"LanuHum" <Lan### [at] yandexru> wrote:
> "And" <49341109@ntnu.edu.tw> wrote:
> > "LanuHum" <Lan### [at] yandexru> wrote:
> > > I looked here
> > > http://www.f-lohmueller.de/pov_tut/modelling/model_000e.htm
> > > and recalled Yoda with a dark side of force...
> > >
> > > Only 32 bicubic_patches build 10 kilometers of the curved railroad.
> >
> > Looks nice.
>
> Thanks.
> Here I bent a rail manually.
> Now I learn to build parallel Bezier curves. (in 3-D)
> It is necessary to automate process of a bend along a Bezier curve.

Parallel Bezier curve... sounds a challenge. I use Inkscape to do this operation
sometimes.(But only 2-D) Keep it up.

Post a reply to this message

"And" <49341109@ntnu.edu.tw> wrote:

> Parallel Bezier curve... sounds a challenge. I use Inkscape to do this operation
> sometimes.(But only 2-D) Keep it up.

The problem is solved.
Inkscape won't help. Parallel curves shall have number of control points equal
to number of control points of the original (sample).
I will use only the Blender tools.
Yesterday I already built a parallel bezier curve(3D) in Blender, using Python
and formulas. :)))))))

Post a reply to this message

"LanuHum" <Lan### [at] yandexru> wrote:
> "And" <49341109@ntnu.edu.tw> wrote:
>
> > Parallel Bezier curve... sounds a challenge. I use Inkscape to do this operation
> > sometimes.(But only 2-D) Keep it up.
>
> The problem is solved.
> Inkscape won't help. Parallel curves shall have number of control points equal
> to number of control points of the original (sample).
> I will use only the Blender tools.
> Yesterday I already built a parallel bezier curve(3D) in Blender, using Python
> and formulas. :)))))))

Wow, that's great.

Post a reply to this message

"And" <49341109@ntnu.edu.tw> wrote:
> "LanuHum" <Lan### [at] yandexru> wrote:
> > "And" <49341109@ntnu.edu.tw> wrote:
> >
> > > Parallel Bezier curve... sounds a challenge. I use Inkscape to do this operation
> > > sometimes.(But only 2-D) Keep it up.
> >
> > The problem is solved.
> > Inkscape won't help. Parallel curves shall have number of control points equal
> > to number of control points of the original (sample).
> > I will use only the Blender tools.
> > Yesterday I already built a parallel bezier curve(3D) in Blender, using Python
> > and formulas. :)))))))
>
> Wow, that's great.

It's a go. Parallel curves are constructed automatically.
Now control of a UV-mapping is necessary.

Post a reply to this message

Attachments:
Download 'rail_guide.blend.jpg' (471 KB)

Preview of image 'rail_guide.blend.jpg'

"LanuHum" <Lan### [at] yandexru> wrote:
> "And" <49341109@ntnu.edu.tw> wrote:
> > "LanuHum" <Lan### [at] yandexru> wrote:
> > > "And" <49341109@ntnu.edu.tw> wrote:
> > >
> > > > Parallel Bezier curve... sounds a challenge. I use Inkscape to do this
operation
> > > > sometimes.(But only 2-D) Keep it up.
> > >
> > > The problem is solved.
> > > Inkscape won't help. Parallel curves shall have number of control points equal
> > > to number of control points of the original (sample).
> > > I will use only the Blender tools.
> > > Yesterday I already built a parallel bezier curve(3D) in Blender, using Python
> > > and formulas. :)))))))
> >
> > Wow, that's great.
>
> It's a go. Parallel curves are constructed automatically.
> Now control of a UV-mapping is necessary.

Do you write them by yourself? I recall that you ever said your math was not
good enough to deal with such a thing.

Post a reply to this message

"And" <49341109@ntnu.edu.tw> wrote:
> "LanuHum" <Lan### [at] yandexru> wrote:

> >
> > It's a go. Parallel curves are constructed automatically.
> > Now control of a UV-mapping is necessary.
>
> Do you write them by yourself? I recall that you ever said your math was not
> good enough to deal with such a thing.

