Two more tutorials (External)

If you haven't seen the short The Monk & The Monkey, by Brendan Carroll and Francesco Giroldini you should really take the time to watch it. It's only 4:27 long.

The Monk & The Monkey from Brendan Carroll on Vimeo.

Francesco Giroldini has created and posted two nCloth tutorials that I think anyone who's interested in cloth simulation should watch, even if they consider themselves to be a complete novice. They're both very easy to understand and you'll learn a lot.

The first is Joint Driven Cloth Simulation. This 29:32 length tutorial "shows you how to easily use joints, blendshapes or any other kind of deformer to control your cloth simulation". -145 MB .mov file

The second tutorial is his Intro to nCloth. At 47:42, it goes over a lot of topics that will gently ease people into cloth simulation using Maya's nCloth. - 212 MB .mov file

Go to the tutorial page:

Note: You will have to download the tutorials.

This guy seems pretty cool. You should look over his blog in general.

~Melissa

Demo reel on DVD - DON'T say no!

I have no idea how I ended up here this morning, but I ran across an interesting video on Vimeo entitled Say NO to DVD Demo Reels.

Say NO to DVD Demo Reels! from Nick Campbell on Vimeo.

Because I know 99% of you won't take the time to watch the 6:11 video, I'll paraphrase his argument. "Your DVD will end up in the trash because potential employers don't want to waste the time it takes to put your DVD in and watch it. Get a website, put it on your business card, and hand these out instead. Direct all of your traffic to your site."

While he makes two arguments that I agree with, I have to call foul on everything else.

1) I'll mirror my comment to his video here by saying I agree that you need a website, or at least a blog that showcases your thought processes, demo reel and other works. I'm sure a lot of larger studio's don't have the time to sit around and read someone's blog and their thoughts as to why they think Cubism is teh sexorz, but I think a lot of employers would like to get into the head of a potential hire.

2) I also agree that DVD's are a waste of money, time and resources. So is paper, but I don't see us getting rid of that any time soon.

When it boils down to it, the following is always true: Luck favors the prepared.

Case in point, I submitted soft copies of my demo reel, breakdowns, resume and cover letter to about a zillion studios. I found out later that one of those companies would also like a physical reel on DVD. Whoops. I didn't have one so I had to rush and make it (not that it takes that long, but still). There are a LOT of companies that still require a physical reel and just because there's a chance they'll throw it away without watching doesn't mean you shouldn't have one. Who cares what they do with it?! Your future is worth more than that $0.50 cent disc.

As a student, I think you should have an up-to-date reel online at all times. You should be working on something often enough so you're able to update and improve your reel once every three or four months. Put it on your blog, Vimeo, YouTube, pimp it out on CGTalk. It's all free and you open yourself up to great feedback. Don't be silly and make an obscene amount of DVD reels because whenever you update your reel, those remaining DVD's become useless. You should, however, have SOME copies of your reel on DVD (my modest range is 5-7) and hard copies of your resume, breakdowns, art work and what have you at the ready in case you are required to mail these things in, or you run into someone who wants them.

One thing I do promote having a good amount of are business cards, but be smart about it. Don't print off 10,000. Also, only put information on these cards that isn't likely to change in the near future.

Name
Specialty/Position
Website/Blog link
E-mail address (yourname@mail.com, nothing crazy like HotMama69SweetCheeks)
Telephone number

Make sure the back of the business card is of a light color and unmatted so the other person can jot down other information if they need/want to.

~Melissa

Simulating Cloth in Sections - mini tutorial

A little while ago, I was handed a shot that another TD was having trouble with. In this shot, a character does an action, bends down, comes up incredibly fast then does another action. Rather than key a zillion attributes and force the cloth to work from beginning to end, I decided that I'd make the cloth work from -50 to 119 (where the character was totally out of frame), cache, then I'd sim from 120 - 195 (the end of the shot) and cache this section. After that, I'd combine the two caches.

When it comes to problem shots, simulating cloth in sections can save you a lot of time and hassle. In this post, I'll walk you through that process.

((This tutorial is meant for those who are somewhat familiar with cloth simulation.))

Prep Work

i. Create three folders where ever you like, for example your Desktop. Name them Cache 1, Cache 2 and Final Cache.

ii. Set up the cloth in your scene and save. Make a copy of this scene and rename the copy SceneName_Clean, or something to that effect to let you know that this is your fresh, untouched scene.

Caching

1. Simulate your cloth. Get it as good as you possibly can for as long as you can. Make a note of the frame where the simulation starts to fail. For my shot, the cloth started to break at frame 140.

