THE FINAL POST

okkkkk so here it is, here is the shot i've worked up over this project. I still think there is obvious room for improvement, but i have learned soo much over this project!

I would still like to address how the petals fall as i think they have the right kind of speed now, but they need to last longer, i dont like it when particles just "disapear" but this is somthin i will be implamenting houdini for.

THINGS THIS PROJECT HAS TAUGHT ME
-i have learned what a voxel is, and by doing this, i have learned what sort of voxel grid resolutions to work with, then step up to render with.

-i have learned about the different render systems such as scanline and raytracing, and why you would prioritize raytracing in voxel fluid rendering as it steps rays right through your fluid to calculate accurate shadow maps based on density

-i have learned that volumetric emmiters work much better with voxel systems rather than the standard omni which is essentially a mathmatical "point" with extended space.

-i have learned all about the fluid shader, and how best to control it for different applications such as fire, smoke or combustion.

- i have learned about the various dynamics (fields) rigid bodies and types of fluids (particle, volumetric and voxel) and their best uses.

-i have learned about ppInstancing in maya and its drawbacks

-i have learned the benefits of using 2D voxel containers with raytracing.

i could go on and on, but basically i have loved this project, and I am more or less pleased with my results. here is a break down of what i think worked well and not so well.

THE GOOD:
-I like the trail from the spell, the body of it atleast.
-i like the look of the flower petals
-i'm sort of hapy with the shockwave

THE BAD (well, not bad, but stuff to improve on)
-the motion of the trail of the spell, more interesting and less uniform
-the flower petals i will need to take into houdini to get more complicated procedural dynamics working upon them
-the shockwave can be improved and i plan to do this soon!

HOW HAVE I INTERGRATED IN A TEAM?
-i have used animation from nick and joe to use as a base for my spell development. without their help, my work would be very boring and linear

-i have used Ollies model, which i rigged and gave to the animators, which feeds back into my first point.

-i have mimicked Stephanie Joy's beautiful matt paintings for my environment creation

---AND i have constantly been asking for feedback from everyone in the group, which they kindly ablidged with and without their feedback again i think my effects would be very bland and boring. so thanks for all of your help guys!

submision to MY group (aka spellbound)

so before i came and showed dan and mike, aswell as the rest of the students on my course, i wanted to pitch this to my group so they had the chance to give last minute feedback.

the feedback i got from this concerned making it a little faster, and have the flower petals fall slower, and die out quicker. I implamented these changes which i will show in the next post.

more workings (11 - 19)

soooo heres a few more videos showing development. I decided to implement a placeholder backdrop so i could start playing with how the spell would stand out from the colours of the plate. i will get around to creating a little environment before presentation time.

I really like how this is going in. I feel that each element of the spell is looking better together than it did before, and many things have changed now.

for instance i have decided to use a particle system to make the wand sparks look a little more like flower petals, as he is quite feminin and vein. these are working ok for a first draft and will be further developed before the last test.



v11



v12



v13

---sorry, didnt conform a few---



v18



v19a



v19b

working up presentation test (1-10)

rather than comenting on every single video of this series i though ti would just talk about what i am, and have been doing. So our presentation is next week, and i believe i now have all of the knowledge i need to get going on implamenting a spell in a scene with a character and an environment (which will come a little later)

so below are the first 10 verisons. I took certain assets from previous tests rather than re-doing everything. i found this efficient, especially for some of the spell containers that hold the "look" which i developed in previous tests.

knowing this means that it will be easy to get good continuity between our shots in the final film!

(the following videos, when watched in order, show the development of this shot. here are 1-10, more will be included in the next post.)

v1


v2


v3


v4


v5


v6


v7


v8


v9


v10

pushing the limits of thickness!

ok so i wanted to see how far i could push the thickness of my fluids. this was following suggestions from my team about getting it more smokey. i didnt just want to immediatly start fine tuning somthing before trying to understand it so i went ahead and did this test to see how solid looking i could get my render out. this in my mind looks like cement!

obviously not what we want for our spell but was a good test and has taught me a lot about the opacity curves and transparency float slider.

developing further again

ok so in this test, i tried making the emmision smaller, but the fluid thicker, with what i think is quite a nice result. Please note that the black patches on the spell are a product of low quality rendering


here i tried to push the movement as much as i physically could, in order to see how much looks good and how much bad. i like the above test but i would like to see a more varied movement still!

longer tests. this one mucked up!

i have decided to start doing some slightly longer tests, firstly to make sure that the fluids im using are consistant enough to be used over long periods of time. seccondly as things like dissapation and diffusion you can only really messure over these sort of time periods. and thirdly i am mimicing this video I found! (shown at the bottom of this post)

as for this spell, i am getting quite happy with the trail. i think it is still a bit tame, but its a great direction and my group seem to like it too. they have offered some very useful feedback as to the shape and motion which i will now be developing up aswell as incorporating a few of my own ideas!