I am very bad mathematician.
Almost all computation is made by Blender. I am able to cause the necessary
functions.
I found something in the Internet.
1/3 t and 2/3 t are found by the Blender. The formula finds P1 and P2.
If you like to simulate on formulas, then use Blender and Python.
Less mathematics will be required.
:)))))

Post a reply to this message

"And" <49341109@ntnu.edu.tw> wrote:
> "LanuHum" <Lan### [at] yandexru> wrote:
> > "And" <49341109@ntnu.edu.tw> wrote:
> > > "LanuHum" <Lan### [at] yandexru> wrote:
> > > > "And" <49341109@ntnu.edu.tw> wrote:
> > > >
> > > > > Parallel Bezier curve... sounds a challenge. I use Inkscape to do this
operation
> > > > > sometimes.(But only 2-D) Keep it up.
> > > >
> > > > The problem is solved.
> > > > Inkscape won't help. Parallel curves shall have number of control points equal
> > > > to number of control points of the original (sample).
> > > > I will use only the Blender tools.
> > > > Yesterday I already built a parallel bezier curve(3D) in Blender, using Python
> > > > and formulas. :)))))))
> > >
> > > Wow, that's great.
> >
> > It's a go. Parallel curves are constructed automatically.
> > Now control of a UV-mapping is necessary.
>
> Do you write them by yourself? I recall that you ever said your math was not
> good enough to deal with such a thing.

How many there is math?

import bpy
scene = bpy.context.scene
ob = bpy.context.object
cu = ob.parent
pd = cu.data.path_duration
frames = 4*[1]
frames[1] = pd/3
frames[2] = pd/3*2
frames[3] = pd
duplis = []
for i in range(4):
    scene.frame_set(frames[i],0.0)
    bpy.ops.object.duplicate()
    du = bpy.context.object
    duplis.append(du)
    cu.select = True
    scene.objects.active = cu
    bpy.ops.object.parent_clear(type='CLEAR_KEEP_TRANSFORM')
    du.select = False
    cu.select = False
    ob.select = True
    scene.objects.active = ob
ob.select = False
bpts = []
for i in range(4):
    bpts.append(duplis[i].data.splines[0].bezier_points)
all_coords = []
for i, pt in enumerate(bpts[0]):
    coords = []
    for p in range(4):
        co = duplis[p].matrix_world * bpts[p][i].handle_left
        coords.append(co[:])
    all_coords.append(coords)
    coords = []
    for p in range(4):
        co = duplis[p].matrix_world * bpts[p][i].co
        coords.append(co[:])
    all_coords.append(coords)
    all_coords.append(coords)
    coords = []
    for p in range(4):
        co = duplis[p].matrix_world * bpts[p][i].handle_right
        coords.append(co[:])
    all_coords.append(coords)
all_coords.append(all_coords[0])
all_coords.pop(0)
all_coords.append(all_coords[0])
all_coords.pop(0)
cu = None

for i, coords in enumerate(all_coords):
    if i % 4 == 0:
        cu = bpy.data.curves.new('MyCurve', 'CURVE')
        cu.dimensions = '3D'
        obj = bpy.data.objects.new('MyCurveObject', cu)
        bpy.context.scene.objects.link(obj)
    x0 = coords[0][0]
    y0 = coords[0][1]
    z0 = coords[0][2]
    x1 = coords[1][0]
    y1 = coords[1][1]
    z1 = coords[1][2]
    x2 = coords[2][0]
    y2 = coords[2][1]
    z2 = coords[2][2]
    x3 = coords[3][0]
    y3 = coords[3][1]
    z3 = coords[3][2]

    p1x=(-5*x0+18*x1-9*x2+2*x3)/6
    p1y=(-5*y0+18*y1-9*y2+2*y3)/6
    p1z=(-5*z0+18*z1-9*z2+2*z3)/6
    p2x=(2*x0-9*x1+18*x2-5*x3)/6
    p2y=(2*y0-9*y1+18*y2-5*y3)/6
    p2z=(2*z0-9*z1+18*z2-5*z3)/6

    bezier = [((x0,y0,z0),(x0,y0,z0),(p1x,p1y,p1z)),
              ((x3,y3,z3),(p2x,p2y,p2z),(x3,y3,z3))]

    spline = cu.splines.new('BEZIER')
    spline.bezier_points.add(1)
    for n in range(2):
        bpt = spline.bezier_points[n]
        bpt.co[:] = bezier[n][0]
        bpt.handle_left[:] = bezier[n][1]
        bpt.handle_right[:] = bezier[n][2]