2. Choose your nDynamics menuset and go to nCache > Create New Cache > click the option box. Here, set your Cache Directory to the Cache 1 folder you created.

3. Name this cache Cache 1. For the file distribution option, I like to choose the one file per frameoption. This way, should you ever need to replace certain frames, you can.

4. For Cache time range, select Start/End and set your start frame to your very first frame, mine was -50, and your end frame will be about 20 frames beyond where you really want to end. For example, I knew I wanted my first cache to end at frame 119, so to give myself a buffer and set this cache to end at 139. You'll see why in a moment.

5. SAVE!!

6. The fun part. You can do several things beside this, and I admit this is the most drastic route, but it's the cleanest. Delete all of the animation on your character up to where you wanted your first cache to end. It's a good idea to give yourself a 20 frame buffer. In other words, if you wanted your first cache to end at frame 119, then delete the animation up to frame 99. Note: Sometimes this isn't possible because the animation before that point is what causes your cloth to break.

7. Duplicate that key and set it 20 frames prior. In other words, frames 99 and 79 are identical for me.

8. Go 40 frames back and put your character in his/her "T" pose. Set a key. Go forward 20 frames, make sure your character is still in his/her "T" pose. Set a key.

9. Simulate the cloth for this half. Note: In your Nucleus, under Time Attributes, you can set your Start Frame to your first "T" pose frame for the time being.

10. Go to nCache > Create New Cache > click the option box. Here, set your Cache Directory to the Cache 2 folder you created.

11. Name this cache Cache 2. For Cache time range, select Start/End and set your start frame 20 frames before you want the second cache to start. I wanted my second cache to start at frame 120, so I gave myself a 20 frame buffer and started the cache at frame 100.

12. After your cloth has finished caching, close this file and open the SceneName_Clean file you made. In this file all of your animation is still in tact.

13. Select your cloth and go to the nDynamics menuset, nCache > Attach Existing Cache File... Navigate to your Cache 1 folder, select the .xlm file and press OK.

14. Repeat step 13, this time navigating to your Cache 2 folder.

15. With your cloth selected, go to Window > Animation Editors > Trax Editor. Hit the "f" key for frame everything in the window. You'll see your two caches in these blocks. As you can see, the extra frames you cached at the end of Cache 1 and the beginning of Cache 2 overlap. While they overlap, both caches are active and are blending together. In the case of this example, I dragged the top number (don't drag the bottom number!) forward until my 20 buffer frames overlapped.

Why would you want the caches to overlap here?

Sometimes you'll want this so the cloth doesn't make a visible snap when it transitions from one cache to the other. You can further finesse this by selecting your cloth, going to the attribute editor and clicking on the Cache Blend tab. You can key the first cache from 1 to 0 during the overlapped time, and the second cache from 0 to 1. This way the cloth is slowly segueing from the first to second cache.

In my case, I didn't really need the caches to overlap or blend because the switch was happening off camera, but it's not a bad habit to get into. However, if you want to snap from one cache to the next, all you need to do it drag the end of the Cache 1 track back 20 frames, and the beginning of the Cache 2 track forward 20 frames. (or whatever you set your buffer to be). Be sure you're clicking and dragging on the top number and not the bottom number!

Also, be sure not to have any gaps between your two trax, other wise your cloth will more than likely pop/break for those frames.

16. The two working cache halves will now drive your cloth object [more or less] seamlessly. The only thing left to do now is to combine these caches to make one working cache. Go to nCache > Merge Caches > click the option box.

17. Set the Cache Directory to the Final Cache folder you created.

18. Name this cache Final Cache. For the Cache time range, select Start/End and set your start frame to the very first frame in the shot, the final frame to the very last frame of your shot. Note: Make sure in your nucleus, your Start Frame is the very first frame in your scene. Technically you don't 'have' to do this, since your first cache is active from the first frame, but it's good to put the start frame back to where it initially was in case you have to do more cloth simulation.

19. That's it! You can go to nCache > Delete Cache > click the option box and delete the first two caches if you like (select keep the files just to be safe). You can also select your cloth object, go to the nClothShape tab in your Attribute editor and deselect Enable so the cloth isn't calculating in the background.

The nice thing about this is, you can combine more than two caches. My maximum so far has been 7 for a difficult :35 second shot and it works really well. You can also cache again and merge your combined cache with this new cache. I don't really advise going too insane with this, but the sky's the limit.

