Day 58

Today I changed the tweaked the explosion's displacement shader.

new displacement

After tweaking I rendered passes of each effect I've got and used matte shading if the geo is overlay.

Day 57

Today I tried animating the Geometry light and I manage to add a expression so that when the explosion ball increase in size the radius would too. But the problem is that my light does not comes out at frame 1, I tried expression but cant seem to work. I tried timewarp but as soon I lay down the node they disallow me to modify with the light.

Pop out

Today I asked Wei Jie for his shader, it seems that he had already finish with the shader and dint informed me. Just wonders what if I dint ask. Time is already running out and so I got started with it. To my depress, the geo of the sea was changed. I had to retweak the whole impact wave. I spent the whole day retweaking and had to stay back in school to watch my render.

Day 56

Today I changed the shader for the shockwave2 and explosion as Mr.Douglas wants something orangy and so I did but Mr.ron dint like it and help me it the shader.

New shader

he gave me another choice and gave me a file full of old houdini shaders. There was a flame shader and it was fabulous. I on the geolight and rendered.

Flame Shader

There shouldn't be any shadow ,so I remove the shadow by unlinking the distant light to the geo.

without Shadow

Day 55

I changed the ShockWave1 to be more deform.

Deformed Geometry

Today I thought of adding light to the explosion and geometry light seems to be the right choice. There was a  problems at first, the lights were dotted spots, I tried increasing the sampling quality but the problem lingers . I realise that the displacement shader plays a part in affecting the light shone. The fractal noise was the reason the lights was in spots.

dotted lights

 

Day 54

Today I tried using a glass shader for the shockwave1, I was trying to achieve the refraction and transparent effect. Then Mr.Ron came, and so I ask him for help with the shockwave 1.

For the shockwave 1, the method used was to render the object separately and deform it in comp later on. This saves me the work and render time from trying to tweak the glass shader to achieve the similar effect.

comp network

The deform is made by having the image to deform (impactwave) connected to the left side of the input of the deform node and the deform plane (shockwave). This effect is then refined by using a blur node and bright node.

w/o blur and bright node.

w/o blur.

Result.

But when I render I realise there is a problem, the blur cause the edge to be rounded when the shockwave reaches the edge of the frame.

So I scaled up the frame of the shockwave hiding the flaw out of the frame. That would change the timing a little bit, but it's not that much of a difference.

Day 53

Today all I did was tweaked and render to view the result and repeated the two steps until lastly I was satisfied. There was a lot of "trial and errors" videos for the testing so I am only showing the end result.

For the secondary dust, I just duplicated the same emission of the dust and delayed the emission of particles and also tweaked the particles to refine the simulation.

I find that this was naturally and good to go for the presentation. Tomorrow I plan on finishing the shaders for both shock wave.

Day 52

I got back to school and checked my render and it crashed half way through the night. There was a problem with mantra. Mr. Ron helped me with the lighting and it seems to solve it. He changed the Ray-Trace shadow to Depth-Map. And also he helped me with the caching of the impactwave so that I can work better without delay.
And so I rendered and went out for lunch and this is what I got.

There are some adjustments I would have to fix. Firstly is the secondary dust isn't coming out and secondly, the too much drag acting on the rock. I decided to remove the drag and increase the attractor a little.

Today Mr. Leong came in and reminded that presentation was on next friday that shocked me and got me to start planning.

Things left to do

-secondary dust

-shockwave1 shader

-shockwave2 shader

-explosion shader

-pyro(optional)


Hopefully this things can be done by this Friday.

Day 51

This is the render with the total effect. I have not done the shader for both the shock wave.

I realise there is problem with the explosion ball, it stopped at frame120. I found that the timewrap for the explosion ball is set to frame120. So I went on to the secondary dust and render. 

And realise there are some adjustment need to be made to the rock as the emission is quite slow.

Day 50

So I changed the geometry as well as the rotation. And it's working fine.

Today when I came back school I found that my render had crashed, I tried re-rendering but the problem still persisted until quite some time. It stated that mantra had stopped working but dint state the cause. Anyway I was able to render the scene without the sea glass shader on. It takes about a minute to render and 2 when it was about after frame70. At the same time when I was rendering I did the timing of the explosion ball and render through flip book.

Day 49

So again I rendered over night with Wei Jie's shader. The rotation of the sprites were still there.

Through trial and error by offing the effect and render, I managed to find the problem. The problem was caused by the deformed geometry surface which emitted the particles.

simple illustration

The black line is the deformed geometry and the red line is the attribute transfer. So as soon as the geometry cross the red line the particles starts to emit. So surface1 will emit first since it passes the red line first then followed by surface2. So the sprites1 rotates first then the sprites2 rotates. This pattern repeats as there are new particles born. This gives a wavy illusion to the sprites.

So I changed the source of emission from the original impact wave.

not deformed geometry

In this case the sprites would not cause the same effect.

I also decided to change the rotation of the sprites to be rotating inwards the source instead of rotating outwards. 

illustration

Since I couldn't find any videos on aerodynamics references, I drew a example.

Day 48

 So during the weekends I left my computer to render and this is what I got.

The Sprites tend to be rotating altogether which isn't good at all. Having this problem, I suspect that the cause of it is due to the axis line which the sprites are rotating along.

The camera perspective is in the middle of Z & X axis. So as you can see the blue line is where the sprites are rotating along and the red line is the place where I want the sprites to be rotating along horizontal to the camera. So I asked Mr.Ron to help me with the changing of position of the line which the sprites rotate along.

Voppop and  attribute node.

Voppop network

Attribute node

 Firstly Mr.Ron added a Voppop node before the sprite node. In there he create a vector where got it from subtracting X & Z axis. He then name it with the attribute node. He also adds the attribute node at the pop network to link it.

Day 47

So today I came back to school to work on shot 3, which Mr. Douglas had done the shot blocking. I had to merge the sea of glass with my impact wave.

Merging both effect.

Deformed geometry.

Deformation of the wave network

 The geometry is deformed using lattice. With the lattice the it is required to have the Geometry to deform (sea of glass), the rest geometry (grid of the impact wave) and the Deformed geometry (impactwave).

Emission of the dust network

As for the emission of the geometry I have to have the surface where the particles are going to emit. I used the original sea of glass and delete the deformed sea of glass to have the deformed geometry only. From there I used the surface to emit my particles.

Day 46

Today I realise that I got all the lights that came from Joyces's  left on, that was why my scene took so long to render.

Anyway today Mr.Douglas came again and scrapped the pyro in shot 5 as it did not go well with the scene. And told that the pyro can be used to implement on shot 3. I was also told to work on shot 3 and Mr. Douglas did the blocking of the camera.

Day 45

Rendering shot 5.

I took a frame of the pyro sequence and did a quick composite with the clouds in photoshop to have a rough visualization.

Updates about the impact wave, I forgot to mention that I was taking reference from the scene from the movie "Deep Impact".

@0:44

Day 44

Today Mr. Ron came and told me I should be rendering the pyro individually so I can have more options when post production. So I "rerendered" the sequence.

Meanwhile I worked on my impact wave and tried to implement the shock wave to it. From the concept learn from ruth's shockwave, I modified and tweak it to suit my scene. The timing and scale have all been changed to fit the sequence. Although there is some problem with the timing, I am sure it could be done.

Day 43

Rendering the pyro explosion for shot 5. It takes around 30mins to render a frame. I also worked on the explosion scene but only able to see the effect through viewport while rendering.

Day 42

Today I tweaked the explosion and rendered the whole sequence. Here are some test renders.

Day 41

Came back to school on a Saturday to have the layout completed. I used Joyces scene to position the explosions.

Rendering only a frame became a nightmare after I composited both together.

Day 40

Day 40

I tweaked the pyro shader to have the colour I wanted. And also Mr.Ron teach me how to duplicate the pyro and changed the timing.

Day 39

Day 39 Shot 5

Today I have been placing the Pyro explosion in Shot 5. Scaling and tweaking the Pyro. Since testing Pyro is time consuming very few things could be accomplish today.

I got the Pyro looking somewhat like the researched explosions on the internet.

Day 38

Happy Deepavali Day

Day 37

Absent.

Day 36

Day 36 Pyro Referencing.

I was trying to get reference on the explosion, so at least I've sort of a target to work towards to. Here are some of the references Mr. Ron wanted.

Picture reference - Volcano eruption in Chile

Video 

There weren't really explosions in the picture so I tried finding for a same effect. I believe the effect Mr.Douglas wanted was a round, dense explosion, more of a "firework" shape. First thing that comes to my mind was underwater mines going off.

Finding Nemo - Underwater Explosion

Super Nova Explosion in space.

With this explosions without the water and with pyroclastic dust.

I found this on a forum. I think I can use it after a tweaking and adding pyroclastic dust.

Day 35

Day 35 Fading sprites & Sprites Rotation

Today Mr.Ron came in and taught me couple of things. Firstly he taught me how to fade of the spries according to its life with a ramp to control the timing.

alpha ramp

And also I asked him about the rotation of the sprites. As normally in other scene files they show the sprites rotation in a direction, clock-wise or anti-clockwise. What I wanted was, as the explosion comes from a source the rotation have to be away the blast.

So, he taught me the how to rotate the sprites in 2 opposite directions. By having a $F * sign($TX)*10, the rotation will be along TX but in opposite sides and the rotation velocity will be constant and not increase.

simple illustration

adding expression to rotation

I notice there are some problems though, as the rotation is only to X axis the rotation does not look good from sideways. The rotation will be inverted if the view is from behind where the X-axis is pointing to the other direction.

He he later then taught me on adding motion blur to the particles. Blurring the hard lines when rotating as well as giving the effect of where the particles are coming from.

Add velocity on point node.

After tweaking

Day 34

Today I worked on the dust and this is what I got.

Fixed the dust.

Mr.Ng came in and viewed my work and advised me to have the randomisation of the shape and sizes of the rocks to be more various.

So the researched began again. Luckily I remembered Lu Sheng did almost the same stuff as me in SESI, but it wasn't what I wanted. From there I linked to other group members, and stumbled upon Jia Bao's Dragon Flood project. There he did some rocks and I took reference from how it was done and redid with my version.

Rock network

It was basically using mountain to deform the sphere and having more surface noise added to it by Vop Sop.

Vop Sop Network

I created the material shader as well.

Surface Shader Network

Render Display

I asked Mr.Douglas about the surface material of the debris and he adviced it to be something like the floor material Wei Jie's been working on but not as detailed.

So far the explosions have had great response from the lecturer's and Mr.Douglas, very glad of that phew~ So determined to finish the project with all I've got! :)