[flow_default] Transcription: 01-Compression Stiffness.json

transcriptions/01-Compression Stiffness.json

@@ -0,0 +1,7 @@
+{
+  "audio_file": "01-Compression Stiffness.wav",
+  "text": "Alright, welcome to lesson 5. So before we get to work on our dry and wet sand, I want to talk about a parameter that flies under the radar for most artists. So let's go to our villain cloth node. We're going to find this parameter that's called compression stiffness in our straightsh tab. So compression stiffness controls the stiffness of the constraints when it's being compressed, obviously. So it helps the cloth to try and keep its initial shape. But since we want our class to actually stretch and fracture and completely lose its initial shape, this is actually working against us. So just getting rid of this parameter will greatly affect our simulation and will make it look a lot better. So let's just disable this and see how this affects our simulation. All right, so you can see on the right we have our old flipbook with the compression stiffness turned on and on the left here we have a compression stiffness turned off. And granted our bag isn't being torn apart while the compression stiffness is disabled, but as discussed in the previous lesson, this is because we don't have enough grains right now to be pushing down on the bag. But one very noticeable difference is if we played this frame by frame or if we just stood on an initial frame, we'll find that in the same frame, so this is frame 5 in old and in the new flipbook, we have a higher resolution bag. You can see that this bag has a lot more divisions in it and it's still the same bag with the same points the same faces the same wireframe everything but just because the bag isn't resisting being compressed we can have this extra level of detail so once we're done working on our grains and we have it at a sufficiently high resolution this will begin to tear up again and we're going to have a much better cloth fracture result than we did right here, even though it's the same bag. So that's how important the compression stiffness parameter is. So without further ado, let's work on appraising our sand simulation and getting that dry and wet mixture look. So to make our sand look clumpy, we're going to be using the Attraction Weight and Attraction Stiffness parameters in the Vellum Solver in the Advanced tab. But first, let's bypass our cloth object so we can focus only on our grains. the So now our grains are sticking to each other and not every particle is free to move on its own. And this could very well pass for wet or clumpy sand. And below that we have the attraction stiffness attribute which controls how strongly the particles are attracted to each other. So even though we now have all the particles clumping together we can vary the strength or the stiffness of that clumpiness. So the attraction weight and attraction stiffness are multiplied by each other and the final result is the final attraction value assigned to each particle. And our goal is to have varying values of each of the weight and stiffness attributes so we have a mix of both dry and wet looks. So if we took the attraction stiffness all the way up to 100, we're going to end up with almost a solid block of grains and the complete opposite happens of course when we type lower values. So let's get our attraction stiffness attribute to its default of 10 right now. I'm going to leave the attraction weight at one so it acts as a multiplier for anything we do at our sub level. And with this information in mind, let's go back to our sub level and create an attribute pop and put it right after our Willellum grains node. Let's dive inside and create a turbulent noise node. So we have a noise map to drive our attraction attributes with. Let's hook our position data to our position input in the turbulent noise and to actually use the result of the noise node, we need to create a bind export to export the resulting values as our attribute. And our attribute in this case is the attraction weight. Also, since our attraction weight is constricted to a range from 0 to 1, let's drop a fit range node in between our nodes to make sure we don't end up exporting crazy high values to our attraction weight and risk breaking our entire SEM. So we have a lot of noise types in our turbulent noise node and I'm going to be using these pass convolution noise, it's my personal favorite. But you should experiment with different noise types and find out your personal favorite. Don't construct yourself to what I'm using here. So before we get to actually set up our noise parameters, let's first promote our parameters to the outside of attribute drop. This way we don't need to dive in the attribute drop every time we want to make a change. So I'm just going to right click on our turbulent noise, vex and warp options and click on create input parameters and What this did is it promoted all of our parameters at the front of the attribute. So it's easier to make any change in the future and I'm going to promote our fit range parameters as well this destination min and max so we can easily Change anything we want. So just click on this wheel and promote parameters and each noise type has an original range that it exports its data as for example if you use the original Pellin noise its original range is negative 1 to 1 and since we're using our sparse convolution noise we know that its original value is from negative 1.7 to 1.7 but it doesn't matter that much in our case so I'm just going to remap it from negative one to one and to actually see what our noise looks like right now we need to create a visualizer and this is very simple just click on the notes information and click on the attraction rate attribute and automatically Houdini will create a visualizer right here we have the red color signifies a value of one in the attraction rates so any particle that has a red color has also an attraction rate of, so it's going to be clumpy and opposite for the blue color, this is going to be a dry sand. So I'm going to push amplitude all the way up to 50, so the difference in the attraction weight are really apparent and I'm going to increase our grain resolution as well. So let's take the particle size down to 0.01 and let's run our simulation and see what we have right now. Alright, so now we have two parts. One part that's very