After exactly one month of development I’m finally calling this project finished. Most of the time went into learning the new technology. Currently I’m waiting for the printers to send me the actual Lenticular print.

I’m attaching the files the sequence was made with. I actually had to do multiple passes. Once just to create the cartoon edge, once to create the actual images, and then using a third file to combine the two. The “birender” file is the one that looks most like the end result. Be carefull though, I used a computer at the Blender Institute to render it because it would have just taken too long to do on my laptop.

Here is a small video recorded on my mobile of what the print looks like

After working on the lenticular for quite a while already, today I truly realised how little time I have left. The affordable art fair is coming up quick! So… that leads me to some decisions.

Printing

I would have liked to print at a local place, which would have saved me quite a bit of money, but that would have been a lot of back and forth, both to the printer and the laminator. There’s no time for this at all. So in stead I’ll spend a bit extra and have the laminator print and adjust my image to get an optimal result. It does limit the scale of the final image somewhat. He can’t print wider than 60cm (2 feet), which is a shame… but saves me rendertime!

Design

I’m sticking with my original idea for the design even though I had my doubts. When a process takes this long, at a certain point you always start doubting yourself. I think this will work rather well. Here’s a single image from the sequence… still needs a little work!

Scripting

I’ve been doing a lot of scripting as well… A nice new little script is one I wrote to add a line to the left of any image. So I can just make a full size interlaced image, crop it… add the line (for aligning the lense sheet), then print. I’ll also attach the latest version of the interlacing script.

I’ve skipped work on this project for a few days. After spending a week on it, working all day every day my mind turned to mush. It’s rather a complex thing to work on. I took two days rest, had to work on another thing for a day, and now am back on it.

Progress is sweet!

The scene I have been working on is progressing well. I’m finding out lots doing my experiments. A big one is that motion should be small… Normally when you animate you think of frames per second and whether something moves quick enough in a second… When you make lenticulars it’s all about motion between frames… the difference between the frames shouldn’t be very big.

What I’m doing now is OpenGL renders from inside Blender 3D to check on camera motion. Here’s my latest little test (for a 40LPI sheet).

You can see that it doesn’t look like as much of a “washed out” mess as the previous tests. That is due to two factors. One is that the motion relative to the frames is much smaller… so the difference between frames is smaller too. And I’ve also cropped the image. So, it’s a full size render (two feet wide), but in this image you only see a detail of the large image.

Cropping

To crop I wanted to use the Border Render option in Blender 3D, but sadly that doesn’t work with OpenGL renders. Thankfully the aspect ratio settings do work there. To then crop all the images I get from OpenGL I wrote a tiny little extra python script which also uses the PIL library

# --------------------------------------------------------------------------
# ***** BEGIN GPL LICENSE BLOCK *****
#
# Copyright (C) 2010: macouno, http://www.macouno.com
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
# as published by the Free Software Foundation; either version 2
# of the License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software Foundation,
# Inc, 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
#
# ***** END GPL LICENCE BLOCK *****
# --------------------------------------------------------------------------
 
'''
This script takes a number of images from the directory the script is in itself,
and composites them into one with alternating strips
 
It is the start of a script for doing composites for Lenticular style prints.
It is still messy, quite experimental, and imperfect, but a first step.
 
Written to run with Python 2.5 and the PIL library
'''
 
import math, glob
 
STATE = {}
 
# Set the word to look for in the tile images to identify them
# (use names such as tilename.0001.png, or tilename0001.png for instance)
STATE['tileName'] = 'stalkie'
 
imX = 14400.0
imY = 8010.0
 
tileX = 960.0
tileY = 8010.0
 
# Crop settings (for this project, testing with openGL renders)
cropX = 3000.0
cropY = 3000.0
 
offX = int((tileX / imX) * 3000.0)
offY = int((tileY / imY) * 2000.0)
 
print 'offsets',offX,offY
 
xStart = offX
yStart = offY
 
xEnd = int(offX + ((tileX / imX) * cropX ))
yEnd = int(offY + ((tileY / imY) * cropY ))
 
print 'cropvalues',xStart, yStart, xEnd, yEnd
 
# Import the PIL library if possible
try:
        import PIL
        from PIL import Image
except:
        print 'error PIL library not found'
        exit()
 
for i, infile in enumerate(glob.glob('*.png')):
 
        if STATE['tileName'] in infile:
                print 'cropping',i,infile
                im = Image.open(infile)
 
                im = im.crop((xStart, yStart, xEnd, yEnd))
 
                im.save(infile)
 
                print 'saved',infile
 
print '-- finished --'