Generally, if you're having to combine caches a lot of times on a shot that isn't very complex, you should revisit your sim and try to make it behave better for a longer span of time.

--

This may sound like cheating but let me make one thing perfectly clear: it is. Don't feel bad, we're all cheating every second we choose to work with animation. These characters and effects and lights and worlds don't exist. We're all faking reality. You should try to make things work the 'right' way, whatever that is, but at the end of the day, you need to do what you have to do in order to get something to work.

I plan to make a video demo on this at some point in time, and type this up and make it available as a PDF, but in the meantime take this for what it's worth.

I hope it helps.

~Melissa

A little break

So this has nothing to do with anything, but c'mon! How awesome is that?!

XD!

~Melissa

Video Tutorial - simming thick cloth

I guess this video that I made a few months ago is relevant.

On a project I'm working on called Rock in the Road, the characters were all modeled with REALLY thick cloth and many of them have many (many, many, many) layers of cloth. This cloth has some pretty intense texture too. Obviously, with all of this going on, we couldn't just select the modeled cloth, apply nCloth and have a nice day. So people ask how did we sim the cloth when it was so thick. Didn't that take forever? Well, it didn't take forever. In fact, the process is pretty quick.

Wow, I said cloth 6 times in the paragraph.

We used Wrap Deformers.

Our basic workflow.

1. Build a proxy mesh based on the modeled cloth object(s). To do this simply, you could duplicate the outer faces of the cloth object, simplify the topology by deleting extra edges until you have something reasonable then delete the history of this simplified mesh. This is your proxy mesh.

2. Make the proxy mesh nCloth and sim this.

3. CACHE for the love of Pete! (I'll get into that later, this step isn't in the video.)

4. Select the hi-res cloth object, select the proxy mesh and in your Animation menuset, go to Create Deformer > Wrap. You may need to adjust the Max Distance setting of the wrap deformer, but there you go. The hi res mesh (your thick cloth) now acts exactly like the proxy mesh (the thin, one sided cloth you made in step 1).

You may need to adjust your volume for this video. I hope you enjoy this full screen and in HD:


~Melissa

Demo Reel Review

I went on and on a couple of months ago about demo reel creation I even got a nice response, which is a shock since I didn't (and still don't) think more than two people will ever read this blog. Anyway, before I graduated this December (yay!) I had to have a portfolio review with three of my teachers. Talk about nerve wracking! You start looking at your demo reel and you notice all of these things you never noticed before, and you go insane trying to tweak them.

I, for one, think that's awesome. I spent a lot of time taking things out, reworking things, I even redid a couple of the shots trying to get everything perfect.

Of course, nothing ever is perfect and my teachers had quite a few things to say about my reel. It needs more diversification, so far I only have one project in there, even though I've worked on more. There are some unnecessary shots, and some that either need to be cleaned up a lot or dumped. I also heard that I should include some of my scripting work/tools I've developed using MEL. Learn Python, etc.

It was great hearing what they had to say. While I'm proud of my work (I wouldn't have put it up if I weren't), I can see where I really need to do better. It's great to get a fresh pair of eyes to look over what you've done. Sometimes you've fallen in love with a piece that you seriously need to divorce. I've sent out my demo reel, because why not, but I really hope I can have some amazing new things in the next couple of months.

~Melissa

Detect N-gons in Maya using MEL

I learned something really interesting today and I just had to share it.

I'm not much of an organic modeler, but sometimes I have to get my hands dirty. Someone asked something that I've been wondering for a while. Is there a way to select ngons, or polygons with more than 5 or more sides. Someone else responded with the following MEL script:

polySelectConstraint -mode 3 -type 0x0008 -size 3;
polySelectConstraint -disable;

After selecting your mesh, run this. All of your 5+ sided faces will be selected. Now you can do with them what you will...

If you're new to MEL like I am, this can look like a string of nonsense, but it's really pretty simple. Let's take a second and dissect this:

polySelectConstraint - "By using the polySelectConstraint command you can select connected components, components within a shell, or components on open borders." -source

-mode 3 - Mode 3 is "all and next", in other words it selects everything meeting the requirement of the constraint. But what is the requirement of the constraint? That's defined next.

-type 0x0008 - type 0x0008 is face. So the type of your selection will be a face, not an edge or vertex.

-size 3 - size 3 is nsided. (fyi, 0 is off, 1 is triangles and 2 is quads)

So the first line of this simple script is plainly read as "Create a polySelectConstraint for every face that is nsided". The second line is read, "disable the constraint". Once you have your selection, there's no need to keep it enabled.