clumpy that has an infection attribute of one. And on the left we have the blue dry sand. So we can manipulate our noise map much more than this so we can have a really interesting pattern just like we did with our mountain node inside of edge fracture node. And just like with the mountain node inside of the edge fracture, I already have some values that I know are going to produce an interesting looking results since I've already experimented with this quite a lot before starting this tutorial. So I'm going to apply my values right here but please feel free to play around and find your own values and what looks interesting to you and your personal project. Alright so let's run our simulation again to see how the different pattern affected the simulation result. Okay I think this looks really good. I think this looks really interesting. Now we have another issue that we need to tackle a couple of issues actually. Now all of our sand has the same mass attribute which results in them having the same weight. But we know for a fact in the real world that wet sand will be for sure much denser and much heavier than dry sand so we need to vary our mass attribute as well so we have different weights we need the width sand to be a lot heavier than our dry sand and the second problem is our friction right now our sand explodes on impact with the ground and it keeps rolling on the ground as well so we need to have a more aggressive friction value than we have right now and that bounce is also really extreme you'll find sand doesn't bounce that much in the real world if it even bounces at all. So let's first take care of the weight problem and then let's take care of the friction problem. So to vary our mass attribute we're going to use the same noise map so it corresponds with the attraction rate attribute and to do this we're going to use the same method we've used to vary our AttractionRate attribute as well. But before going into AttributeWarp and dropping down another FitRange and another BindExport, we need to first organize our parameters right here so things don't go out of hand very quickly and we have a very messy setup that we don't know which parameter controls which attribute. So let's go to Edit Parameter interface. And the first thing I'm going to do is get rid of these default nodes and folders. I'm just going to turn them invisible, hit apply, and I'm going to drag a folder and let's name this noise for the label and for the name, let's also name it noise with all small letters, apply, and I'm going to drag all of our noise parameters into this folder, hit apply again. And now we have a folder that contains all of our noise parameters. And let's do the same thing with our min and max values for the attraction weight. Let's drag another folder called this attraction weight and drag our min and max values for the attraction weight. Hit apply. And now everything is neatly organized in its folder. If you don't like this classic look and in fact, I don't like it, we can change our folder types to be simple folders and a strip of tabs. And this will create a much better looking menu. So let's dive in again, add another fit range, hook the value to the noise output and do another bind export. This time we need to export this as our mass. But remember we've already set our mass in the vellum constraints green as 2. So we don't have to set our mass from scratch. We can take our pre-existing mass attribute and just manipulate it using this noise map. So let's this time just use a bind. So to bring our attribute into a VOP, we need to bring our mass attribute and let's actually multiply our mass attribute by the noise values. So let's drop down a multiply node and multiply our mass with the noise values. So right now the mass is multiplied by zero for the dry parts and by one for the clumpy parts. And of course we don't need to multiply our mass by zero. The dry parts also need to have mass so let's multiply this by 0.1. So the dry parts are actually 0.2 in the mass attributes and we keep our mass as 2 for the clumpy parts or the wet parts. And let's set our assortment to negative 1 just like we did with the other fit node and let's promote our parameters as well. So if we need to change them in the future they are more easily accessible. Make another folder and this time we'll call this mess and drag our mess parameters inside. Change this to sample folder type, apply, accept. Alright so let's see what this gets us right now if we run our simulation. Alright so as you can see now the clumpy parts are falling faster to the ground while the dry parts are lagging behind. And this is because our pop drag effect, if you remember, we have it affecting different masses respectively so that's why we chose to not ignore our mass attribute. So that our heavy masses and light masses don't get all affected in the same way. We actually do have our drag affecting our lighter masses in the dry parts more intensely so they lag behind the heavier parts and it helps still the illusion that the wet parts are actually heavier since their impact has more weight than the dry parts but they all still have the same friction so we need to fix this now. So let's dive down into our attribute VOP once more. And you might have guessed that by now, we're actually dropping another fit range node. This time controlling our friction. So let's hook up a noise to our value. Once again, negative one to one. Let's promote the destinations, the min and max and drop a bind export for this time a fiction. Let's create another folder to hold our fiction parameters in. Call this fiction. If we have our joy and clumpy send both having a fiction of one, they're going to stop immediately on impact. But actually, as you can see, we still have our sand sliding across our ground plane. And that's because there's another multiplier for the friction attribute inside of our DAWP that's actively working against us in this case. So even though we've set our friction as one for all of our grains, We have an attribute in the comment tab here in the friction options called dynamic scale, which actually controls how much of our sand's tangential velocity gets reduced every second. So in this case, our tangential velocity gets reduced 10% every second that passes in the simulation. And this is multiplied by our friction attribute. So right now, even though