Display clean python code for copying

# --------------------------------------------------------------------------
# ***** BEGIN GPL LICENSE BLOCK *****
#
# Copyright (C) 2010: macouno, http://www.macouno.com
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
# as published by the Free Software Foundation; either version 2
# of the License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software Foundation,
# Inc, 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
#
# ***** END GPL LICENCE BLOCK *****
# --------------------------------------------------------------------------

'''
This script takes a number of images from the directory the script is in itself,
and composites them into one with alternating strips

It is the start of a script for doing composites for Lenticular style prints.
It is still messy, quite experimental, and imperfect, but a first step.

Written to run with Python 2.5 and the PIL library
'''

import math, glob

STATE = {}

# Set the word to look for in the tile images to identify them
# (use names such as tilename.0001.png, or tilename0001.png for instance)
STATE['tileName'] = 'stalkie'

imX = 14400.0
imY = 8010.0

tileX = 960.0
tileY = 8010.0

# Crop settings (for this project, testing with openGL renders)
cropX = 3000.0
cropY = 3000.0

offX = int((tileX / imX) * 3000.0)
offY = int((tileY / imY) * 2000.0)

print 'offsets',offX,offY

xStart = offX
yStart = offY

xEnd = int(offX + ((tileX / imX) * cropX ))
yEnd = int(offY + ((tileY / imY) * cropY ))

print 'cropvalues',xStart, yStart, xEnd, yEnd

# Import the PIL library if possible
try:
	import PIL
	from PIL import Image
except:
	print 'error PIL library not found'
	exit()

for i, infile in enumerate(glob.glob('*.png')):

	if STATE['tileName'] in infile:
		print 'cropping',i,infile
		im = Image.open(infile)

		im = im.crop((xStart, yStart, xEnd, yEnd))

		im.save(infile)

		print 'saved',infile

print '-- finished --'

For testing the final render in Blender I do want to use the Border render function in Blender, the script for that is in my previous post.

Decisions decisions

Since I’m getting closer and closer to the final design I called the printers today with a few more questions. The sad thing is that the printer who can do lenticular, doesn’t have a great deal of choice in materials to print on. But… he says it doesn’t actually matter much for lenticulars.

There is a printer nearby here in Amsterdam who can do amazing quality prints… but they’ve never done lenticular. And they don’t have what’s needed to stick the lense sheet on.

So what I want to do, is have the printer nearby do a gorgeous print, and have the dude in Rotterdam stick on the lense sheet.

There’s a problem with that though. Apparently each sheet can be slightly different. And so can every printer give different results. Thus… there’s no guarantee that a print made here in Amsterdam, that I test with the sheet I have, will line up perfect with the sheet at the company near Rotterdam.

That said… it’s actually cheaper if I do it that way.

Now the only decision left… do I buy an epson printer, which everyone says is best for lenticulars, and is also the brand the printers here in Amsterdam use…. Then I can test properly at home at higher DPI than I have been doing (1200dpi vs 600dpi)

I might hop on the bike right now to look at some prices… though I’m not sure it’s wise to spend even more money… To be continued! Definitely!

These last 2 days have been all about getting my scene set up in Blender 3D. And guess what… it’s been a challenge. Here’s why: I have 8660 objects, which make up 720 characters. Each character has 2 rigs, and at least 8 constraints, plus two sets of animation data. Oh, and I’m working on a laptop, which means, the computer basicly can’t handle it, becomes uncooperative, and everything takes for ever. Here’s me at my geekiest…

Solution

Thus, the only way to nicely edit things in my scene is to do everything by python script. Luckily Campbell Barton has done an amazing job with the Python API for the new Blender version (2.54). It is just magic. So I thought I’d share the little scripts I’ve made to mess around. I’ll try to comment them, but they’re not real “released” scripts, just stuff I’ve hacked together, but they might be of use to you.

