making a jetwall "à la tron"

[J.Santos]

Hi guys!

For my "Flynn's Ride" project I want to make the jetwall generated by any LightCycle worthy of that name. I have the "Technical Effects" from Peter Claes and yes, the 1st project is called "Trail", it could be used...

However, I would like to do it without copying... So I started making some experiments with the... suspense... trail SOP so I could generate some geo and skin + sweep it...

So I made an experiment but things aren't working out the way I expected...

Maybe this isn't the right way...

I know that the curve I'm using to sweep is being copied into the points... Maybe I should try a line copyed onto the points and then fused???

By the way, why is it that when I do a path and make my object follow it, it goes wrong (I mean it doesn't follow the path correctly...) I tried to adjust the Orient Along path and the orient Up Vector without great sucess... However, if I apply the follow path (or use the path object) in a primitive object, it goes correctly down the path...

So... It's something I did wrong, is it the normals direction or what? When I try to fix the pivot I'm unable to use the $CE(x,y,z) variable so I correct it by eye...

Thanks for any tips or paths to follow...

I'll attach my file so you can take a look...

The backWheel is a .bgeo file I generated...

TrailTest.hipncbackWheel.rar

Edited April 28, 2010 by J.Santos

[anim]

i don't know what is the method you want to use, since your file gives me messy geometry

but there are few thinks i can tell you

1. you are animating object along path but inside you want to get its trail using trail SOP

be aware that geometry inside is not yet animated so it'll not give you any trail

2. turn off Auto-Bank Factor parameter on path constraint since it produces bad results on the corners

3. change your Up Vector for path constraint to 0,1,0 because Y-axis is the only logical one since you are animating mainly on XZ plane

to compensate rotation offset you can adjust ry parameter (in your case to -90)

here is example file with example trail using Trail SOP, since I do not know your needs for the trail i keep it very simple with just a minimum node count

(there is a little misalignment of the trail with source path because the wheel is rotating as it turns and since trail source is not in the center of the wheel it offsets from source path)

i have changed the path to wheel geometry so you'll need to correct it otherwise you'll get an error

sorry that I didn't lock it out but since you attached it in previous post i thought it would be more economic this way

TrailTest_fix.hipnc

[Akabane]

Hi! I didn't see your file, but i did a try in making the jetwall tron-like.

There's the file, see if that's what youre searching for, sorry if not.

Peace

EDIT: busted by anim. Anyway, there's a different method i think, you're using trail sop, i'm using particles to generate the curves. Trail maybe can be more optimized, but hey...

wall01.hipnc

Edited April 28, 2010 by Akabane

[J.Santos]

Hi and thanks anim and Akabane!

I'll look into your examples!

anim:

Yeap, it was giving messy geometry... I think that since I wanted to sweep to that trail a square, and it was being applied to the points everything was a bit chaotic... I left it on on purpose so I could get feedback, like yours

I was aware of fact 1, that's why I used an object merge. But since I don't know another way (except particles) to produce the result I wanted, I decided to give it a go. Is a trail a bad option, because of these "extra" steps? (I mean, because you animate in a place and them import back the animation and so on...)

Thanks for all your help, both of you!!!

[anim]

trail sop is the easiest method and there are no extra steps compared to particle method (actually a lot less)

but you do not get all the flexibility particle method gives you so it's up to you

i personally prefer particle method for this type of things since then i can have lot of attributes from particle sim used for shading especially "life"

[J.Santos]

Thanks for the input anim!

I was so close and yet so far... I just saw your file and it's all so clear! Thank you!

You may be right about particles... I'll follow this route a bit more, but I'm already thinking that I might get into a corner... The trail tends to go "after" the bike, making like an accordeon...

With particles you can generate them in point A and go all the way to point z and if they have a long life expectancy they'll still be at point A.

Akabane:

Your example will take a bit more to "digest" but works like a charm... probably will go for something similar, like I posted above!

Thanks for your files!