we have our friction value as 1, it's actually considered as 0.1 to solve it. So let's put this all the way up to 1 so it multiplies correctly and run a simulation again. All right, so that's exactly what I wanted to happen with all the particles coming to a complete stop. So the way I want this to look eventually is the wet or clumpy parts coming to a complete stop on impact since they have water inside them. But for the dry sand I want to give them a little bit more of a room to slide across any surface that they get in contact with. So let's take our friction value for the dry sand down to let's say 0.25 so they have a lot more room to slide across our surface. Alright so we can very clearly now see the differences between the width parts and the dry parts. Before we get the bag into our simulation once again, let's do another layer of details for our sand. So I'm going to dive in the attribute VOP once more, add another fit range. And this time I'm going to vary our attraction stiffness attribute. So all the red parts that are attracting and sticking together, each particle will have a slightly different attraction stiffness attribute. So, even though they're all sticking together, the amount of stickiness is going to be slightly different for each particle. And this will add yet another layer of detail that will really help sell this wet and dry illusion of a sand. So, let's drop another folder and don't worry this is going to be the last folder we'll create in this lesson. The Attraction stiffness actually goes much higher than 1 since we've seen that the default is 10 and we've already experimented with high values like 100. So we can actually go a bit extreme with this one ranging from 0 to the dry sand and all the way up to 1000 for the clumpy sand. And before we can see the results of this, we need to dive into AdoptNetwork once more and make sure the attraction stiffness value is 1 so it acts as a multiplier and it doesn't multiply everything that we have by 10. So now we can take a look at what we have right now. Alright this is very convincing right now as a mix of dry and wet sand. I'm very happy with the behavior and look right here. We can get rid of our visualization right now since it's already served its purpose. And this is a good place to set up our caching system to desk since our simulation is starting to get heavier and heavier with all these variations and we're going to push our quality of the send up as well. So we need to start caching our data to desk instead of relying on our RAM. So what I'm going to do is drop down adopt import fields to import our data from the solver into our SOP level. We need adopt network and adopt node to do this and this is our adopt network. I'm just going to drag this and drop it here and I'm going to copy paste this at the forward slash and the object I'm looking for is the Vellum object since it's the object that has all of our data inside and we are now importing zero fields, so we have no data right now So we need to add a field and this will automatically import all of our data Everything that's coming through our Vellum solver is going to end up in this node right here And let's drop down a Vellum IO node, which is a fancy name for file cache node that's responsible for caching out Vellum data And it has three inputs just as any Vellum node does the The first input is for geometry which is what we have here and the second input is concerned with constraints and this is helpful if we need to do any Vellum post-process like detangling our simulation or adding subdivisions or smoothness to it. So let's cache our constraints out as well just in case we need them in the future. I'm going to copy this over and in the field right here I'm going to type constraint geometry so this node knows that I want only the constraints not the actual geometry and as you can see here we get the constraints for our sand I'm going to hook this to the second input and before caching our simulation out let's make sure we unbiopass our bag so we cache both our bag and the sand and I'm only going to cache 72 frames at this time. So I'm going to head save to this and when this is done, I'm going to do another flip book so we can take a look at what we have now. All right, now that our cache is done, let's check load from desk. And before doing our flip book, let's actually split what we have. We now have our sand and bag cached out together. So we need to split these two so we can work on each element separately. So let's drop down a split node. And in the group right here, we're going to find a stream cloth and a stream grains. And these are actually made inside of our dotnet. So here in our Vellum source, we have a stream tab and the stream is essentially a group. And it gets name whatever we name our node by default. So we should expect to have stream class stream grains and whatever else we have is going to have its own stream group. So let's select our stream cloth and whatever we select right here is going to come from the left output of the node. So let's drop down on null, call this bag and another null for the grains from the right output. Let's call this grains and let's drop down a color node to better visualize our greens since we got thread of our visualizer. So the same thing our visualizer did for our greens we are going to do again, but this time using actual color data. So let's change our color type to ramp from attribute and the attribute we're going to use right here is the attraction weight. And let's make a dry sand have an off-white color and our clumpy sand let's give this a brownish color so let's merge these two elements back together and see what we get okay so this is a much better result than anything that we had up till now I'm very happy with the way that the bag is tearing apart and this is in a very large part due to getting rid of the compression stiffness parameter and I'm also very happy with how our grains are looking right now. I think this is a good place to end our lesson. I'm very happy with the progress that we've made right here. We still have a lot to do of course. This has been a long lesson I know so thank you very much for paying attention and keeping up with me. I hope I've explained everything clearly and I'll see you guys in the next lesson.",
+  "language": "en",
+  "confidence": null,
+  "duration": 1298.06
+}