Setting subdivision levels, and removing subsurf modifiers

import bpy
 
for ob in bpy.data.objects:
 
    try:
 
        # Set the subdivision level for all objects with a mesh named 'iris'
        if ob.data.name == 'iris':
 
            ob.data = bpy.data.meshes['simpleiris']
 
            print('set data for',ob.name)
 
            #for m in ob.modifiers:
 
                if m.type == 'SUBSURF':
 
                    m.levels = 1
                    m.render_levels = 1
 
        # Remove all subsurf modifiers from objects with a mesh named 'simpleiris'
        if ob.data.name == 'simpleiris':
 
            ob.name = 'iris'
 
            d = 'x'
 
            for i, m in enumerate(ob.modifiers):
 
                if m.type == 'SUBSURF':
 
                    d = i
 
            if not d == 'x':
 
                ob.modifiers.remove(ob.modifiers[d])
                print('removing from',ob.name)
 
    except:
        print('no data or error')
 
print(' -- finished --')

Display clean python code for copying

import bpy

for ob in bpy.data.objects:

    try:

        # Set the subdivision level for all objects with a mesh named 'iris'
        if ob.data.name == 'iris':

            ob.data = bpy.data.meshes['simpleiris']

            print('set data for',ob.name)

            #for m in ob.modifiers:

                if m.type == 'SUBSURF':

                    m.levels = 1
                    m.render_levels = 1

        # Remove all subsurf modifiers from objects with a mesh named 'simpleiris'
        if ob.data.name == 'simpleiris':

            ob.name = 'iris'

            d = 'x'

            for i, m in enumerate(ob.modifiers):

                if m.type == 'SUBSURF':

                    d = i

            if not d == 'x':

                ob.modifiers.remove(ob.modifiers[d])
                print('removing from',ob.name)

    except:
        print('no data or error')

print(' -- finished --')

Enable disable layers on specific objects

import bpy
 
lTargets = [ob for ob in bpy.data.groups['localtargets'].objects]
 
print(len(lTargets),'to do')
 
#lTargets = [ob for ob in bpy.context.selected_objects]
 
#Loop through the target objects and enable/disable specific layers
for c, ob in enumerate(lTargets):
 
    tLayers = []
 
    for i, l in enumerate(ob.layers):
 
        if i == 5:
            tLayers.append(True)
        elif i == 15:
            tLayers.append(False)
        elif i < 20:
            tLayers.append(l)
 
        print(i,l)
 
    print(len(tLayers))
 
    ob.layers = tLayers
 
    print(c, ob.name)
 
print('--finished--')

Display clean python code for copying

import bpy

lTargets = [ob for ob in bpy.data.groups['localtargets'].objects]

print(len(lTargets),'to do')

#lTargets = [ob for ob in bpy.context.selected_objects]

#Loop through the target objects and enable/disable specific layers
for c, ob in enumerate(lTargets):

    tLayers = []

    for i, l in enumerate(ob.layers):

        if i == 5:
            tLayers.append(True)
        elif i == 15:
            tLayers.append(False)
        elif i < 20:
            tLayers.append(l)

        print(i,l)

    print(len(tLayers))

    ob.layers = tLayers

    print(c, ob.name)

print('--finished--')

Set the subdivision level on objects lower the further they are from the camera

import bpy
 
camLoc = bpy.data.objects['Camera'].matrix_world.translation_part()
 
for ob in bpy.data.objects:
 
    try:
 
        n = ob.data.name
 
        # Only work on objects with a specific mesh linked in
        if n == 'neck' or n == 'eyeball' or n == 'pupil' or n == 'foot':
 
            gLoc = ob.matrix_world.translation_part()
 
            # Get the vector from the camera to the object so we know how far away it is
            relLoc = gLoc - camLoc
 
            if relLoc.length > 40:
 
                nLevel = 0
 
            elif relLoc.length < 20 and not (n == 'pupil' or n == 'foot'):
 
                nLevel = 2
 
            else:
 
                nLevel = 1
 
            # Set the subdivision level
            for m in ob.modifiers:
 
                if m.type == 'SUBSURF':
 
                    m.render_levels = nLevel
                    m.levels = nLevel
 
                    print('set level',nLevel,'for',ob.name)
 
    except:
        print('no data or error')
 
print('--finished--')

Display clean python code for copying

import bpy

camLoc = bpy.data.objects['Camera'].matrix_world.translation_part()

for ob in bpy.data.objects:

    try:

        n = ob.data.name

        # Only work on objects with a specific mesh linked in
        if n == 'neck' or n == 'eyeball' or n == 'pupil' or n == 'foot':

            gLoc = ob.matrix_world.translation_part()

