#dynamicsvfx

To complete the assessment.. 

I imported everything I had into Nuke and created a bash comp - integrating all the different pieces together. 

I only ended up changing a few things to change the overall look in my composite, them being the timing slightly (using a retime) to get them to appear and disappear when needed and also a translate node.  I don't know why but my particle systems were a little off when putting them together and didn't fit in with the ufo opening exactly, but once I did this it blended together nicely. 

I also incorporated my heat plumes, which I rendered out separately (back and front) and then put them together. I had a little trouble with getting these working and had to render them out multiple times and refer to the class example we did. But in the end, I got them working the way I wanted them to and was happy with the result. 

With the mannequin, I ended up rendering that in with the background (maybe this was why the particle system was a bit off centre, I don't know why). But found this worked well for me because I tried doing it separately, and when I did the mannequin wasn't integrating with the background (getting sucked up and disappearing into the ufo). Anyway, I completed this and was happy with the result. :) 

For the second part of my assessment, I looked at the feedback given by the client and pretty much just recreated the particle system in my new scene, changing a few things along the way. 

Firstly I created the same kind of look as the first scene, using toon shapes again. Only this time I used a cylinder shape instead of a cone shape - following the clients feedback from my first submission. 

Other than that, the only thing I really changed was the timing and the colour. I made it so that there could be two rings descending from the ufo at the same time - rather than just one. And also made the colour of one of the rings slightly lighter than the other. 

I felt I could do this to add more variance, and only because the systems together wouldn't look too over the top due to the now stationary coreBeam. I was really happy with this look overall and stuck to what the client wanted so hopefully they're happy too ! 

For the third particle system, I did the same thing: Looked at what the client wanted changed and did that. Duh.  So for this, instead of having the particle system rotate in different directions over the time, I made them rotate just the one way. This also made the tractorBeam less complex (another reason as to why I changed the colour up a bit on the system written about above). I copied over most of the values of the last system I made in the first scene as I was quite happy with how this turned out and the client didn't say anything about changing it, so hopefully they were happy with it as well. 

I also changed the shape of this particle system so it integrated nicely with the other three, now that they were the cylinder instead of the cone / pyramid shape. I felt that this gave the tractorBeam a lot more variation compared to the previous scene, and was happy with how it turned out. 

To complete the scene, I rendered out the particles in different layers so that I could have some rendering at uni and others rendering at home so it could be done quicker and then I'd have more time to play around in Nuke - editing them slightly so I was happy with them. 

Overall, I was quite happy with the final output of my particle systems and felt I adhered to the brief and clients feedback as much as I could. 

To begin working on completing the second part of the assessment, I've re created the scene and written down all the main values and edits on the emitters so I can recreate them in the new scene. I wanted to completely start from scratch, so that none of the issues that were arising in the first assessment would do so again. 

I just took into account the things I needed to change and revised the way I built up my systems. 

Firstly I did the coreBeam.  For the coreBeam, I filled a cylinder full of particles (balls) and worked with nDynamics. I found that this approach was good and did what I wanted it to do in the first scene, so therefore kept it going throughout the assignment. I just simply changed the shape of it, by creating a cylinder geo the correct size and filled it with the particles. I mucked around with the particle size, the opacity, colour, count and start frames for quite a while, trying to get my clients desired look. 

I also added a turbulence field to add a randomness to the appearance and makeup of the particles in the filled object. Otherwise they were extremely uniform and quite frankly just looked ridiculous. This is where I played around with the start frames and things to get the system to look the way I wanted it to by the time it emitted down from the ufo. 

I tried making the particle system slowly descend from the source of the ufo, although had difficulty in doing so. I decided to make this easier I could just edit it in Nuke afterwards. 

After rendering out my coreBeam the first time, I had a lot of trouble trying to get it to actually move the way it was doing so in the scene. The bulkiness of the scene was getting too much and the computer couldn't deal with it - so I just decided to playblast until I could get something useful and interesting.  After it's first render, the coreBeam was flicking rather than moving slowly (effects of the turbulence beam) which was annoying the hell out of me because I couldn't actually get it to do it the way I wanted. I tried using mental ray and maya software to do so, and mental ray was just a joke. It took so so long and the result wasn't even worth it.

I really liked the look of the beam while it was stationary and thought that maybe I could use that without it moving. I revised my work and looked over my past tractor beam and came to the conclusion that I didn't really like how all of the beams were constantly moving - it was almost too much for the scene and too sickening for the viewers eye. That's when I decided to just render out the system at a stationary position and up the particle count even more to make it look like a ray coming down from the ufo. 

I was happy with this by the end and transferred it to Nuke to make sure it fit into the scene correctly. 

With getting my grading rubric back, I've revised a lot of things that I want to improve before I submit my final assessment. 

One thing I have done, is edited and expanded more so on my own references collected throughout the semester. Here you can find the same process taken, as what I did with Chris' references - taking notes and recording them which I can then later refer to when doing my assignment. I thought this would further help me as I found my notes on Chris' extremely helpful when trying to fill the clients brief.

I can see where my work needs improvement now and am going to actively keep trying new approaches to achieve what my clients feedback is. 

