorry for my very late reply. If I understood correctly, then sadly this addon doesn't work for what I wanted. Like it was mentioned, I was looking for a way to animate breasts and maybe other soft parts of the body, but if you can't make the softbody selections follow the rest of the genesis figure in an animation I don't see how else you can do it. Yeah, it works to simulate other types of objects or a squish in a character for a still render (like the nose render) but I don't see how it could work for character animation =/

Alberto

mazinkaiserzero said:

Alberto said:

mazinkaiserzero said:

Hellboy said:

Hi.

I'll be that person...

Will this work to properly animate breasts or male anatomical elements?

Thanks.

Came here to ask this, so I'll second it.

I already bought it and have been trying to do a simple animation with a G8F figure moving up and down, I'm using springs with the active facets only option to select only the mesh for the breasts. Sadly, they deflate completely. I tried using the skin young presets and same results. I did notice increasing density helps a bit in keeping the shape but then gravity affects them even worse. Also, the selected mesh now doesn't seem to follow the figure at all. Turning off the gravity stops the breasts from deflating but then the breasts just stayed in the same place while the figure moved up and down. Then I tried doing a collision with a rigid body and I just couldn't get any bouncing back effect, it almost just feels like a regular dforce sim.

So I'd like to ask if you could show how to do it like in the nose press promo picture and if this can actually be used for bouncy animations with figures (and how)

Thanks! 

Use Shell instead of Spring. There are some presets for the Fisio Materials (in Smart Content - Props - Fisio SB - Materials (they're not surface material for renders, but properties preset applicable to the Fisio Materials.

If you use Spring, use the Stiffness of the Spring property to control the stiffness.

You can't animate with keyframes the Soft bodies, but you can transform the simulation into morphs and then disable the soft body (or unsign it as a soft body) to use the morphs.

The soft bodies and the rigid bodies can collide, but when using the Non Smooth contacts System type, the objects tend to stick. What are the settings of the bodies and the system you used?

Hi, sorry for my very late reply. If I understood correctly, then sadly this addon doesn't work for what I wanted. Like it was mentioned, I was looking for a way to animate breasts and maybe other soft parts of the body, but if you can't make the softbody selections follow the rest of the genesis figure in an animation I don't see how else you can do it. Yeah, it works to simulate other types of objects or a squish in a character for a still render (like the nose render) but I don't see how it could work for character animation =/

I sent an update to DAZ (it's in the queue for testing now). The update allows the user to animate the soft bodies after the simulation, so the soft body deforms created by the simulation will follow the figure.

bashton

I'm just wondering, are you going to update it so that soft bodies that touch each other bounce off of each other rather than being sucked up into each other?

Alberto

bashton said:

I'm just wondering, are you going to update it so that soft bodies that touch each other bounce off of each other rather than being sucked up into each other?

Yes, I'm planning to improve the soft-bodies collisions, among other things. But probably it will take some months.

second_technician_rimmer_9571136c47

I wish I could figure out how to use it to make a mattress to squish under a character who's sitting or laying on it. 

barbult

I gave up and returned it. It was just too hard for me to understand, even though I had some experience and success with the original Fisio. I hate to let a technical challenge beat me, but I could see that this one was over my head.

Alberto

second_technician_rimmer_9571136c47 said:

I wish I could figure out how to use it to make a mattress to squish under a character who's sitting or laying on it. 

The mattress should be your soft body; use the Tetrahedron model for the whole body, or use the Shell model and select only a group of facets at the top of the mattress. 

The character should be registered as a rigid body (mesh shape type). Set the figure on top of the mattress or a little above. Let gravity force do its job in the simulation.

algar

https://ibb.co/fdJMMJV

All i get when i try to run a sim followed by a crash everytime, main plugin works no problem

My install path for Addon is same as fisio: C:\Daz 3D\Applications\64-bit\DAZ 3D\DAZStudio4

Alberto

algar said:

https://ibb.co/fdJMMJV

All i get when i try to run a sim followed by a crash everytime, main plugin works no problem

My install path for Addon is same as fisio: C:\Daz 3D\Applications\64-bit\DAZ 3D\DAZStudio4

It seems that it's a problem with the installation. Did you install the FisioSB addOn after the installation of the main plugin?

Could you check the instructions here?: https://www.daz3d.com/forums/discussion/comment/8276751/#Comment_8276751 and https://www.daz3d.com/forums/discussion/comment/8277136/#Comment_8277136

Let me know what you see, please.

 

Alberto

There is a new Fisio SB tutorial: