Klockworks.net

New reel for the end of the year. No Side Effects work at this point. Please enjoy!

I realize my reel is rather old, I cannot put things on it from my time at Side Effects in an open space, yet. However, I can send you my latest reel and breakdown sheet by request.

That is all! Thanks.

You may or may not notice, but I disabled the links to my scripts pages. The reason for this is, they really contain nothing. I have never, found a good way to display scripts on a reel or my site, and until I do, the links will be dead. I really am at a complete loss for how to display them, short of posting code, which I consider bad style and a waste of time.

Though in all honesty I should update some of my code. After a conversation with a developer at SESI he showed me the error of my scripting style. Just little things(that I was never taught), about how to make better code that in the long run will be easier to maintain and read.

Not that I expect many people read this that I don’t talk to regularly(or every once in a while.) I thought it would be a good idea to give a slight update. I am, and have been, at Side Effects Software Toronto, doing an artistic(vfx) internship. This will last until mid December when I head back to the states and then the west coast.

So… Yah. Not dead, just busy.

So, for some reason two of the links were missing in my nav bar on my site. I put them back. If you look in the portfolio section you should now see links to my RenderMan 1 and 2 classes. Why they dissapeared, I don’t know. The pages were still there. If it isn’t there for you just force refresh your page(ctrl+shift+r) and it should show back up.

I got this guys.

I have a new reel(the front page has been updated with a local copy) and would love to share it with you all!

There is one person who consistently does more silly things in hypershade that work than anyone else here. That person is Jared Reisweber. The part that makes it so awesome is that most of his sillieness works. One of the silly tricks he has figured out is a way to fake thin walled translucency. Yes mia_material_x has translucency built in, but his method produces better results more consistently. It also has the added benefit of being much simpler to artistically control. I have taken his method and altered it slightly for a simpler network.

The network is simple. You can find the misss_lambert_gamma under the textures section in the mental ray shaders and the misss_fast_shader in the materials section. On the misss_lambert_gamma node open up the lightmap content tab and set the flip to 1. As stated in the node documentation “flip can be one of the following: 0, normals are not flipped; 1, normals are flipped; or 2, both the unflipped and flipped-normal sides are lightmapped. This can be useful for translucency in thin objects such as leaves.” When you add the fast_shader to the network you can go ahead and delete the mentalrayTexture and the fast_lmap, you don’t need them. Plug the misss_lambert_gamma into the diffuse shader box.

You will notice I turned off Front and Back SSS, you won’t need them either. In the misss_fast_shader the Diffuse color will be the “translucency color.” As in the translucency contribution to your final look. The Diffuse Weight becomes the multiplier of the translucent effect when it is added on top of your shader. From there you can plug in the misss_fast_shader to the additional color slot. You will have to scroll down in the connection editor to find result as well as additional_color. From the node docs “The additional_color is an input to which one can apply any shader. The output of this shader is simply added on top of the shading done by the mia_material and can be used both for “self illumination” type effects as well as adding whatever additional shading one may want.” It can almost be thought of as an ambient contribution to the material_x.

From there you can just tweak the Diffuse weight until you get the translucency right for your scene. Enjoy!

Description to follow!

You read my last post yes? No? I will give you a moment. Go read it… Now. Done? Good. So the process is innately the same with a few KEY changes. One of them being the new render is unstable. I know how to fix it, but that weird speed up slow down is because of the data acting funny. I didn’t realize what it was until after I had rendered out all 126 of these beautiful frames. Even the random dark ones in the middle. Those two frames? No idea.

One thing you will notice is it feels much more like liquid is shooting out from the tendrils as they drift downward. I need to tune it as well as add back in the curl noise. I wanted to see what pure advection would look like and its very pretty, but not quite what I have planned. Oh another thing, this test is only 3.9 million particles.