What the client said and what I need to improve:

- Rotate vertical beams in one direction.

- Shape of the UFO. Needs to be much more narrower (more cylinder than cone shaped).

- Bulk up the particle count in the pyramid (or cylinder shaped core beam). 

- Make sure the tractor beam only appears when the hatch opens and disappears when it closes. 

Notes from Chris:

Dynamics – Rag Doll Effect

As part of the Assessment for this subject you are required to add a dynamic simulation of a mannequin as it gets lifted into the UFO (aka rag doll effect). This technique will not be taught in class as you are required to create this work as a research task. Once you have found a technique that allows you to dynamically lift the mannequin into the UFO, you will then create a tutorial on your blog that goes through the workflow you used to create the dynamic mannequin. This task is due at the start of your Dynamics and Visual Effects class in Week 13. Work that is handed in after the due date will have marks deducted according to the rubric for Assessment 3 as found in Appendix1.

Bullet. 

Soft Bodies. 

Allows us to deform the shape of a piece of geometry by altering the position of vertices in the geometry.  - Calculates collisions on vertex level.  - Deforms based on collision.  - Flags, table cloths, sheets, pillows. 

Good idea to have shapes visible in the outliner. (faster selection of nodes)

Bullet solver node = controls how the simulation behaves.  Bullet soft body shape = controls how the soft body behaves in the simulation.

Collider shape = make sure it closely matches the shape you're making the collision happen with. 

Soft collision = allows the solver to detect when different parts of itself collide and prevents that geometry from totally collapsing on itself.

Max number interations = make it lower to get a more stable simulation where it won't freak out. 

Up the hz to get more calculations within the simulation - which allows it to be less buggy. 

Create a soft body anchor:  Eg: making a flag - to make parts of the soft body attached to something eg: a pole that the flag holds onto. Use locators and use that as the anchor. 

Aerodynamics = for objects that have air / space trapped inside them. 

Drag = similar to dampening. Pulls the fabric / geo away from the anchor. 

Making a flag:

-Make plane and flag pole.  - Up the subdivision - Create plane as a soft body.  - Create soft body anchor points.  - Put wind magnitude = 15.000. Or 'create new expression'.  bulletSolverShape1.windMagnitude = rand(10,20); - Put drag at 1.000 and lift at 5.000.  - Play with the soft body properties and turn on generate bend constraints and self collisions.  - Up the mass to 30.000.  - In expression, bulletSolverShape1.windDirectionX = noise (time*3)*0.5); - Create expression for drag and possibly lift.  bulletSoftBodyShape1.drag = rand(0.02, 1.5) bulletSoftBodyShape1.lift = rand(2,8);

Open gremlin animBake:

More troubleshooting...

When piecing every thing together, I also found that when saving and re opening files, the ufo was disappearing. I ended up having to save the file as a binary file, otherwise it would just completely disappear from the scene each time I opened it and would therefore have to re import it. :( 

Also..

When I was rendering out my assessment, I had to do each particle system separately and then compile them all together using Nuke. This is why my final .mov looks unfinished and is overlapping at the source of the tractor. 

Before beginning my assignment, I've watched Chris' videos and have taken notes on what he thinks of each reference image and recorded them here.

WHAT THE CLIENT LIKES:

1. Camera angle. 2. More energy in the center illustrated by the 'hot' colour. 3. Colours of these palettes.  4. The non-threatening varience in blue. Darker, less energy on the outer part. 5. Circular form shown looking up into the cone. 6. Cloud like variance / moving like a cloud.

1. People within the 'cone' or 'volume'. 2. Feelings of curiosity, wonder, amazement.

1. Variance within the cones / volumes of light. Lighter towards the edges and further away. More energy towards the middle of the rays. 2. Consistent non-threatening colours paired with the shadowing from the main light source.   3. Sense of motion in all the beams.  4. Feeling of energy conveyed through the light source.  5. Visual representation. Other things / objects floating upwards enhancing this. 

1. Variance in volume (the 4 rays) - Could possibly move?

1. Randomness in shapes coming off.  2. Brightness of light source fading as it gets further away. 

1. In motion - even though the image is still, it still looks like it's moving.  2. Whispy trails coming off the main beams. Trailing behind, as though the main beams are travelling away from the source. Almost cloudy vapor like - drifting off.  3. Colour - the same blue palette, non-threatening.  4. Not enough whispyness, or cone shaped enough.

1. Gentle misty vapor.

1. Spread becomes less and less noticeable until it disappears completely.  2. Opacity and how it fades slowly from the main part to the less effected areas.  3. The texture, ribbons within the moving light. Creates a more dynamic, and interesting flow to the movement.  4. Variety in the shape. 

1. Likes how it is layered. Could use a main cone combined with another particle system that creates the sense of whispyness we're after on a large scale.  2. How it scatters / dissipates.  3. Thicker to thinner. Maybe randomize these shapes coming off using an instancer. 

1. Darker / Lighter tones (higher contrast).  2. Thicker, more variance in the mist.  3. Smoke / mist / vapor.  4. Noise as it comes down away from the source and fades out. Note the emptiness in between the main rays as well.