XM Magdalena 3D print, GGeorgy (3D)
LC #42 Pipers Alley

View Full Version : FXWars! Avalanche!: CoryC

09-02-2004, 04:01 PM
I think I'll have some time to give this a shot. A tough challenge this time! We'll see how it goes.

09-03-2004, 06:53 AM
Good luck, Corey. I'll be following you.

09-03-2004, 09:53 PM
Good luck, Corey. I'll be following you.
Thanks. I hope not like a stalker:scream:

I'm not sure when I'll start having things to put up but the tests so far are going good. Maybe I'll get an animatic up in the next few days.

Cory(no E)

09-05-2004, 05:33 AM
I started working on this today and here is my first test. This is only a test scene and nothing in it will be used in the final. Mostly I was working on getting an idea of how I'll put things together and see where I might have problems. The leading edge of the snow looks wierd at the end but I think that can be taken care of by adding an emitter up front and some turbulant wind to roll the front.

http://beta.usavgroup.com/cgtalk/aval/avdroptest.jpg (http://beta.usavgroup.com/cgtalk/aval/AVdroptest.mov)

http://beta.usavgroup.com/cgtalk/aval/AVdroptest.mov (http://beta.usavgroup.com/cgtalk/aval/AVdroptest.mov) (430k)

09-05-2004, 06:12 AM
Hah! That looks really cool already!

I forget, are you using Maya? Or are you a Lightwaver?

09-05-2004, 11:44 AM
Looking good there CoryC

nice impact! :bounce:

09-05-2004, 04:35 PM
Hah! That looks really cool already!

I forget, are you using Maya? Or are you a Lightwaver?
I am using Lightwave for this challenge. I also use XSI but I just started and I am not up to speed enough with it yet.


09-05-2004, 08:33 PM
Looking good!

09-06-2004, 06:01 PM
very cool!!

the destruction effect is very well done:thumbsup:

09-10-2004, 06:04 PM
Nice start:)

09-15-2004, 12:28 AM
I've started working on the beginning. Here is how the initial cracking of the snow to start the avalanche is going to look. It is still a WIP. Need to tune it some more.

http://beta.usavgroup.com/cgtalk/aval/cracking.jpg (http://beta.usavgroup.com/cgtalk/aval/cracking.mov)

http://beta.usavgroup.com/cgtalk/aval/cracking.mov (475K)

09-15-2004, 01:04 AM
wow thats awesome!

09-15-2004, 02:05 AM
very impressive :thumbsup:

my guess is its geometry with normal displacement revealed by a wipe

but I`m not very confident about that

very nice indeed

09-15-2004, 02:17 AM
my guess is its geometry with normal displacement revealed by a wipe
Thanks but your guess is wrong;) It is all geometry and dynamics. No wipes or maps of any kind.

09-15-2004, 04:26 AM
Playing around with it a little more, just trying a few things out. No, the sun doesn't match the shadows.

http://beta.usavgroup.com/cgtalk/aval/cracking2.mov (http://beta.usavgroup.com/cgtalk/aval/cracking2.mov)

09-15-2004, 04:30 AM
Wow that's really showing promise. Dynamics and geometry only? Wow! How??

Triple G
09-15-2004, 04:31 AM
Damn....I was going to guess it was done by unwelding points and setting up a morph map for the cracks, then using Prem's Relativity Morph to reveal the morph. Hmm....now I'm kinda wishing I had upgraded to LW8...

*wonders how to fake something like this with 7.5* :wip:

09-15-2004, 09:47 PM
Here is another update to the first scene.


09-15-2004, 11:17 PM
lookin good... someone did there homework:)

09-16-2004, 01:53 AM
Awesome Cracking effect Cory! I can tell i am working on one cracking effect right one...not for an avalanche though.

So did you "Split Poly" all the way through that plane for the main crack? I can't figure how you did the multiple shatters..

Promising WOrk!

paul k.
09-16-2004, 06:32 PM
So Cory this may be a bit passe, but..... How ya doin it? It looks like you modleled some big chunks in Modeler, and then maybe used a plug for the small stuft......then trigger it all by an event?

I would really love to know the particulars as I am doing a series of animations involving cracking eggs.

You use LW right?

09-16-2004, 07:23 PM
I'm surprised by all the interest in the crack effect. It is basically a lot of boring work in modeler and then some dynamics.

First I built the crack by starting with a 200 foot long box divided into a lot of sections. Then I pulled points around until it looks descent. Next I added some points and moved tose around to break up any looks of a bunch of squares.

Above the crack there is a secondary breaking of some hanging areas. I made polys of those areas and then one by one, used crackit on them. I get rid of all but the top polys, merge, unweld, and extrude for less poly count.

There is an area below the crack in the shape of a large half circle that will break away. I built big poly chunks, maybe 50 or so, and then did the same thing with crackit to them. C4 will not work as well because it makes too many trianles and the poly count is insane. All these sections were reduced down to 3 layers.

The rest of the static landscape was built, including a layer below the break-away sections.

For dynamics, a couple collision events control the main crack. Another fires off the extra parts above the crack. and a few more break up the parts below.

