Closed Thread Icon

Preserved Topic: an edding Pages that link to <a href="https://ozoneasylum.com/backlink?for=11690" title="Pages that link to Preserved Topic: an edding" rel="nofollow" >Preserved Topic: an edding\

 
Author Thread
GRUMBLE
Paranoid (IV) Mad Scientist

From: Omicron Persei 8
Insane since: Oct 2000

posted posted 06-20-2002 16:48

i dont know if people ouside of europe even know what an edding is, but anyways. (it can be very usefull to kill time and live out creativity on posters!)


the original (excuse the bad quality of my webcam)

the fake


enough the beautiful words, lets get to the questions:
*oh, slime! you are here too!*
1. why is my image-map gettin so dimmed down. the red im using is original THE red. ff0000. but on the cylinder it looks kinda darker and also blured.
2. ok, since the back side of the cylinder is so beautiful i want to animate it. it should roll into the spotlight and stop at the position where it is now. so how (mathematically) do you roll an object depending on one variable (clock). and further, how do you stop it smoothly?
3. any further suggestions? especially on the texture of the cap/tip?


edit: who thought i could post without edding? wuhahahaha.

[This message has been edited by GRUMBLE (edited 06-20-2002).]

DL-44
Maniac (V) Inmate

From: under the bed
Insane since: Feb 2000

posted posted 06-20-2002 19:05

Well, I don't know 3D programs, but the simple answer to question #1 is that it would realistically appear slightly dimmed and or blurry because of the shiny surface and the light bouncing off of it.

Looks quite good I must add Grumble - nice work.



Slime
Lunatic (VI) Mad Scientist

From: Massachusetts, USA
Insane since: Mar 2000

posted posted 06-20-2002 19:13

Mmm, if you really want to see why bright red comes out dark, make a sphere with color rgb<1,0,0> (which is equivalent to #FF0000). The problem is that your light source, which I assume has a color of rgb <1,1,1>, isn't hitting the red straight on. You multiply the color of the object times the color of the light source times the cosine of the angle between the normal and the direction the light is coming from, and that's the color that appears there. Since the light source and sphere both have a value of 1 for red, the amount of red is entirely dependent on the angle, and this is almost always less than 1. Probably closer to .5 or so.

To fix this with a normal pigment, you just specify a color higher than rgb<1,0,0>. Such as rgb <1.5,0,0>. Of course, you can't do this with your image (no image file format allows values above 100%). So you might try making your light source brighter. Just remember that 1 is not an upper limit - nor is zero a lower limit (negative light sources can be helpful from time to time).

I'm not sure what would cause blurring, though I don't see any myself.

[edit: by the way, If you want an indoor room lit by the sun, I find that a very high light source value works well, especially if you use radiosity too. I've gone as high as 3.]

...will answer your second question later if no one beats me to it...

[This message has been edited by Slime (edited 06-20-2002).]

tomeaglescz
Paranoid (IV) Inmate

From: Czech Republic via Bristol UK
Insane since: Feb 2002

posted posted 06-20-2002 19:31

Grumble cant you set an expression controller to the pen where say start_Pen_rotate_speed=6 and then every second in the frame that passes until the last key frame is reached reduces the controller value by 1 or something, sorry my max scripting sucks but thats the only way i could think of doing it, you would get a gradual decleration and say the anim is 4 seconds long ya have something like this

frame 0 start_pen_rotate_speed=6

frame 40 speed =4

frame 60 speed =2

frame 80 speed =1

frame 120 speed=0

and apply that to the rotation, its not scientific i know but would give ya a way of doing it with max script, i had a great tutorial in a book by fx&design that covered something like this and it was for a mini submarine and they applied this principle to the hydroplanes where as one movement increased or decreased the speed of another movement on the same object would increase or decrease relatively, if i can find the cdrom and grab the files including the max script i will post the link here..the book was 3d studio max fx and design by creative professionals press publishers coriolis.

if ya know max script which ya probably do well ya should be able to tear through this a lot better than i can explain it


edit just called home its sat on my cd shelf i have the book in the office will bring it in a zip it all and fire it onto my server for ya tomorrow

ah nuts gimme 10 mins i will scan the pages



[This message has been edited by tomeaglescz (edited 06-20-2002).]