            # Get the vector from the camera to the object so we know how far away it is
            relLoc = gLoc - camLoc

            if relLoc.length > 40:

                nLevel = 0

            elif relLoc.length < 20 and not (n == 'pupil' or n == 'foot'):

                nLevel = 2

            else:

                nLevel = 1

            # Set the subdivision level
            for m in ob.modifiers:

                if m.type == 'SUBSURF':

                    m.render_levels = nLevel
                    m.levels = nLevel

                    print('set level',nLevel,'for',ob.name)

    except:
        print('no data or error')

print('--finished--')

Set a whole bunch of objects to a specific “local” location

import bpy, mathutils
 
newLoc = mathutils.Vector([0.0, -5.0, 10.0])
scene = bpy.data.scenes[0]
 
for ob in bpy.data.groups['localtargets'].objects:
 
    ob.location = newLoc
    ob.update(scene)
    print('reset',ob.name)
 
print('-- finished --')

Display clean python code for copying

import bpy, mathutils

newLoc = mathutils.Vector([0.0, -5.0, 10.0])
scene = bpy.data.scenes[0]

for ob in bpy.data.groups['localtargets'].objects:

    ob.location = newLoc
    ob.update(scene)
    print('reset',ob.name)

print('-- finished --')

Assign a specific action to a list of objects

import bpy
 
oList = [ob for ob in bpy.data.groups['localtargets'].objects]
 
ac = bpy.data.actions['LocalTargetAction']
 
for ob in oList:
 
    try:
        ob.animation_data.action = ac
    except:
        ob.animation_data_create()
        ob.animation_data.action = ac
 
    print('set',ob.name)
 
print('-- finished --')

Display clean python code for copying

import bpy

oList = [ob for ob in bpy.data.groups['localtargets'].objects]

ac = bpy.data.actions['LocalTargetAction']

for ob in oList:

    try:
        ob.animation_data.action = ac
    except:
        ob.animation_data_create()
        ob.animation_data.action = ac

    print('set',ob.name)

print('-- finished --')

Setting a whole heap of bone rotations and keyframes for them at specific frames

import bpy, math, random
 
scene = bpy.data.scenes[0]
boneList = [[ob.pose.bones['iriscontrol'],0] for ob in bpy.data.objects if 'iriscontrol' in ob.name]
startList = []
blinkList = []
endList = []
cnt = 0
 
# Set a random frame for a blink
for b in boneList:
 
    r = 15.0
 
    while r > 13 and r < 17:
 
        r = int(300.0 * random.random())
 
    startList.append(r)
    blinkList.append(r+1)
    endList.append(r+2)
 
    if r < 30:
        cnt += 1
 
print('cnt',cnt)
 
#print(boneList)
 
# Loop through all the frames
for f in range(31):
 
    fr = f+1
 
    scene.frame_set(fr)
 
    for i, b in enumerate(boneList):
 
        if startList[i] == fr:
 
            b[0].rotation_euler = [math.radians(-15.0), 0.0, 0.0]   
 
            b[0].keyframe_insert('rotation_euler')
 
            print('start blink',i)
 
        elif blinkList[i] == fr:
 
            b[0].rotation_euler = [math.radians(-0.0), 0.0, 0.0]   
 
            b[0].keyframe_insert('rotation_euler')
 
            b[1] = 3
            boneList[i][1] = 3
 
            print('blink',i)
 
        if endList[i] == fr:
 
            b[0].rotation_euler = [math.radians(-15.0), 0.0, 0.0]   
 
            b[0].keyframe_insert('rotation_euler')
 
            b[1] = 0
            boneList[i][1] = 0
 
            print('end blink', i)
 
    print('-- set',f)
 
print('-- finished --')

Display clean python code for copying

import bpy, math, random

scene = bpy.data.scenes[0]
boneList = [[ob.pose.bones['iriscontrol'],0] for ob in bpy.data.objects if 'iriscontrol' in ob.name]
startList = []
blinkList = []
endList = []
cnt = 0

# Set a random frame for a blink
for b in boneList:

    r = 15.0

    while r > 13 and r < 17:

        r = int(300.0 * random.random())

    startList.append(r)
    blinkList.append(r+1)
    endList.append(r+2)

    if r < 30:
        cnt += 1

print('cnt',cnt)

