Week 9 continued with the Digital Tutor lessons on Fluid Containers:
Display Attributes of the Fluid Shape Node
- Display – As render – this is as you would see it! All the other options have opacity setting.
- Voxel quality – the default is the ‘faster’ setting, ‘better’ is more advanced but slows the simulation down.
- Boundary draw – outline/full/bounding box for 3D containers, they are intuitive enough to use – ‘bottom’ or ‘reduced’ are usually the favoured options
- Numeric display – Density – this option shows the numeric values on screen, lots of numbers! It would be the same for Temperature and Fuel if there were any in the scene!
- Wireframe display – this is good if the system is struggling; particles are just points; rectangles give more visual information and seem to work more quickly
- Velocity draw – this turns the arrows on and off which show the direction of the simulation. If you have a high density simulation, use velocity draw skip to show the direction of simulation.
Dynamic Simulation Attributes
(You can create containers and emitters directly from the shelf as well – in this case the, the 5th one along – 2D with eintter)
Gravity – if you want your fluid to fall down, but not all everything else in your scene, you need to set the buoyancy to a negative number, not the gravity! (-9.8 gravity will pull everything down).
- So, use the contents details attribute – Density – set buoyancy to -1 (gravity still positive!).
- Viscosity – as in nParticles, this makes it thicker
- Friction – this slows down contact with the surface of the container
- Damp – the amount of energy it is losing as it is moving – this might be useful for creating a slow moving ooze or drip
It’s is possible to create textures with these containers – i’ll discuss this later!
- Solver – spring mesh is best for things moving in and out. For example, the ocean or a pond. Navier stokes – use this as the default.
- High detail solve – all grids = high quality, but slow! All grids except velocity – density = HQ
- Solver quality – again, the higher the number you choose, the longer it takes!
- Grid interpolator – linear is default; hermite is an accurate simulator setting. You can offset the start of the simulation by using the start frame setting, e,g, strart at frame 100
- Simulation Rate Scale – this speed up the rate of the simulation. NB – if you’re doing more than 2, you might want to rethink your simulation! (It doesn’t look good!)
- Disable evaluation – only need to use this once you’ve cached your simulation.
The other check boxes, usually leave on (conserve masss etc.).
As was mentioned in last weeks blog, a task was set to create a head in Mudbox, the purpose of which was to familiarise ourselves with using the basic tools available.So, starting with the basic imported head made available to us, I used a variety of the sculpt, grab and opinch tools etc, as well as a couple of the stamp and stencil tools to add wrinkles and spots etc. Useful points to remember when using Mudbox include the shortcuts that relate to the various sculpting tools:
- 1, 2, 3, 4, etc – relate to the sculpting tools
- Shift left mouse drag – smooth tool
- CTRL + left mouse – invert
- B + LMB drag changes the size of the brush
- M + LMB drag changes the strength of the brush
Anyway, for my first attempt, here is my alter ego:
Fluid Containers – Adjusting the Density of Particles
In order to control the density of the particles in a container, first change the density scale:
- Density scale: 0 = can’t see it!; 2 = very dense!
- Buoyancy – as mentioned earlier, if set to 0 the particles will not go anywhere, if a negative number they will drop and if positive they’ll go upwwards. If you wanted to animate an explosion, for example, animate the containers buoyancy and scale. You can set it to 0.001 for a slow explosion!
- Dissipation – this is how it fades out and disappears
- Diffusion – this controls the spread!
- NB – it should be noted that there are additional options in Maya 2013/14 that are not covered in these lessons. They include:
- Density pressure (good for explosions!)
- Threshold – how far it can go!
- Density tension
- Tension force
- Gradient force (the last 2 seem to affect the ‘mushroom cloud’ look)
Introduction to Nuke
To break things up a little this week, I decided to have a look at an introduction to using Nuke. The project used includes a robot doing a little somersault, a background plate, a shadow for the robot and a highlight for the robots gun, all on their own layer. I have used After Effects to a very limited degree before, but Nuke is node based compositing software rather than the Adobe layer based type.
The man screen is known as the viewer, and there is a viewer node available before you read any other attributes into your file. All of the buttons above the viewer relate to ways in which you can change the view. On the right of the screen is the properties bin – this is where the attributes of whichever node you have selected will appear. The node graph is at the bottom and the toolbar on the left.
To get images or video into the scene, use the image node – read – if you have a sequence of images you can check or uncheck the box to read in the whole sequence or choose individual images. Alternatively, you can drag the whole folder into the node graph and Nuke will recognise it.
To view the images on the screen you have to connect the arrow, known as the pipe, from the viewer node to the chosen images, for example, ‘read 4’.
- Playing back – there’s a green line at the bottom of the playback that indicates the level the video has been cached. Once it’s solid green it will play back at a reasonable speed.