here is that video i mentioned:

some further development

OK, SO I HAVE BEEN DEVELOPING THE FLOW OF THE SPELL FURTHER. i HAVE FOUND THAT USING MOTION AND VOLUME EMMITERS SEEMS TO BE A GOOD AWAY OF GETTING INTERESTING FLOW TO THE SPELL. i STARTED OFF BY JUST USING SPHERES BUT NOW I HAVE STARTED ACTUALLY SIMMING SOME CLOTH FIRST, WHICH I THEN CACHE OUT AND USE AS AN EMMITER ITSELF. THIS SEEMS TO GET SOME REALLY NICE RESULTS SO I WILL CONTINUE TO DEVELOP THIS FURTHER!

my influences - revisited

I have mainly been looking at 3 sorces for my research.
1.) harry potter!



or link
http://www.youtube.com/watch?v=aEYkVedmboQ

harry potter is the most contempory source of inspiration into creating abstract spell effects!

2.) cloudy with a chance of meatballs!



or link
http://www.youtube.com/watch?v=h92oK9u3pzI

watch for the smoke trails left by his contraptions! THIS hehe

3.) this video of effects i found!



or link
http://www.youtube.com/watch?v=7c4WYzr30B0&feature=player_embedded

these are amazing! but i think they were probably cached and rendered over days and weeks, i just dont have the time to spare for this!

I love the second effect though, i believe this would be good for our crude wizzard!

lightning...y?

i thought i would do a little breakdown of the end of the spell at the moment. i am experimenting with the dynamic lightning built into maya, which creates a curve, introduces a noise/fractal into its structure, then applies a seed value on a per frame basis. this allows some variables to change the look and shape of the curve over time. by then extruding a tubular surface along it you can give it substance, and applying a shader with inbuilt incandesense and glow, you can get quite an electrical look.

I'd just like to bring up a point about PRESETS.

I hate PRESETS.

my honest opinion is whats the point. If i can get an amazing looking spell by using other peoples work, i dont want to use it. it would kill me for someone to highlight a preset in my work and i would have to own up. I also think it teaches you nothing to hit a button and click render. no point whatso ever. at the end of the film, autodesk would have to stand up and take a bow, you know.

BUT

i do understnd the value of taking a preset and breaking it down. for this lightning, i applied it, observerd it both in the view port and the hypershade, then looked at the node structure that goes into creating it. its not a special thing that you can only get by clicking the preset button, it can be built from scratch, and thats what I will be doing for the rest of the project. i think this method does teach you somthing because i see no difference in observing a preset (or copying or just clicking it in for that matter) to observing anothers painting if you are a fine artist. you cant just copy it, its plagerism, and the least satisfying feeling in the world. imagine just downloading an image and using it directly in your work.

not good enough.

getting there,, but no cigar!

here is a test that i spent a total of 13 hours on. and guess what, i'm not happy with it.

i like the movement, shown better in the play blast below, but the rendered substance i think could use some work. Again, this was also stated by my team, and I will completely focus on this rendering issue from now on. I think it may have something to do with the fact that im using incandescense values to display my fluids, this makes some quite dynamic looks, but possibly a more smokey fluid rather than a reactive fiery fluid may be best.

heads and tails

this is a test that I did with getting two fluids into a single scene by referencing them, and importing the cache files. This isn't really a test on how good i can get the fluids to look, but of how we can utilize referencing and cache to our advantage over the project.

spell head!

so going on my new method of using different pieces to make up the final spells, i have tried crafting the spell head. I was reasonably pleased with these results, i like the way it moves, but i dont think there is enough substance to the spell. i would like a higher level of detail, with a thicker overall body. This point has also been raised by my group, so i will be concentrating on this for the next few attempts

primary research!!!!!!!!!

ok so today was an early start. I went to a WIP Session at Framestore, in which they went over some of the latest films they have worked on such as harry potter, jonny english 2, captin america and.....the smurfs!

now ive never been a fan of the idea of a 3d/live action  smurf movie. I have seen it to be a kitch, rehash of an idea, but I must confess, the visuals sure look nice!

Also it was watching this wip that I saw a spell! The spell itself was of two pieces, the mussle flash effect at the point of emmision, and the jet/trail. the two were so different from each other, that this has prompted me to think differently about my approach.

in ANALYSIS of my previous attempt (the animated scene embeding a fluid spell within) has led me to the conclusion that I'm doing it soo wrong.