The biggest changes from last post to this one are internal structures and, obviously, the fact the ink is now colliding with a “container.” I actually came across how I have the system set up by accident. I was trying to get two smoke systems to mutually collide(which they were not doing even with the mutual flag on). So I instead made a static system from a SDF that acts as a container. No forces act upon it, and it just sorta hangs out. I will probably go back and only use the fields I need with a frakensteined solver and object.

Something I haven’t integrated into the DOPS system, but it is created, is a ripple for the top of the beaker fluid surface. This will give the impression the ink is actually pushing against the top and add another level of complexity to the shot.

One other thing I changed was the base particles. I moved a lot of things around, made the system more efficient and happen in half the frames. This is good as before that major change, the entire shot was around 960 frames from start to reaching the bottom of the glass. Now it reaches the bottom in 400 frames and then goes into singularity mode at around 550. Yes, I am having it go into a fake singularity. While putting that together I found a weird set of values that not only compresses it how I want but also creates a nearly flat orbiting ring around the attraction point. Oh and the ring wobbles like it is in water. Purely accidental and it looks awesome. That and I did it all in VOPS. Why? Because I like low level control. And figuring out exactly how I wanted it to work was a fun puzzle while it lasted.

I may post pictures of the vops networks as well as a description of it in a few days. I have been pulling all nighters and getting about 3 hours of sleep a day for the last 3-4 days. This will continue to happen until I can get most of the major things I need done. I want to be at a point where I am waiting around for full simulations by the end of this coming weekend. Why? Simple. I need to have SOMETHING to show to Pixar when they come. Which is soon.

Also, when I do post the final video, be on the look out for some easter eggs. I like hiding things within scenes.

Figured I should post a test of what I am actually doing. A short explanation will come after.

So what is going on in this awesome animation you may ask? Particles. Millions of them. Actually around 5.9 million. This is just one pass of the simulation. My ultimate plan is to have 3-4 layers of this and push 15million particles in the final render. Which is going to be awesome. So, I have a core of particles, around 25k, that drive 90% of the final image. These particles are used as a source for smoke as well as advecting the created smoke and then advecting another smoke sim. The secondary smoke sim simply moves up and gets pushed around by the aforementioned smoke. This entire smoke system, that is both of them, then advects particles that are split off from the base particles. Not quite a two way advection but close.

Now, keep in mind that is just a test. One I am working every day to improve. One of the major issues, is if you pay attention to the upper left corner of the sim you will see some particles get a little… over-zealous. This is due to there being no data in the fluid advection in that area and the curl noise pop gets a little silly. One of the things I am adding as we speak is collision with the fluid that this entire sim is happening inside. This should help cut off that silliness and help the smoke advection to better control the overall sim. Adding all the water information in will also be good for when I have particle fluids for the surface.

I am also talking with Ray Popka, a fellow SCAD Houdini dynamics artist, about this project and as always he has great ideas. You should go look at his stuff. He has some crazy ideas for helping efficiency.

I had a rather frustrating weekend last week. One of the problems with wanting to do an epic project is it can get away from you. One of my goals was to get the movement of actual ink within water. Yah, that doesn’t look like it is going to happen at this point. Maybe for a moment on the initial droplet, but not past a VERY brief occurrence. I have tried many methods and so far have gotten no where closer to even getting a semblance of the look of it. As of right now I am dropping that from the list of wants from this project and moving on with the rest of it. If I have time I do plan to attempt to get more movement out of it. Deadlines are a pain when you really want to get something to work.

Though I was just shown a shot from Harry Potter which gave me new hope not only for the shot but for the effect itself. It also gave me and idea for a pretty neat camera angle. Stuart was thankfully sitting next to me and we discussed how to best apply this to my short. I have to say having Stuart sit next to me and help me with the camera work was awesome. I have never been the best with cinematography, so having his input and ideas are invaluable.

At this point, I am working on perfecting the system that does the effect so that I can go back and art direct certain parts.

I will finish my previz for this and post the movie up before the weekend is out(and I mean that). As well as a test render for the effect so far. Though it will only be probably a 10th of the amount of particles in the final effect. Which is going to be awesome.

More to come!