- MM Scroll can be used to zoom in and out of the screen – there is a measure at the top of the screen to let you know how far you have zoomed in/out, e.g. ‘x 1.7’
- MM drag will pan the image
- MM click will centre the image – the above are the same for actions in the node graph
- To see the project settings, move the mouse over a blank area in the properties bin, press ‘S’
NB – Make sure that you set the appropriate resolution to be the same as your clips.
- You can remove all of the properties windows by clicking ‘X’ at the top of the bin. (or CTRL + Shift + A)
Merge Node – this is used to merge read nodes together (like merging layers together in AE). Press ‘M’ in a blank space in the Graph node area or use the toolbar to access the nodes. With a merge node there are 2 pipes available – A and B pipe – the B pipe is generally used as the ‘background’ with the A pipe as the foreground. So, in this case:
- A – this is for the robot node
- B – this is for the background plate
- Viewer – plug this in to the Merge Node
Please note that keyboard shortcuts are panel dependent! So, be careful to press M over the node graph area – you may open up something else otherwise!
Merge nodes have an operation attribute which determines which order the image is presented – for example, the default is over, but you could have it under or any number of other options. You need to work out how best to arrange your node tree. It seems there are a number of ways to do this. My version of the digital tutors version was slightly different in that I swapped my read 1 and 2 nodes compared to them and used an over operation rather than an under operation, as seen above in the image. The same result was created.
If you want to disable a node to see what it looks like before and after, select it and press D.
As the tree grows, it’s essential to keep it tidy and organised! It’s a good idea to only use straight lines not diagonal ones – we can do this by adding in Dot nodes. So, the idea is, that B pipes go down, and A pipes go across. To add in a dot node you can do so by either holding down the CTRL key, pick a dot on the available pipes, let go the CTRL key and position as appropriate. Alternatively, hit the tab key over a blank space and type in dot to the text space and the node will be made available.
Here is a tidied-up version with three merge nodes and three dot nodes:
Fluid Containers – Velocity and Turbulence
- The velocity attribute – e.g. 0.25 oon the y axis will give a slimline stream due to the decreased velocity (like cigarette smoke) – if you open up the y boundary to ‘none’, and increase the x velocity to 2. Z velocity is important for 3D containers!
- Swirl attribute (low setting – wide patterns/ high setting – more chaotic!).
- Turbulence – strength – like noise, while frequency – the higher the number, the tighter the pattern. Speed – of noise – the higher is more like a windy day!
For all of the above attributes, it’s better to add just a little for effect!
Temperature, Fuel and Color
- this is greyed out until you turn it on from Contents methods – turn on to dynamic grid – need to change display to temperature to see it. Now you can adjust Temperature Scale. The emitter has temperature as its own emitter, so you can have it at a different rate, e.g. buoyancy etc. Set the temperature attribute as you see fit!
- turn it on and make it dynamic grid. Fuel scale – this is very quick, it burns out fast, so crank it up to 2 to see it! Also, slow down the reaction speed to make it last.
- Ignition temp and Max temp work together. It’s akin to saying ‘don’t ignite until you reach this temperature!’ (ignition temp), ‘keep going, using up the fuel, till you reach max temperature!’. Anything after the max temperature will quickly burn off the fuel.
- Heat released – you may want to shade differently
- Light released – the light color will multiply whatever the color is with the incandescence shader. So, if the shader is showing temperature for incandescence, then the light released is a modifier of that temperature.
- To view – set display – shaded display – as render
- Color – change from white to black as this will multiply the shader by 0 to let the incandescence show through (transparent)! (So whatever the colours in incandescence are set to will show through)
- So, we can now tweak the temperature and fuel settings. For example: increase buoyancy (temp); reduce fuel scale (fuel); decrease reaction speed (fuel); increase ignition temp (fuel); decrease heat released (fuel).
- After all of that, it doesn’t seem to make much difference!
Color – (not within shading!) you need to turn this on from contents methods again.
- Color – dynamic grid – this has a slightly different look – it kind of mixes the colours
- Color dissipation – blends colours together
- Color diffusion – spreads out colour (blur)
- With no dissipation or diffusion it seems to be hazier!
For this exercise I created a 3D container with emitter, with a resolution of 25 x 50 x 25 (pretty low!)
- Density scale – (contents details/density) buoyancy set to -1
- For the time being, go to shading and remove all transparency
- Surface attributes – activate surface renderer button – you’ll need to up the resolution for a better render!
- NB – hard surface – sharper edges; soft surface – blurs edges (not much!)
- Although there is surface threshold (blobbyness) and suface tolerance (the lower the number, the more it has to compute that frame – fuzzy edge – change render settings to production quality)
- Specular color – highlights!
- Cosine power – this one was not explained!
- Environment – you can could create a gradient ‘sky’ – the colours will reflect in the specular
- Shading quality attributes – up the quality to e.g. 5
- change from linear to smooth – much better but slower!
- If you want to change the general colour, go to the shading attribute – color input – constant
- You have the ability to override the output mesh settings (resolution) – output mesh attributes – mesh resolutions 1.00/2.00/3.00
- Can also use gfradient normals checkbox for higher quality – in Maya 2013 you have a number of adittional options as well [refractive index (environment attribute) – transparency – this will refract the image or object behind it, although it’s best to do this in post or as a separate pass.