Even the analysis i previously did of my concept work should have given me this idea. Let me explain the rest visualy.

so, incase I havnt been clear enough in my drawings, I will be looking at using different fluid containers for different parts of the spells. This may have been obvious to someone else, but I'm glad the idea has finally made its way around to me!

a maya test in-scene!

ok so this is 4 hours work compared to the houdini 12 hour fail. This is not really what I was going for, so much more development needed with this, but its a start in my further learning and understanding!

Back to the loverly familiar world of Maya Fluids

Ok, so basically, I have analysed how I am going to forefill the role of Effects TD for our film, and I have come to an unfortunate conclusion. It seems that learning houdini in just a few weeks is an unrealistic task, and i think I would do better returning to maya and spending the rest of this term developing the skills I possess in Maya's fluid systems.

That is not to say that I wont be using Houdini anymore. I have learned a VAST ammount about the rigid body networks available in the software, aswell as how adaptable the particle effects are. I will also probably use Houdini for any water effects that appear in our film, as it seems houdini is more or less sut up for this kind of thing from the word go. Its just voxel fluids are soooooooooooooooo in depth that I just cant get my head around how houdini has set them up, which attributes to effect, and which nodes and forces I can use to effect them in the same exact ways That i already know in maya. So this is for now on houdini, I will have a maya test up on here soon! :)

believe it or not, this is 12 hours of work

Ok so I spent practically the whole day on this, and its soo not what I wanted. I have done tons of tutorials in housini, from the side effects website, but there is soo much to take in. Im really thinking of switching back to maya for this project.

anyways heres the non-results

Soooooo Houdini is HARD

wow where to begin... so I have been learning houdini for the last couple of weeks. Ive picked a lot up but it is soooooooooo in depth. I have done many many tests, Apologys that I dont have renders, or screen grabs from most of them, but let me explain why. I have been so focused on getting my head around this program, i havnt thought about anything else for days. And when your working ur constantly CACHEing your files, then deleting the cache once your done. I didnt think to screenshot until after I deleted the cache, so its either wait 3 hours to get a picture to put up on here, or i'll just try my best to explain what Ive been up to. I have a couple of rendered Items which Ill put up at some point

to help prove my point though, here is the node tree for the default fluid emmiter in houdini (its massivly soomed out)

So what am I doing with this test?

I have set up an animation curve, which will carry a sphere along the path of the wand, then jet off in the direction of the spell. This will drive a fluid which will be the essence of our spell. Im very new to this so it will just be an expreriment really!

Houdinis interface

i thoought it would be good to talk a little about houdinis interface. The interface is made up of 7 main areas.
1.) Dual Tool Shelves: Houdini has 2 tool shelves that you can customise with diffeent menu sets, or make your own. this abillity to have two open at once can be useful when you are starting to lay out your main nodes, and can help quickly set up your scene

2.) Network Pannel: This area is where your nodes sit. Every new item you create in houdini has or rather, IS a node. These nodes exist in different levels: Scene Level, Geometry Level, Dynamics Level, Shop Level and Output Level. There are a few more but I havent fully explored them yet.

3.) Peramaters pannel: This is like the attribute editor in Maya. It updates as you select different nodes, however you can PIN it to one node when ever you like to enable you to keep the focus on it whilst editing other nodes in other windows.

4.) Selection Tools: Similar to Maya's tools pallet, the Houdini tools are slightly more context sensitive. For instance when you are in Dynamics Level, the move, scale and rotate tools are greyed out as they cannot be used in this context.

5.) Time Line and Playback Features: These are very much similar to Mayas Playback interface, just some diferent options in different places y'know!

6.) Simulation Options: Here you can select how you would like your simulations to be solved, and how they play back. Basic stuff really. This dosnt effect how your effects actually are simulated, just how you view them as a user.

7.) View Pane: The customary 3D scene view. What can I say, operates more or less the same motion as Maya, however there are far more options for viewport display, featuring the option to turn all nodes to bounding boxes for low-cpu display, or hidden line which can be really useful when needing to select fine detailed objects

Volume Fluids

In houdini, there are 3 different types of volume fluids you can work with. These are:

Pyro, Flame, Smoke and Liquid. To help me understand the kinds of things these could be used for, I decided to simply set up a scene with one of each container, then source from 4 identical spheres and analyse the results. Here is a screen grab of this:



as you can see, they all give very different effects, considering that they are all set to default values. It seems that:

Pyro: gives a combusterbal smoke cloud, with the abillity to easily add fuel in areas to increase the reaction.

Flame: will actually burn and deform an object it is applied to. The reaction given off is a mixture of heat and density, driven by the rate of decay from the geometry.

Smoke: this seems to simply emit a density from the selected geometry, with options to effect bouency ect. Fuel and temperature are possible, but this is better for cloudy, low reaction fluids

Liquid: this is by far the most different from the group. It seems to use the voxels to generate a surface based on the density. Obviously impracticle for any smoke or gassy fluids, but very useful in simulating the flow of water.