I wanted a looping animation that appears to deform space continuously. This restricts the mapping to be continuous and bijective. Arbituarily, I choose the domain to be the unit disc and the mapping to be $z = r \exp{(ix)} \rightarrow \text{frac}(r+c) \exp{(ix)}$. Perhaps other mappings might be interesting.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
from PIL import Image, ImageDraw, ImageFont, ImageChops
import numpy as np
def trim(im):
i = np.array(im)
i[i[:,:,3] == 0] = 0
while True:
if i[0].max() == 0: i = i[1:,:,:]
else: break
while True:
if i[-1].max() == 0: i = i[:-1,:,:]
else: break
while True:
if i[:,0].max() == 0: i = i[:,1:,:]
else: break
while True:
if i[:,-1].max() == 0: i = i[:,:-1,:]
else: break
return Image.fromarray(i)
def str2img(s:str, fz:int):
img = Image.new("RGBA", (300*len(s),300), (255,255,255,0))
draw = ImageDraw.Draw(img)
font = ImageFont.truetype(font_path, fz)
draw.text((10, 0), s, (0,0,0), font=font)
return trim(img)
def to_img(arr):
return Image.fromarray((np.clip(arr, 0, 1).astype(np.float)*255).astype(np.uint8))
def to_arr(img):
return np.array(img).astype(np.float) / 255
def place_text(canvas, s, fz, col):
t = to_arr(str2img(s,fz))
t[t[:,:,-1] > 0] = col
x,y,_ = t.shape
cx,cy,_ = canvas.shape
assert cx >= x and cy >= y
dx = int(.5*(cx-x)); dy = int(.5*(cy-y))
canvas[dx:dx+x,dy:dy+y,:] += t
return canvas
K = 2
def transform(canvas, c, _cache={}):
_cache = {}
cx,cy,_ = canvas.shape
nc = np.zeros((cx,cy,4))
def finv(x,y):
x,y = x/cx * K - K/2, y/cy * K - K/2
z = complex(x,y)
r = abs(z)
if r == 0: return True, 0,0
r = (r+c)%1
if r < 0: return False, 0,0
z = z/abs(z) * r
x,y = z.real, z.imag
x,y = (x+K/2)/K * cx, (y+K/2)/K * cy
return 0<=x<cx-1 and 0<=y<cy-1, x,y
if (cx,cy,c) in _cache:
finv_arr = _cache[(cx,cy,c)]
else:
finv_arr = [
[finv(x,y) for x in range(cx)]
for y in range(cy)
]
_cache[(cx,cy,c)] = finv_arr
for x in range(cx):
for y in range(cy):
if not (0<x<cx-1 and 0<y<cy-1): continue
sxy = [finv_arr[y+j][x+k] for j in [0] for k in [0]]
if sum(s for s,_,_ in sxy) != len(sxy): continue
for _,nx,ny in sxy:
iy,ix = int(ny), int(nx)
fy,fx = ny%1, nx%1
nc[x,y] += (
canvas[ix,iy] * fy * fx +
canvas[ix+1,iy] * (1-fx) * fy +
canvas[ix,iy+1] * (1-fy) * fx +
canvas[ix+1,iy+1] * (1-fy) * (1-fx)
)
nc[x,y] /= len(sxy)
return nc
J = 1
canvas = np.zeros((1800//J,1800//J,4))
canvas += np.array([7, 42, 95, 255], dtype=np.float)/255
canvas = place_text(canvas, "hello", 260//J, np.array([0, 1, 0.7, 1.]))
canvas = transform(canvas, 0.5)
canvas = place_text(canvas, "world", 260//J, np.array([1, 0.66, 0, 1.]))
canvas = transform(canvas, 0.5)
nc = to_img(canvas)
nc = to_arr(nc)
nframes = 80
cs = np.linspace(0,1,nframes)
for idx,c in list(enumerate(cs)):
print(idx, end="\r")
_nc = transform(nc, c)
_nc = to_img(_nc)
_nc.save(f"out/{idx}.png")
def gen_frame(path):
im = Image.open(path)
alpha = im.getchannel('A')
im = im.convert('RGB').convert('P', palette=Image.ADAPTIVE, colors=255)
mask = Image.eval(alpha, lambda a: 255 if a <=128 else 0)
im.paste(255, mask)
im.info['transparency'] = 255
return im
imgs = [gen_frame(f"out/{idx}.png") for idx in range(nframes-1)][::-1]
imgs[0].save('dist/out.gif', save_all=True, append_images=imgs[1:])