Archive: Why Does Avs Keep Changing!?!?!??!!?


24th March 2003 04:28 UTC

Why Does Avs Keep Changing!?!?!??!!?
YUS@@


24th March 2003 06:36 UTC

lol :D

Not bad, but the scope movement is slow and the vibration is too small to notice properly.


24th March 2003 08:06 UTC

looks scarily like the beginning to platonic liquid (tff1.0)... :igor:


24th March 2003 08:39 UTC

but can you explain a bit more about the question


24th March 2003 09:47 UTC

quite cool but could be a bit more dynamic.Oh and when i saw the topic i pressed the button with anger and was sure i would post a big reply that you should have searched and forums and it aint nice and blah.You understood that milkdrop is uglier than avs ? *GRIN*


24th March 2003 12:12 UTC

Don't knock teh milkdrop! :D


24th March 2003 12:17 UTC

/me knocks milkdrop

Milkdrop is alright, but AVS has a better feel/control to it IMO.


24th March 2003 12:37 UTC

YUS@@@


24th March 2003 12:50 UTC

Mmm. I liked the first version better. Pretty nice and simpe AVS. Still not true 3D though. :P


24th March 2003 12:54 UTC

I don't think I'll ever make anything 3D, My math skills are somewhat limited.


24th March 2003 13:11 UTC

You're an idiot anyway :p

making 3d is not too hard as it's quite simple to do with x,y,z coordinates but thy have to beĀ“mapped back to 2d, thats what makes it a bit difficult.

Furthermore it's not that easy to make a really interesting looking 3d preset. I get sick on seeing simple rotating cubes, this has been done way too often.

Let's face it we can't all be little UnConed's

;)


25th March 2003 20:21 UTC

If you want to understand 3d/2d changing: Try to d/l Elvis' old Superscopes pack (no. 6 or so). It has an example of it totally explained. Rip that apart and try to understand it - it helps a LOT, at least it did for me. Write every part down and try to follow all variables. It's a lot of work, but well worth it.


28th March 2003 01:28 UTC

If you want to understand 3D avs then read Atero's AVS Primer.


28th March 2003 03:20 UTC

I didn't say I wanted to. But I haven't read AVS Primer since the first one came out, so maybe I will.


28th March 2003 03:41 UTC

His second one hasn't come out yet.

Atero, when will that be done?


5th April 2003 16:56 UTC

Originally posted by ;-c ,rattaplan
If you want to understand 3d/2d changing: Try to d/l Elvis' old Superscopes pack (no. 6 or so). It has an example of it totally explained. Rip that apart and try to understand it - it helps a LOT, at least it did for me. Write every part down and try to follow all variables. It's a lot of work, but well worth it.
actually, DON'T use the El-vis 3D method. REad the Tips and Tricks for better ways.

6th April 2003 14:18 UTC

Originally posted by ;-c ,rattaplan
If you want to understand 3d/2d changing: Try to d/l Elvis' old Superscopes pack (no. 6 or so). It has an example of it totally explained. Rip that apart and try to understand it - it helps a LOT, at least it did for me. Write every part down and try to follow all variables. It's a lot of work, but well worth it.
Igonre this advice. Its wrong. The EL-VIS 3D code may have been used to create some very amazing presets but codewise it is cack. Long winded, inefficient and difficult to read. Try using some much *much* simpler code. Tons of us have written much better 3D code and explained it in the forums, preset comments and such.

If you can't be arsed to search here is how to do it:


per frame:

rx=...;ry=...;rz=...;crx=cos(rx);srx=sin(rx);cry=cos(ry);sry=sin(ry);crz=cos(rz);srz=sin(rz);asp=h/w;

per point:

x1=...;
y1=...;
z1=...;

x2=x1*crz-y1*srz;
y2=x1*srz+y1*crz;
x1=x2*cry+z1*sry;
z2=-x2*sry+z1*cry;
y1=y2*crx-z2*srx;
z1=y2*srx+z2*crx;
z1=1/(dst+z1*scale);
x=x1*z1;
y=y1*z1;
x=x*asp;


x1, y1, z1 set at the top are used just like x and y. dst and scale have to be replaced by constants and these represent the distance into the screen that the origin is and a scaling factor on the z-axis to control how much perspective there is. Usually I use dst=2 and scale = 0.5 to give z1=1/(2+z1*0.5);. rx, ry, rz are the rotations around the three axes, when these are all zero the z axis points through the screen and x and y are their usual selves.

You don't need to understand this code to use it, just write expressions for rx, ry, rz, x1, y1 and z1 as you would write for x, y, d, r or anything else.

The crx,srx... variables are simply used to optimise the code. If you really want to make the code neater (but slower) you can replace crx with cos(rx).. etc in the per point box.