Andro

After getting the hang of jit.gen with basic matrix operations I've decided to try and step away from jit.gl.gridshape @matrix_ouput 1 and learn how to create basic geometry with equations and jit.gen.

http://www.iquilezles.org/www/articles/distfunctions/distfunctions.htm

I'm just having some trouble converting some of what I see there into the world of jit.gen.

I'd like to make an entry on the jit.gen wiki for creating most of those shapes and functions for newcomers like me (because documentation is sparse )

So my 1st challenge is trying to convert the code for a torus - signed.

I'm pretty close (like theres definitely something going on with 2 cylinder things in there )

Can anyone point out what I'm doing wrong with this patch ?

After getting the hang of jit.gen with basic matrix operations  I've decided to try and step away from jit.gl.gridshape @matrix_ouput 1 and learn how to create basic geometry with equations and jit.gen.
I've been checking out the website
http://www.iquilezles.org/www/articles/distfunctions/distfunctions.htm
for generating models.
I'm just having some trouble converting some of what I see there into the world of jit.gen.
I'd like to make an entry on the jit.gen wiki for creating most of those shapes and functions for newcomers like me (because documentation is sparse )

So my 1st challenge is trying to convert the code for a torus - signed.
I'm pretty close (like theres definitely something going on with 2 cylinder things in there )
Can anyone point out what I'm doing wrong with this patch ?


modelling-shapes-with-distance-functions-and-jit-gen

just remove the final 'l": http://www.iquilezles.org/www/articles/distfunctions/distfunctions.htm


Updated the original link so it works now.

The code used from the website is inside the jit.gen object with a comment box.

Updated the original link so it works now.
The code used from the website is inside the jit.gen object with a comment box.


very cool page! this makes me want to immediately dig into GLSL...

I tried using gen's codebox and the translation was very easy (plus some manual tweaking to get the right values for the "t" vector..

...and here's the same formula translated into gen objects (even if the results are a bit different... can anybody tell why?)

Thank you so much Lska !!! This has made a lot of simple operations with jit.gen (especially codebox !) much clearer for me :)

I'm going to try and do the rest of the primitives (to improve my jit.gen skills) and post them on this page (plus any failed attempts)

Thank you so much Lska !!! This has made a lot of simple operations with jit.gen (especially codebox !) much clearer for me :)
I'm going to try and do the rest of the primitives (to improve my jit.gen skills) and post them on this page (plus any failed attempts)
Thanks !


thanks for this , gonna dig more and try to get better @ gen .

Woah LSKA! smooth operator :-)
Thanks andro for bringing this up 

thanks for this , gonna dig more and try to get better @ gen .
Thanks again you guys :-) 

Alright, I've converted 2 more primitives into codebox, hexagonal prism and cone-signed.

Stuck right now with how to declare the variable length8

It then doesn't work as it says that length8 is not defined.

Original Code and comments are inside the jit.gen objects

Alright, I've converted 2 more primitives into codebox, hexagonal prism and cone-signed.
Stuck right now with how to declare the variable length8
I set
length8 = length(8);

It then doesn't work as it says that length8 is not defined.
Not sure how to set this in codebox.

Original Code and comments are inside the jit.gen objects


https://cycling74.com/wiki/index.php?title=Gen_Patch-A-Day#Day_5

Cool -- and see also https://cycling74.com/wiki/index.php?title=Gen_Patch-A-Day#Day_5


@Andro: length() is a function, not a variable.

As written inthe original webpage, the lenght() function is defined as: (x^2+y^2+z^2)^(1/2) , thus length8() should be (x^2+y^2+z^2)^(1/8)

And here's the translation in gen (BTW, I just learned how to define parameters inside the codebox):

@Andro: length() is a function, not a variable.
As written inthe original webpage, the lenght() function is defined as: (x^2+y^2+z^2)^(1/2) , thus length8() should be (x^2+y^2+z^2)^(1/8)

And here's the translation in gen (BTW, I just learned how to define parameters inside the codebox):


Thanks Lska, was way too tired yesterday to even notice that.

Heres another version of the patch I posted with the following shapes

Trying to get my head around the rest of the page, I'll post more as I get there.

Thanks Lska, was way too tired yesterday to even notice that.
Heres another version of the patch I posted with the following shapes
-torus
-hexagonal prism
-cone
-Triangular prism
- Box with variables for x y z scale
- Torus 88 signed (your example)

Trying to get my head around the rest of the page, I'll post more as I get there.


sorry to dig up an old thread but wondering if anyone tried the operations / deformations on this list further down?

Got a basic example somewhere. I'll see if i can dig it up tomorrow.

If I'm correct its just multiplying one result with another in its badic form.

Got a basic example somewhere. I'll see if i can dig it up tomorrow.
If I'm correct its just multiplying one result with another in its badic form.


Cool thread... thanks for sharing. This is helpful.

…and here’s the same formula translated into gen objects (even if the results are a bit different… can anybody tell why?)

@lska It looks like you flipped the final operation in the patch.. Instead of subtracting t.y from the length of q and outputting the result, you subtracted t.y from q and outputted the length of the difference.

Cool thread...  thanks for sharing.  This is helpful.

@lska  It looks like you flipped the final operation in the patch..  Instead of subtracting t.y from the length of q and outputting the result, you subtracted t.y from q and outputted the length of the difference.


modelling shapes with distance functions and jit.gen