#print(boneList)

# Loop through all the frames
for f in range(31):

    fr = f+1

    scene.frame_set(fr)

    for i, b in enumerate(boneList):

        if startList[i] == fr:

            b[0].rotation_euler = [math.radians(-15.0), 0.0, 0.0]   

            b[0].keyframe_insert('rotation_euler')

            print('start blink',i)

        elif blinkList[i] == fr:

            b[0].rotation_euler = [math.radians(-0.0), 0.0, 0.0]   

            b[0].keyframe_insert('rotation_euler')

            b[1] = 3
            boneList[i][1] = 3

            print('blink',i)

        if endList[i] == fr:

            b[0].rotation_euler = [math.radians(-15.0), 0.0, 0.0]   

            b[0].keyframe_insert('rotation_euler')

            b[1] = 0
            boneList[i][1] = 0

            print('end blink', i)

    print('-- set',f)

print('-- finished --')

Offset NLA strips relative to lots of position calculations

import bpy
 
offs = [ob.matrix_world.translation_part() for ob in bpy.data.groups['offsets'].objects]
ofDist = 10
 
cLoc = bpy.data.objects['Camera'].matrix_world.translation_part()
 
minDist = 'x'
maxDist = 'x'
 
start = 11
end = 511
 
for ob in bpy.data.groups['localtargets'].objects:
 
    t = ob.animation_data.nla_tracks['NlaTrack']
 
    s = t.strips[0]
 
    offset = 0
 
    obLoc = ob.matrix_world.translation_part()
 
    cDist = obLoc - cLoc
    cDist = cDist.length
 
    offset -= (cDist * 0.4)
 
    if minDist == 'x' or cDist < minDist:         minDist = cDist     if maxDist == 'x' or cDist > maxDist:
        maxDist = cDist
 
    for i, of in enumerate(offs):
 
        cDist = obLoc - of
        cDist = cDist.length
 
        if cDist < ofDist:
 
            # Use modulo to set a nice maximum deviation
            if (i%2) == 1:
                if (i%3) == 1:
                    max = -5
                else:
                    max = -10
            else:
                if(i%3) == 1:
                    max = -15
                else:
                    max = - 20
 
            cFact = (1.0/ofDist) * (ofDist - cDist)
 
            offset = offset + (max * cFact)
 
    s.action_frame_start = start + offset
    s.action_frame_end = end + offset
 
    print('set', ob.name)
 
print('minDist',minDist)
print('maxDist',maxDist)
print('-- finished --')

Display clean python code for copying

import bpy

offs = [ob.matrix_world.translation_part() for ob in bpy.data.groups['offsets'].objects]
ofDist = 10

cLoc = bpy.data.objects['Camera'].matrix_world.translation_part()

minDist = 'x'
maxDist = 'x'

start = 11
end = 511

for ob in bpy.data.groups['localtargets'].objects:

    t = ob.animation_data.nla_tracks['NlaTrack']

    s = t.strips[0]

    offset = 0

    obLoc = ob.matrix_world.translation_part()

    cDist = obLoc - cLoc
    cDist = cDist.length

    offset -= (cDist * 0.4)

    if minDist == 'x' or cDist < minDist:         minDist = cDist     if maxDist == 'x' or cDist > maxDist:
        maxDist = cDist

    for i, of in enumerate(offs):

        cDist = obLoc - of
        cDist = cDist.length

        if cDist < ofDist:

            # Use modulo to set a nice maximum deviation
            if (i%2) == 1:
                if (i%3) == 1:
                    max = -5
                else:
                    max = -10
            else:
                if(i%3) == 1:
                    max = -15
                else:
                    max = - 20

            cFact = (1.0/ofDist) * (ofDist - cDist)

            offset = offset + (max * cFact)

    s.action_frame_start = start + offset
    s.action_frame_end = end + offset

    print('set', ob.name)

print('minDist',minDist)
print('maxDist',maxDist)
print('-- finished --')

Messing with Constraints

import bpy, math
 
grp = bpy.data.groups['controllers']
 
ctrl = [ob for ob in grp.objects]
 
cam = bpy.data.objects['Camera']
camLoc = cam.matrix_world.translation_part()
 
minDist = 'x'
maxDist = 'x'
 
minF = 10
 
print(len(ctrl))
 
for ob in ctrl:
 