09-16-2004, 08:40 PM
again, VERY NICE!

09-16-2004, 11:39 PM
Starting on the next shot (with the tail of shot 1). This is the same set as the first shot. Particle effects will be added to this next.

http://beta.usavgroup.com/cgtalk/aval/shot02.jpg (http://beta.usavgroup.com/cgtalk/aval/shot02.mov)

http://beta.usavgroup.com/cgtalk/aval/shot02.mov (411K)

Triple G
09-17-2004, 12:23 AM
Damn dude, you are definitely bringin' it for this competition... :buttrock:

09-17-2004, 12:36 AM
Sensational! I too would love to know how the heck you have done this.

09-18-2004, 12:34 AM
Another test of moving snow to see if I can pull this off:) This won't be part of the final. Lots of comped layers (all dynamics though) and I cheated with a little bit of avalanche audio.

http://beta.usavgroup.com/cgtalk/aval/avaltest2.jpg (http://beta.usavgroup.com/cgtalk/aval/avaltest2.mov)

http://beta.usavgroup.com/cgtalk/aval/avaltest2.mov (285K)

the sparky
09-21-2004, 07:32 AM
that cracking effect looks great!

In that latest test it looks a bit weird how the snow seems to appear out of nowhere.

09-21-2004, 01:07 PM
Yea, that last test doesn't look right. Trying to work out how to make it look like snow from the ground and not just moving above it like a crazy fog bank.

09-21-2004, 11:42 PM
Another test. making progress...

http://beta.usavgroup.com/cgtalk/aval/hilltest.jpg (http://beta.usavgroup.com/cgtalk/aval/hilltest.mov)

http://beta.usavgroup.com/cgtalk/aval/hilltest.mov (874K)

09-22-2004, 06:53 AM
Waow, in many ways that's the best test I've seen in this FXBattle, so far.


- In a still picture, you've really nailed the look of a real avalanche
- Playing the movie though, the puffy clouds still look a bit like rising 2D sprites, and not a volumetric cloud
- There are some issues with bouncing stones inside the cloud (they seem to flicker a little bit)
- It looks like you applied some kind of physics, so when the avalanche particle drop and land hard, they release bigger puffs? That looks cool.
- But most of all, this looks more like a snow avalanche than anything else we've seen here, because of those tumbling snow elements in the front of the avalanche, they're awesome! :thumbsup:

- Jonas

09-22-2004, 01:41 PM
Thanks Jonas,

on the comments:

1 - The still is just a frame of the video
2 - They are volume particles but because it is straight on with no shadows from the avalanche to the ground, it looks 2D yet. There is no easy and reliable way to get just a shadow pass of the particles. I think I have a way of doing it but that would require another pass and some playing around.
4 - The particles are released on the speed of the emitter so when it lands it slows down and more particles are released plus the bounce sends them floating higher for that look.
5 - thanks. I am glad to have have figured out how to do this right. I think the snow debris up front still rolls too far too fast yet but that is just a matter of increasing friction. Now to get those stones to look right and to get good shadowing.

09-22-2004, 04:51 PM
Here is that same test with the shadows added.

http://beta.usavgroup.com/cgtalk/aval/hilltest2.mov (873K)

09-22-2004, 08:06 PM
just wondering how long the render time is for most of the video/

09-22-2004, 08:18 PM
It is built in layers but I would guess overall there is an average of 30 seconds per frame or render time or 3 hours for the previous 2 clips.

09-22-2004, 09:06 PM
Very nice work!!!

But you should put on some turbulence on those hypervoxels. Now when they rising towards the sky they dont move and look unreal.

09-22-2004, 09:45 PM
yes I should. Thanks for reminding me. I noticed that yesterday but after staring at it for so long I forgot.

09-22-2004, 10:39 PM
Here you go Ztreem - http://beta.usavgroup.com/cgtalk/aval/hilltest3.mov (873K)

09-22-2004, 10:48 PM
But you should put on some turbulence on those hypervoxels. Now when they rising towards the sky they dont move and look unreal.That's exactly what I meant when I said they look 2D - you just said it better than I did. :)

And it looks MUCH better in the newest render, Cory!


09-23-2004, 03:14 AM
a one man effects team. a very impressive display, very realistic looking ( i think...no snow where i am) should be proud of what you have so far. definantly watching this one.

best of luck, doubt you'll need it.

the sparky
09-23-2004, 03:36 AM
that's looking great. The only problem I have with it is that it doesn't look like the clouds are settling back down. Right now it kind of looks like foam because the avalanche clouds grow and then stay there.

09-23-2004, 04:12 AM
it doesn't look like the clouds are settling back down. Right now it kind of looks like foam because the avalanche clouds grow and then stay there.
The snow clouds don't settle back down for quite a while (depending on the type of snow). One of my resource videos has an avalanche that has a snow cloud running at least 500 yards and the beginning still has snow that is a good 100 feet tall at least. It kind of just billows around like regular clouds in high speed.

Thanks Stallion and Jonas. We get snow here but no avalanches. I've shoveled enough of it to know I would rather watch it on the computer though.