Post a reply to this message

"LanuHum" <Lan### [at] yandexru> wrote:
> "And" <49341109@ntnu.edu.tw> wrote:
> > "LanuHum" <Lan### [at] yandexru> wrote:
> > > "And" <49341109@ntnu.edu.tw> wrote:
> > > > "LanuHum" <Lan### [at] yandexru> wrote:
> > > > > "And" <49341109@ntnu.edu.tw> wrote:
> > > > >
> > > > > > Parallel Bezier curve... sounds a challenge. I use Inkscape to do this
operation
> > > > > > sometimes.(But only 2-D) Keep it up.
> > > > >
> > > > > The problem is solved.
> > > > > Inkscape won't help. Parallel curves shall have number of control points
equal
> > > > > to number of control points of the original (sample).
> > > > > I will use only the Blender tools.
> > > > > Yesterday I already built a parallel bezier curve(3D) in Blender, using
Python
> > > > > and formulas. :)))))))
> > > >
> > > > Wow, that's great.
> > >
> > > It's a go. Parallel curves are constructed automatically.
> > > Now control of a UV-mapping is necessary.
> >
> > Do you write them by yourself? I recall that you ever said your math was not
> > good enough to deal with such a thing.
>
> How many there is math?
>
> import bpy
> scene = bpy.context.scene
> ob = bpy.context.object
> cu = ob.parent
> pd = cu.data.path_duration
> frames = 4*[1]
> frames[1] = pd/3
> frames[2] = pd/3*2
> frames[3] = pd
> duplis = []
> for i in range(4):
>     scene.frame_set(frames[i],0.0)
>     bpy.ops.object.duplicate()
>     du = bpy.context.object
>     duplis.append(du)
>     cu.select = True
>     scene.objects.active = cu
>     bpy.ops.object.parent_clear(type='CLEAR_KEEP_TRANSFORM')
>     du.select = False
>     cu.select = False
>     ob.select = True
>     scene.objects.active = ob
> ob.select = False
> bpts = []
> for i in range(4):
>     bpts.append(duplis[i].data.splines[0].bezier_points)
> all_coords = []
> for i, pt in enumerate(bpts[0]):
>     coords = []
>     for p in range(4):
>         co = duplis[p].matrix_world * bpts[p][i].handle_left
>         coords.append(co[:])
>     all_coords.append(coords)
>     coords = []
>     for p in range(4):
>         co = duplis[p].matrix_world * bpts[p][i].co
>         coords.append(co[:])
>     all_coords.append(coords)
>     all_coords.append(coords)
>     coords = []
>     for p in range(4):
>         co = duplis[p].matrix_world * bpts[p][i].handle_right
>         coords.append(co[:])
>     all_coords.append(coords)
> all_coords.append(all_coords[0])
> all_coords.pop(0)
> all_coords.append(all_coords[0])
> all_coords.pop(0)
> cu = None
>
> for i, coords in enumerate(all_coords):
>     if i % 4 == 0:
>         cu = bpy.data.curves.new('MyCurve', 'CURVE')
>         cu.dimensions = '3D'
>         obj = bpy.data.objects.new('MyCurveObject', cu)
>         bpy.context.scene.objects.link(obj)
>     x0 = coords[0][0]
>     y0 = coords[0][1]
>     z0 = coords[0][2]
>     x1 = coords[1][0]
>     y1 = coords[1][1]
>     z1 = coords[1][2]
>     x2 = coords[2][0]
>     y2 = coords[2][1]
>     z2 = coords[2][2]
>     x3 = coords[3][0]
>     y3 = coords[3][1]
>     z3 = coords[3][2]
>
>     p1x=(-5*x0+18*x1-9*x2+2*x3)/6
>     p1y=(-5*y0+18*y1-9*y2+2*y3)/6
>     p1z=(-5*z0+18*z1-9*z2+2*z3)/6
>     p2x=(2*x0-9*x1+18*x2-5*x3)/6
>     p2y=(2*y0-9*y1+18*y2-5*y3)/6
>     p2z=(2*z0-9*z1+18*z2-5*z3)/6
>
>     bezier = [((x0,y0,z0),(x0,y0,z0),(p1x,p1y,p1z)),
>               ((x3,y3,z3),(p2x,p2y,p2z),(x3,y3,z3))]
>
>     spline = cu.splines.new('BEZIER')
>     spline.bezier_points.add(1)
>     for n in range(2):
>         bpt = spline.bezier_points[n]
>         bpt.co[:] = bezier[n][0]
>         bpt.handle_left[:] = bezier[n][1]
>         bpt.handle_right[:] = bezier[n][2]

You say that it doesn't have many mathematics. Ok.
By contrast, I can't understand this code.

Post a reply to this message

"And" <49341109@ntnu.edu.tw> wrote:
>
> You say that it doesn't have many mathematics. Ok.
> By contrast, I can't understand this code.

:)))))))
In any source code of the math looks equally:
* / + - sin cos tan atan sqrt and so on...
In my code you will see only 6 lines math.

Post a reply to this message

"LanuHum" <Lan### [at] yandexru> wrote:
> "And" <49341109@ntnu.edu.tw> wrote:
> >
> > You say that it doesn't have many mathematics. Ok.
> > By contrast, I can't understand this code.
>
> :)))))))
> In any source code of the math looks equally:
> * / + - sin cos tan atan sqrt and so on...
> In my code you will see only 6 lines math.

Well.

Post a reply to this message