    # Get locations and rotations to use
    obLoc = ob.matrix_world.translation_part()
    pRot = math.degrees(ob.parent.rotation_euler[1])
 
    # Calculate the distance to the camera
    obDist = obLoc - camLoc
    obDist = obDist.length
 
    # Set the minimum and maximum distance from the camera
    if minDist == 'x' or obDist < minDist:         minDist = obDist     if maxDist == 'x' or obDist > maxDist:
        maxDist = obDist
 
    print('setting',ob.name)
 
    # Add new constraints
    ob.constraints.new('TRACK_TO')
    ob.constraints.new('TRANSFORM')
    ob.constraints.new('LOCKED_TRACK')
    ob.constraints.new('LOCKED_TRACK')
 
    # Retrieve the constraints
    tr = ob.constraints[2] # track to
    tf = ob.constraints[3] # transform
    tr2 = ob.constraints[4] # track to
    tr3 = ob.constraints[5] # locked track x
    tr4 = ob.constraints[6] # locked track z
 
    ob.constraints.remove(tr2)
 
    # Use the rotation of the parent to set the influence of the camera position to world or local
    if pRot > 360:
        pRot = pRot - 360
 
    if pRot > 270:
        inf = (1.0/90.0) * (90 - (360 - pRot))
    elif pRot < 90:
        inf = (1.0/90.0) * (90 - pRot)
    else:
        inf = 0.0
 
    # Set the influence of the constraints
    tr.influence = 1.0
    tf.influence = 1.0
    tr2.influence = inf
    tr3.influence = 0.0
    tr4.influence = 0.0
 
    # Insert a keyframe
    tr.keyframe_insert('influence')
    tf.keyframe_insert('influence')
    tr2.keyframe_insert('influence')
    tr3.keyframe_insert('influence')
    tr4.keyframe_insert('influence')
 
    # Set the track axis
    tr.track_axis = 'TRACK_NEGATIVE_Y'
    tr2.track_axis = 'TRACK_NEGATIVE_Y'
    tr3.track_axis = 'TRACK_NEGATIVE_Y'
    tr3.lock_axis = 'LOCK_X'
    tr4.track_axis = 'TRACK_NEGATIVE_Y'
    tr4.lock_axis = 'LOCK_Z'
 
    # Set the target for the constraint
    tr.target = cam
    tf.target = cam
    tr2.target = cam
    tr3.target = cam
    tr4.target = cam
 
    # Set values for the transform constraint
    tf.from_min_x = -50.0
    tf.from_min_y = -50.0
    tf.from_min_z = -50.0
 
    tf.from_max_x = 50.0
    tf.from_max_y = 50.0
    tf.from_max_z = 50.0
 
    tf.to_min_y = -0.5
    tf.to_max_y = -0.5
 
    if obDist < 11.0:
 
        obDist = obDist - 6
 
        n = obDist * 0.20
        n = -n
 
        if n < -1.0:             n = -1.0         elif n > 0.0:
            n = 0.0
 
    else:
        n = -1.0
 
    tf.to_min_z = n
    tf.to_max_z = n
 
    tr.owner_space = 'LOCAL'
    tf.owner_space = 'LOCAL'
 
print('minDist',minDist)
print('maxDist',maxDist)
 
print('--finished--')

Display clean python code for copying

import bpy, math

grp = bpy.data.groups['controllers']

ctrl = [ob for ob in grp.objects]

cam = bpy.data.objects['Camera']
camLoc = cam.matrix_world.translation_part()

minDist = 'x'
maxDist = 'x'

minF = 10

print(len(ctrl))

for ob in ctrl:

    # Get locations and rotations to use
    obLoc = ob.matrix_world.translation_part()
    pRot = math.degrees(ob.parent.rotation_euler[1])

    # Calculate the distance to the camera
    obDist = obLoc - camLoc
    obDist = obDist.length

    # Set the minimum and maximum distance from the camera
    if minDist == 'x' or obDist < minDist:         minDist = obDist     if maxDist == 'x' or obDist > maxDist:
        maxDist = obDist

    print('setting',ob.name)