09-23-2004, 10:18 AM
Yeah! Now it looks more like snow clouds. Very good work! :bounce: :thumbsup:

09-23-2004, 06:31 PM
Wow really coming along, really like the fracture shot...

09-27-2004, 06:57 PM
The snow clouds don't settle back down for quite a while (depending on the type of snow). One of my resource videos has an avalanche that has a snow cloud running at least 500 yards and the beginning still has snow that is a good 100 feet tall at least. It kind of just billows around like regular clouds in high speed.

I had the same question. If that is indeed the case, shouldn't there still be movement in the trailing clouds? It seems like it grows to a point and then freezes so, it seems unnatural.

It looks great BTW.


09-27-2004, 09:13 PM
I'm not sure what you mean. The snow clouds are kicked up and then slow down and slowly (probably a couple of minutes at least) fall back to earth. There is always some turbulance but the snow clouds don't continue to change size very quickly. That is how my last test was. They grow and then slowly shrink with some turbulent movement continuing.

I'd love to work on it more but I have run out of time that I can spend on the challenge so this is as far as I will get.

09-28-2004, 03:09 AM
that's a shame you can't finish... but if you do somehow find the time, maybe if you had the clouds thin out as they age it would look a little more realistic... and as they disapate they should grow as well... i haven't seen you reference image but it just stands to reason that they will grow as they get older not shrink... and they will become more transparent as they grow... my 2 cents anyways.

09-28-2004, 04:27 AM
Here is a good link with a video that might help clear some of the confusion of what a real avalanche like I was working on should look like - http://www.nationalgeographic.com/ngkids/0301/. Remember, those snow clouds are made up of solid snow powder, not water vapor. They grow from the initial blast of wind and energy from hidden snow, ice, and debris then settle down slowly. They wouldn't continue to expand due to gravity and a loss of the energy that created them. they wouldn't thin because they are a solid mass slowly falling at a constant state. Physics alone say that it would take almost 4 seconds for the particles to return to earth from 200 feet and that doesn't factor in the wind resistance that will keep them up a lot longer.

09-28-2004, 04:48 AM
Hey Cory, it's a shame you won't be able to finish this but good work on getting as far as you did! Hope you enter again in the next FX Challenge :thumbsup:

09-28-2004, 08:31 AM
check out the video on this sight http://juneaualaska.com/stories/avalanche/ if you think about it, snow is fine particles that are influenced by wind and gravity and such, so there will be drifting and disipation as the avalanche progreses, also, i would imagine that the great amount of friction caused by large amounts of snow moving very quickly would liquify some of the snow causing more of the expansion and disapation that i was talking about. as you will see in that video there are two visable layers in an avalanche there is the thick base full of debris and there is the cloud or "wash" that is carried off the surface and into the air, this cloud acts like any other gas, it drifts and expands and becomes less and less visable as it ages.... anyhoo check out the vid, you will see what i mean

09-28-2004, 09:47 AM
There's been a lot of clever tips on avalanches in the last handful of posts, but IMO they're all microscopic details compared to actually making something that "basically" looks like an avalanche, AND has the right physics to throw things around. When you have THAT, THEN you can start worrying about how snow particles settle.

Anyhow, I think Cory has one of the best-looking avalanches in the challenge so far, gravity or not.

I hope your running out of time doesn't mean you won't submit.

Cheers everybody,
- Jonas

09-28-2004, 04:52 PM
I hope your running out of time doesn't mean you won't submit.
No, I won't be submitting. I don't have all the elements required. I'll be going out of town in a few days so I'll be missing the deadline and the start of the voting as it is.

10-05-2004, 05:44 AM
Hey Cory, seeing as there's an extension will you be doing any more on this one?

10-06-2004, 06:54 AM
Hi Cory,

Excellent work so far! Even if you don't enter it you should still post whatever you finish up with. I'd love to see it.

If I could suggest something, your avalanche seems to be lacking momentum. Unlike dust kicked up in your path, which rises but doesn't move forward, an avalanche has tremendous forward inertia. The bottom snow-slide doesn't always move forward the fastest, sometimes the cloud overshoots itself and then the rest needs to catch up. The cloud itself is moving forward all the time, some slowly, some quickly, but always downhill. Also, the front edge tends to roll backwards onto itself. The forward movement of the snow on the ground causes a slight suction which pulls the snow-cloud in backwards. Imagine a large barrel rolling downhill on top of several smaller tubes. The large barrel would be spinning the opposite direction to it's travel. This is often covered up by spurts of snow jutting forward, but should be underneath.


10-06-2004, 03:18 PM
Hey Fasty and DJ Nicke,

No, unfortunately (or fortunately depending on how you look at it) I have real work I have to get done then I'll be out of town this weekend for some photography. I postponed my trip from last weekend so I would have had some time to work on it more if I knew about the extension but probably not enough to finish anything. It would take me a long time to pull off what DJ Nicke is suggesting. Anyway, I have posted everything that I have done so far.


10-11-2004, 07:52 PM
Very nice indeed now just finish it lol i want to see it already :D

10-12-2004, 06:52 PM

Just a pointer: make it slower, it'll look more realistic

CGTalk Moderation
01-19-2006, 04:00 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.