+- LDraw.org Discussion Forums (https://forums.ldraw.org)
+-- Forum: Models and Parts (https://forums.ldraw.org/forum-18.html)
+--- Forum: Parts Authoring (https://forums.ldraw.org/forum-19.html)
+--- Thread: Elliptical torus-es? (/thread-27877.html)
Elliptical torus-es? - N. W. Perry - 2023-12-13
Before I bend over too far trying to figure it out, is it possible to make a torus, using primitives, that is elliptical in the major radius but without changing the minor radius? I think it is not, because the minor radius is in ratio to the major radius, so scaling the primitive differently in X and Y will change both radii proportionally.
The same problem exists with rings, but you can get around that using different primitives, such as chords and n-discs, as I've done in the two files below. (I'm trying to make both a flat and a rounded version; I've figured out the L and the E, but not how to make the round-topped version of G and O.) So maybe there is some way using another kind of primitive?
RE: Elliptical torus-es? - Willy Tschager - 2023-12-13
(2023-12-13, 2:02)N. W. Perry Wrote: Before I bend over too far trying to figure it out, is it possible to make a torus, using primitives, that is elliptical in the major radius but without changing the minor radius? I think it is not, because the minor radius is in ratio to the major radius, so scaling the primitive differently in X and Y will change both radii proportionally.
The same problem exists with rings, but you can get around that using different primitives, such as chords and n-discs, as I've done in the two files below. (I'm trying to make both a flat and a rounded version; I've figured out the L and the E, but not how to make the round-topped version of G and O.) So maybe there is some way using another kind of primitive?
It's too early in the morning for maths, but those logos look really nice!
w.
RE: Elliptical torus-es? - Philippe Hurbain - 2023-12-13
(2023-12-13, 2:02)N. W. Perry Wrote: The same problem exists with rings, but you can get around that using different primitives, such as chords and n-discs, as I've done in the two files below. (I'm trying to make both a flat and a rounded version; I've figured out the L and the E, but not how to make the round-topped version of G and O.) So maybe there is some way using another kind of primitive?With some drawbacks (flat area, variable width bevels) you can apply a similar method. Attached your file with my proposal for O. Far from perfect, but looks rather good...
RE: Elliptical torus-es? - N. W. Perry - 2023-12-13
(2023-12-13, 5:45)Willy Tschager Wrote: It's too early in the morning for maths, but those logos look really nice!
w.
Thanks.
At first I tried using 1-4 chords and n-discs for the flat G and O, but the result was some ugly, squashy triangles. I was pleasantly surprised at how much more elegant it looks with 1-8 variants!By the way, these logos would be used for the underside of various early bricks (i.e., 1950s). There are several minor variations of the logo from that period: flat, rounded (though some look more bevelled than rounded), long G, short G, etc. So there will be the separate question of which, if any, of these are worth actually submitting to the library, since we don't usually get hung up on non-visible details.
But it's still a worthwhile exercise, since I think that any text-based patterns should employ primitives as much as possible, so that the text can be re-used and scaled up for other applications.
(2023-12-13, 8:30)Philippe Hurbain Wrote: With some drawbacks (flat area, variable width bevels) you can apply a similar method. Attached your file with my proposal for O. Far from perfect, but looks rather good...
Hmm, very nice…the flat area bothers me much, much less than a variable-width letter O would!
And as I say, some of the photos I've seen of this logo look more bevelled than rounded anyway. And with condline shading in LDView, you really don't even notice the flatness.I wonder if it is worth creating the full-circle version of these tori (since you have to create the inner one anyway)?
RE: Elliptical torus-es? - Philippe Hurbain - 2023-12-13
(2023-12-13, 14:35)N. W. Perry Wrote: I wonder if it is worth creating the full-circle version of these tori (since you have to create the inner one anyway)?I tend to create smaller ones (1/4 or even 1/8): more flexible / smaller files. Reuse of tori is lower than for other prims...
RE: Elliptical torus-es? - Magnus Forsberg - 2023-12-13
(2023-12-13, 16:12)Philippe Hurbain Wrote: I tend to create smaller ones (1/4 or even 1/8): more flexible / smaller files. Reuse of tori is lower than for other prims...
Yes.
1/8 are enough. More versatile and reusable.
Any prim beyond 1/4 is waste of harddisc space.
RE: Elliptical torus-es? - N. W. Perry - 2023-12-17
Well, I started work on the G, and quickly realized my problem…its curves don't all have the same center and the scale factors aren't symmetrical. So I'm guessing it won't be possible to torus-ize the whole letter without normalizing the geometry…
RE: Elliptical torus-es? - Philippe Hurbain - 2023-12-17
(2023-12-17, 4:41)N. W. Perry Wrote: Well, I started work on the G, and quickly realized my problem…its curves don't all have the same center and the scale factors aren't symmetrical. So I'm guessing it won't be possible to torus-ize the whole letter without normalizing the geometry…
Might be possible to cheat a bit using segments of cyli/cyls, slanted or not...
RE: Elliptical torus-es? - N. W. Perry - 2023-12-18
(2023-12-17, 7:30)Philippe Hurbain Wrote: Might be possible to cheat a bit using segments of cyli/cyls, slanted or not...
Hmm…something like this, maybe?
Not bad, but I don't like some of the sharp lines around the 1-4sphe.
RE: Elliptical torus-es? - [email protected] - 2024-01-11
Hi, I think you should try rotating it the other way around.