GRUMBLE
Paranoid (IV) Mad Scientist

From: Omicron Persei 8
Insane since: Oct 2000

posted posted 06-20-2002 19:42

thanks everyone for the comments!

dl, you are probably right. but its difficult to control these values.

slime, before i try things out, some comments: wont a brighter light source not affect the whole scene to be brighter?
with image_maps i often have the problem that they get a bit transparent automatically. maybe this is causing the dim? (see idcard-sig)
blur? if you've seen the texture image itself, you see that when rendered its much more blurred.
(maybe i should try things out before i complain? )

tom, i understand your theory, but that's not easy to realize in povray because it isnt talking frames but a variable named clock which goes from 0 to 1. but yes, i could trigger the speed somehow using #if (clock>0.6) or so.

tomeaglescz
Paranoid (IV) Inmate

From: Czech Republic via Bristol UK
Insane since: Feb 2002

posted posted 06-20-2002 19:53

oops sorry didnt realise ya were using pov....max would be a great tool for this

but i will bung the pages up now on there server that way ya can maybe apply the same theory in pov

ok heres the zip its 2mb but i thought i'd leave uncompressed the images so text didnt lose definition
http://www.webfx.cz/helpgrumble.zip


it really explains the maths behind how ya can apply a rotation to an objects z axis depending on how far or fast it is travelling in its x axis, just what ya need so if ya can apply it to pov cool, if not hehehehe ya can do it in max straight of the shelf so to speak

[This message has been edited by tomeaglescz (edited 06-20-2002).]

GRUMBLE
Paranoid (IV) Mad Scientist

From: Omicron Persei 8
Insane since: Oct 2000

posted posted 06-20-2002 23:29

thank you tom! maybe i can use it! though it might be difficult to translate that to pov.

meanwhile i figured out how to do a basic roller. the speed is just 2*radius*pi.

and for the smooth stopper i need a sinus or cosinus right?

ahhhh... funny!

edding: uploaded the ani. 400k divX5 video.

[This message has been edited by GRUMBLE (edited 06-21-2002).]

Slime
Lunatic (VI) Mad Scientist

From: Massachusetts, USA
Insane since: Mar 2000

posted posted 06-21-2002 00:21

Sine or cosine might work, but it wouldn't be entirely realistic.

First of all, don't think that the speed of movement of the object is always 2*pi*r units per time unit. That's *only* the speed if the object is rotating at a constant speed of one rotation per time unit.

It's better to say that the position of the object is 2*pi*r times the number of times that the object has rotated. That is,

x = 2*pi*r * R

where x is the position of the object and R is the amount of rotation that the object has gone through.

Now, all you have to do is express x in terms of time. You want to do this in such a way that x decreases less at later periods of time, in order to make it look like it's decelerating.

The most phsyically realistic function for this is x = c*e^(-t) (where c is some constant having something to do with the initial velocity at time t=0, t is the time that has passed, and x is the same thing it was before). This is because the faster it's rolling, the more it will be slowed down. (The speed decreases more if it's faster, and less if it's slower.) I could solve the integral and all, but i'll just tell you that it turns out to be this =)

This can be expressed in POV-Ray like:

#declare T = clock;
#declare X = 5*exp(-T);
#declare R = X/(2*pi*r); // replace 'r' with the radius

object {
marker
//must start positioned along the Z axis so we can rotate it around the Z axis...
rotate -R/pi*180*z // (converted from radians to degrees)
translate x*X
translate y*r // (move it up along the y axis so that it rests on the floor, assuming the floor is at y=0)
}

I may actually be wrong, now that I think about it. It might be a parabolic function rather than exponential. Not sure.

Slime
Lunatic (VI) Mad Scientist

From: Massachusetts, USA
Insane since: Mar 2000

posted posted 06-21-2002 00:25

Oh, I didn't see you had already created the video =) my bad =)

downloading and watching...

Oh! It looks very realistic! Nicely done.

GRUMBLE
Paranoid (IV) Mad Scientist

From: Omicron Persei 8
Insane since: Oct 2000

posted posted 06-21-2002 10:35

thanks for your explanaition!
i will try that as soon as i have time!

what i did so far, the rolling doesnt perfectly fit to the moving. i did something wrong...

but i'll try that with e^x. sounds reasonable.

« BackwardsOnwards »

Show Forum Drop Down Menu