Pretty useful, huh?

~Melissa

Art Advice for Auditory Learners and the People Who Have to Deal with Them

I'm an auditory learner.

I've always needed things explained to me before I really understood them. I got a B in Physics without cracking open the book the entire semester to study. How? People explained the proper steps to figuring out a certain type of problem and I remembered. Elementary, Middle, and High School was simple for me. Teachers explained, I did the work, I got an A.

This all changed when I got to college and decided I wanted to be an animator.

The teachers were all visual, and so were 98% of the students. If I had a question, the teacher would butcher their attempt at an answer before they just gave up, went to the board and drew something. Everyone else but me would understand what the heck was going on. Not only that, but I was often ridiculed or chastised for needing steps to be verbalized in a clear manner.

"In the work place, no one is going to sit you down and hold your hand through this process."
"Figure it out for yourself."
"Look it up."

I don't know why it's so difficult for many artists to understand this concept:
Verbalizing a basic process is not the same as giving someone the answer!

Example:
2n + 7 = 21; solve for n.

How do I figure out what n is is not the same as asking "what is n"? When responding, don't treat these questions as if they are the same!

The first is a valid question to someone who has never come across a problem like this before. Some people can pull up Google, read up on basic algebra and boom, they know how to do everything. For others like me, you'll kind of get it, but you won't be sure until you can discuss it with someone. And don't treat us like we're idiots. A lot of people ask before reading, but there are people like me who pour over every available resource before asking a question.

I realize art can be a difficult thing to explain with words.

It's really something that needs to be demonstrated. However, I believe that art is the purest form of communication. Anything that is communicated from one person to another needs to be able to BE communicated from one person to another. Even great artist were taught the basics before they discovered their own personal style and developed their personal techniques.

To people like me: It's going to be rough, get over it now and save yourself the pain. You may never be able to sit down and brute force your way through something complicated and that's ok. Knowing how you learn is half the battle. There are resources available to you to make learning easier. I, for one, love forums. It's one of the closest ways you can get to speaking with someone. People explain things differently in forums than they would if they were writing a book. Read posts, watch video tutorials, recreate tests (I find this helps a LOT because I speak to myself when doing this) and things will start to fall into place. This process can be slow, so be patient.

Don't be put off by people who don't like to share or people who look down on your for needing things to be verbalized. Remember how this feels, and when you acquire the knowledge you're seeking, don't be afraid to share your basic workflow. Showing someone how to approach something is not the same as showing someone how to do something. (I swear that makes sense in my head, I hope it makes sense to you, heh).

Oh, and n = 7 by the way ^_^;

~Melissa

Second Nuke Test

So I realize up-resing your fluid tests when you're 3% sure of what you're doing is a big no-no, but I was just so curious.

http://www.youtube.com/watch?v=CcJU5-v3A_M

This is my second nuke test, done using Maya Fluids in Maya 2011
Container size - 50 90 50
Resolution - 200 360 200
Playblast time - ~4.5 Hours (174 frames)

I did this before I got a lot of really good advice from the people in the Ka-boom thread in the CGTalk Forums.

By the way, this thread is AWESOME. There's a lot of great information and people willing to help you improve. I really suggest reading every post, doing a few tests, looking at the sample file hidden in the thread, and asking for ways to improve once you've got a basic understanding of what you're doing.

Anyway, I've gone back to testing at a modest resolution (50, 90, 50), the computer I use can handle this really well and I can get tests out pretty quickly. From now on, I think I'll combine three or four tests in one video, and I can talk about them here.

Thanks everyone!

~Melissa

Maya nuke explosion! My first one!!

What have I been up to lately...

I got a freelance job, doing what I did this summer. Animating cubes to music. Sweet.

I also started learning/teaching myself how to create nukes/nuke like explosions in Maya. I'm having a blast ...no pun intended. I think this is the most fun I've had doing something in Maya since I started really getting into Cloth simulation. After a few hours of reading and struggling, I got a presentable result, which I've posted here:

http://www.youtube.com/watch?v=5lkwDjJ9aow

1 emitter, 1 volume axis field. When I get the hang of this, I'll try 2, then 3 and so on until I get the look I want.

I realize this is a WIP but it's WAY smaller than I want it to be. This explosion doesn't FEEL big at all. I think it's going to fast from the middle to the end. I need to figure that out.

Anyway, I'll post more tests as I go. If someone in the know happens across this, PLEASE offer your suggestions and criticisms. I'm just here to learn ^^;

~Melissa