[Lyonz]

trail sop is the easiest method and there are no extra steps compared to particle method (actually a lot less)

but you do not get all the flexibility particle method gives you so it's up to you

i personally prefer particle method for this type of things since then i can have lot of attributes from particle sim used for shading especially "life"

Im curious, how would you fix up a particle method for the wall ? as i have tried by using a lineage sop in pops to add the lines together by there origins but the problem comes when skinning and making uvs for it, they turn wavey compared with the other method ?

[Lyonz]

The file, its annoying if you press accept you carnt put the file on the same post by using the edit....

Skin_&_Displacement.hipnc

[anim]

you have wrong number of particles born at each frame and therefore they are not distributed evenly and produce those waves

you need to birth by impulse and set it to $NPT

here is the file with following changes:

impulse birth $NPT

origin is calculated directly by source pop

sorting by y is moved to source line instead of resulting particles, allowing the bike to do various crazy flips and so on

deleting the particle primitive is done in add sop

uvs are build according to arclength so they will not average in u with increasing segments when bike is not moving but particles are born

uvs are scaled by trail length so they are not stretching as the trail gets longer

and maybe other changes I don't remember

Skin_&_Displacement_fix.hipnc

[Lyonz]

you have wrong number of particles born at each frame and therefore they are not distributed evenly and produce those waves

you need to birth by impulse and set it to $NPT

here is the file with following changes:

impulse birth $NPT

origin is calculated directly by source pop

sorting by y is moved to source line instead of resulting particles, allowing the bike to do various crazy flips and so on

deleting the particle primitive is done in add sop

uvs are build according to arclength so they will not average in u with increasing segments when bike is not moving but particles are born

uvs are scaled by trail length so they are not stretching as the trail gets longer

and maybe other changes I don't remember

Thanks for the reply anim think this answers alot of questions for people.

From this how could the custom glass shader be applied, so that the displacement only effects the front of the trail ? As atm it is a standard glass shader has be edited by replacing the displacement with a displace along normals vop but wondering how to control the falloff, Could this be done with a ramp parameter ? curious on the setup?

Thanks In Advance

[anim]

i have added falloff control based on trail age at SOP level

it can be tweaked by ramp and Max Age parameter

this will reate falloff attribute which is used in customized glass shader (just to multiply displace scale)

(I have turned off True Displacements to make it quicker, turn it on if you need to)

ps: noise on glass shader is driven by surface position which may be ok, but if you want to use UVs as the coordinates for noise, you need to alter displacement shader that way

Skin_&_Displacement_fix2.hipnc

[Lyonz]

i have added falloff control based on trail age at SOP level

it can be tweaked by ramp and Max Age parameter

this will reate falloff attribute which is used in customized glass shader (just to multiply displace scale)

(I have turned off True Displacements to make it quicker, turn it on if you need to)

ps: noise on glass shader is driven by surface position which may be ok, but if you want to use UVs as the coordinates for noise, you need to alter displacement shader that way

Late reply i know thanks for the advice, I understand most of the network but curious why two converts have been used before the Vop sop ? I did notice when rendering sharp edges maybe this is the reason.

[anim]

just to make NURBS smooth with the order of 4

one convert or basis sop didn't do the trick, but two were ok (to poly through hulls and back to NURBS with higher order)

[Lyonz]

just to make NURBS smooth with the order of 4

one convert or basis sop didn't do the trick, but two were ok (to poly through hulls and back to NURBS with higher order)

Was wondering as i put down a subdivide and worked the same as tried to rebuild what you had done, however getting an issue with the Vop sop not putting the ramp the correct way wondering if you had problems ?

Lyonz

[anim]

if you have clean NURBS surface you don't need subdivide

can you be more specific about ramp problem?

[Lyonz]

if you have clean NURBS surface you don't need subdivide

can you be more specific about ramp problem?

Sorted the problem with the vop sop merged the scene into a new file not sure what was up but working now after trail and error. Essential im trying to create the wave displacement on the tron trail now, thats why im thinking if the trail was subdivided it could be used to deform from the 4 starting points as they remain behind the travelling object, like a ripple.

[anim]

in my file there are correct UVs which start at 0,0 and grow to traillength,1 so UVs can be used to drive any kind of deformation along with the ramp which controls intensity of the deformation

the displacement is done that way so you can use the same technique for ripple if you want or whatever, either in vop SOP or displacement shader