    # Add new constraints
    ob.constraints.new('TRACK_TO')
    ob.constraints.new('TRANSFORM')
    ob.constraints.new('LOCKED_TRACK')
    ob.constraints.new('LOCKED_TRACK')

    # Retrieve the constraints
    tr = ob.constraints[2] # track to
    tf = ob.constraints[3] # transform
    tr2 = ob.constraints[4] # track to
    tr3 = ob.constraints[5] # locked track x
    tr4 = ob.constraints[6] # locked track z

    ob.constraints.remove(tr2)

    # Use the rotation of the parent to set the influence of the camera position to world or local
    if pRot > 360:
        pRot = pRot - 360

    if pRot > 270:
        inf = (1.0/90.0) * (90 - (360 - pRot))
    elif pRot < 90:
        inf = (1.0/90.0) * (90 - pRot)
    else:
        inf = 0.0

    # Set the influence of the constraints
    tr.influence = 1.0
    tf.influence = 1.0
    tr2.influence = inf
    tr3.influence = 0.0
    tr4.influence = 0.0

    # Insert a keyframe
    tr.keyframe_insert('influence')
    tf.keyframe_insert('influence')
    tr2.keyframe_insert('influence')
    tr3.keyframe_insert('influence')
    tr4.keyframe_insert('influence')

    # Set the track axis
    tr.track_axis = 'TRACK_NEGATIVE_Y'
    tr2.track_axis = 'TRACK_NEGATIVE_Y'
    tr3.track_axis = 'TRACK_NEGATIVE_Y'
    tr3.lock_axis = 'LOCK_X'
    tr4.track_axis = 'TRACK_NEGATIVE_Y'
    tr4.lock_axis = 'LOCK_Z'

    # Set the target for the constraint
    tr.target = cam
    tf.target = cam
    tr2.target = cam
    tr3.target = cam
    tr4.target = cam

    # Set values for the transform constraint
    tf.from_min_x = -50.0
    tf.from_min_y = -50.0
    tf.from_min_z = -50.0

    tf.from_max_x = 50.0
    tf.from_max_y = 50.0
    tf.from_max_z = 50.0

    tf.to_min_y = -0.5
    tf.to_max_y = -0.5

    if obDist < 11.0:

        obDist = obDist - 6

        n = obDist * 0.20
        n = -n

        if n < -1.0:             n = -1.0         elif n > 0.0:
            n = 0.0

    else:
        n = -1.0

    tf.to_min_z = n
    tf.to_max_z = n

    tr.owner_space = 'LOCAL'
    tf.owner_space = 'LOCAL'

print('minDist',minDist)
print('maxDist',maxDist)

print('--finished--')

A couple days ago I started development on Lenticular prints. I have to say it’s been going rather well, though it’s not exactly easy. You have to keep your wits about you with tile sizes, LPI, DPI and so forth. Here’s what I tried so far!

First prints

1. I started with a basic black and white image, alternating stripes. To see if it works at all.

2. That seemed to work so I tried a circle and a square

3. I found out that it’s nice to add a little bar on the left to line up the lenticular sheet. one weird thing is that from left to right the animation loops (so you can see the circle multiple times). It’s not just the left = circle and the right = square. So I decided to try something a bit more complex.

4. That was ABCDE, and logically the letters came past backwards, so I need to reverse the array with the images in it. I also tried the lentikit open source software, and found it does exactly what my script already does. And then Beorn stopped by and dedicated a fraction of his considerable brain power to think with me. It now seems obvious that you should not have the DPI of the printer as the variable, but in stead use the image size, and use multiple pixel strips per lense (5 px per tile for instance). So just now I did a rather crazy colourful test.

Because I’m now using 600DPI irrelevant of the number of tiles… the image got quite large… 3000 x 3000 px. This may be an issue later on. This is only a 12cm print… ish. What if I want a huge print at 1200DPI? That’ll be insane… though I think the PIL library can manage.

One odd thing I keep noticing is that the sequence does seem to loop when looking left to right… I need to check that, but I’m hoping to get a little advice.

In the meantime

Whilst I’m “waiting” for help I’ll start with the design/modelling of the image I have in mind. Here’s a first little sneak peak….

There’s going to be LOTS of these. Today I want to continue texturing, and hopefully rigging will follow tomorrow.

I’m still very excited about the project… there’s lots left to figure out!

You can download the current version of the script at the bottom of the article