Converting a Fluid Simulation to Polygon Surface
Modify – convert – Fluid to Polygons
You need to hide the fluid simulation, tesselate the polys etc, and shade it! Add in bump and displacement maps as necessary.
Thursday 7th November
Jeanette’s class today took a further look at research methods, in particular – Questionnaires, interviews, observations and focus groups. We were asked to identify which methods we would likely be using, following the clarification of our research question. So, at this moment:
- Practice/experimentation -as this is a new area of research, practice is essential to test and try innovative approaches using current techniques in 3D dynamics
- Literature search – current literature of physics of sound and results of cymatic research will be essential for accurate modelling of the idea
- Observation – observation of the images and videos produced by the cymascope team of the piano scale of 12 musical notes will be essential, and if I can build my own version of a cymascope, my own observations may well help to further visualise the idea
- Email/Interview – I intend to contact current researchers for information on any additional data they may be able to send me, especially data on 3D co-ordinates of basic cymatic patterns, if they exist!
Clearly, as the research develops, these will no doubt change.
For the past couple of weeks, I had begun modelling the assignment for Jin, which included the creation of a pair of something, like shoes or gloves etc. I had taken a series of photos of a boot, opened them up in Nuke and removed the lens distortion. Because, for some reason, I could only open up a training version of Nuke, the updated reference photos had an array of coloured squares! Nonetheless, I continued as they were not too much of a distraction.
Because I had begun the modelling in my home version of Maya 2014, I had to export the files as .obj files in order to continue the work in the university’s 2013 version. I completed the modelling part of the assignment, so next up I will be looking at using uvlayout, as long as I can get a copy soon, hopefully without having to pay for it!
Friday 8th November
This friday, Kieran Baxter took then reigns for the tutorial on multi-pass set up and rendering basics. So, as an example, Kieran used one of his own files to demonstrate rendering a single static image, but the same rules would apply to a series of animated images. We were going to separate the scene into different render layers:
- layers for the lighting (Key, Fill and Backlight)
- a basic beauty pass (master layer)
- an ambient occlusion pass
- an ID pass (for depth)
The first thing to do was turn off the displacement map for the frog as this would slow down the system.
- In hypershade – copy the frog shader and paste in a new one as this will remove the displacement map, then re-apply to all objects in the scene
- to see the nodes in the working area, right click over them and select graph network
- In the Maya viewport, select view – camera settings – resolution gate (we had set it to 720p). This will provide a suitable gate to frame the scene
- Turn on mental ray shadows for the key light – the umbra and penumbra are the way that light shadows spread out. To control this, use the light angle
- If the shadows are a bit grainy, up the shadow rays to around 12 (or suitable number)
- Select the models and the key light and create a new render layer (as seen above)
- Do the same for the fill and back light
- Make sure you have the geometry and the light on each layer
- The ID layer (is like a mask) but it uses the RGB channels. Create a layer with the camera and geometry only (wood shader).
- Set overrides – right click over ‘color’ and select ‘create layer override‘. Then RMB and select ‘break connection‘, and slide the color slider down to black
- Do the same thing for the specular color
- Override Incandescence – then load in ID matte (In photoshop we can break the channels up by using RGB)
- To create an Ambient Occlusion shader – create a maya surface shader, name it AO, open it up in the graph network, click drag in an mib_ambient_occlusion texture from the mental ray textures and connect the 2 nodes (MMB drag – default)
- Once the shader has been created, right click over the shader – assign material override for AO (turns black)
- No displacement on this! AO layer – RMB – attributes – RMB – shading groups – break
- Assign material to selection (floor!) (updated AO shader)
- Can change the max distance of the AO shaders attributes depending on the size of the scene – e.g. 5?
- Up the samples as well to 32 or 64
- So, once all of this has been done, open up the render settings – passes tab – adds scene passes
- Master layer – diffuse material color
- Layers for each of the lights – diffuse and specular
- ID matte – depth
- Render settings – common – recommended settings here: TIFF format; name_hash.ext ; frame padding 4; renderable camera – camera 1 in this case
- Render menu – batch render
Once your passes have been rendered out, in photoshop add your layers by using Add (linear dodge) or multiply. The order is not importanmt as the layers are being added together. We will do a shader rebuild first of all. CTRL + M is the shortcut to add curves to layer (image menu – adjustments etc).
For the ID matte – RGB channels:
- load in the red selection
- turn RGB back on
- turn the layer off
- add adjustment to change as you see fit
- then do the same for the other channels!
Layer mask for the ID matte – before this create a copy of all the other layers into 1 new layer. Add a lens blur on the depth pass – Filter menu – blur – lens blur. Drag the mask to the bin – delete! Simliarly, in Nuke add the nodes together as has been seen earlier in this post! You can switch between viewing different nodes by using the numerical keys, 1 up to 9!