In [1]:
%pylab inline
import warnings
import numpy as np
import matplotlib.pyplot as plt
import rayopt as ro
# ignore matplotlib and numpy warning each other
warnings.simplefilter("ignore", FutureWarning)
# ignore floating point exceptions
np.seterr(divide="ignore", invalid="ignore")
# by default only print 4 significant digits
np.set_printoptions(precision=4)
In [2]:
# It appears that rayopt requires distance rather than
# thickness quoted in patent literature
# So we have to move the thickness value one row down
#
# The first surface represents the Object (O)
# The last surface represents the Image (I)
description = "Nikkor-Z-35mmf1.8-JP2019090947A"
columns = "type distance roc diameter material"
# Following is original data
# from patent with one correction of distance
# The data is scaled because the focal length as per the
# patent is only 1.57 mm.
text = """
O 20 0 3.0 AIR
S 0.5 4.6232 1.664 1.51680/64.13
S 0.102 0.8784 1.346 AIR
S 0.354 4.0730 1.314 1.95375/32.33
S 0.258 -2.0290 1.28 1.60342/38.03
S 0.083 3.2358 1.138 AIR
S 0.2595 -1.2945 1.096 1.68893/31.16
S 0.172 1.7517 1.262 1.85150/40.78
S 0.329 -1.7517 1.29 AIR
S 0.018 1.1917 1.306 1.49700/81.61
S 0.368 -3.5528 1.262 AIR
S 0.009 4.1556 1.174 1.83441/37.28
S 0.057 1.9836 1.114 AIR
A 0.187 0 1.066 AIR
S 0.514 -0.9917 0.956 1.61293/36.94
S 0.062 -5.7790 0.99 AIR
S 0.012 2.1244 1.088 1.59282/68.62
S 0.282 -1.3918 1.124 AIR
S 0.209 -8.4171 1.198 1.69350/53.20
S 0.098 -2.8176 1.228 AIR
S 0.229 -1.8973 1.29 1.58313/59.46
S 0.083 -6.9348 1.408 AIR
S 0.751 0 2.4 1.51680/64.13
S 0.074 0 2.4 AIR
I 0.0425 0 2.0 AIR
"""
# Following data is scaled by a factor of 21.6
# to get to an image radius of 43.2
text = """
O 432 0 64.8 AIR
S 10.8 99.86112 35.9424 1.51680/64.13
S 2.2032 18.97344 29.0736 AIR
S 7.6464 87.9768 28.3824 1.95375/32.33
S 5.5728 -43.8264 27.648 1.60342/38.03
S 1.7928 69.89328 24.5808 AIR
S 5.6052 -27.9612 23.6736 1.68893/31.16
S 3.7152 37.83672 27.2592 1.85150/40.78
S 7.1064 -37.83672 27.864 AIR
S 0.3888 25.74072 28.2096 1.49700/81.61
S 7.9488 -76.74048 27.2592 AIR
S 0.1944 89.76096 25.3584 1.83441/37.28
S 1.2312 42.84576 24.0624 AIR
A 4.0392 0 23.0256 AIR
S 11.1024 -21.42072 20.6496 1.61293/36.94
S 1.3392 -124.8264 21.384 AIR
S 0.2592 45.88704 23.5008 1.59282/68.62
S 6.0912 -30.06288 24.2784 AIR
S 4.5144 -181.80936 25.8768 1.69350/53.20
S 2.1168 -60.86016 26.5248 AIR
S 4.9464 -40.98168 27.864 1.58313/59.46
S 1.7928 -149.79168 30.4128 AIR
S 16.2216 0 51.84 1.51680/64.13
S 1.5984 0 51.84 AIR
I 0.918 0 43.2 AIR
"""
In [3]:
s = ro.system_from_text(text, columns.split(),
description=description)
s.object.angle = np.deg2rad(32.5)
s.fields = 0, .7, 1.
#s.wavelengths = [587.5618e-9]
# following are the original aspheric coeffcients
s[11].conic = -4.9288
s[11].aspherics = [0, -9.582E-02, 5.043E-01, -4.618E-01]
s[12].conic = -0.4693
s[12].aspherics = [0, -8.355E-02, 5.689E-01, -2.913E-01]
s[18].conic = 15.3255
s[18].aspherics = [0, -2.063E-01, 6.890E-02]
s[19].conic = -0.9347
s[19].aspherics = [0, -2.416E-03, 1.158E-01, 1.983E-01, -1.130E-01]
s[20].conic = -0.1889
s[20].aspherics = [0, -1.143E-01, -1.549E-01]
s[21].conic = 0
s[21].aspherics = [0, -9.359E-02, -1.873E-01, 1.909E-01, -1.298E-01]
# Following is scaling done by hand in the same way that
# rayopt does it
scale = 21.6
s[11].aspherics = [ai/(scale**(2*i + 1)) for i, ai in enumerate(s[11].aspherics)]
s[12].aspherics = [ai/(scale**(2*i + 1)) for i, ai in enumerate(s[12].aspherics)]
s[18].aspherics = [ai/(scale**(2*i + 1)) for i, ai in enumerate(s[18].aspherics)]
s[19].aspherics = [ai/(scale**(2*i + 1)) for i, ai in enumerate(s[19].aspherics)]
s[20].aspherics = [ai/(scale**(2*i + 1)) for i, ai in enumerate(s[20].aspherics)]
s[21].aspherics = [ai/(scale**(2*i + 1)) for i, ai in enumerate(s[21].aspherics)]
print(s[11].aspherics)
print(s[12].aspherics)
print(s[18].aspherics)
print(s[19].aspherics)
print(s[20].aspherics)
print(s[21].aspherics)
In [4]:
s.update()
print(s)
In [5]:
ro.Analysis(s)
Out[5]:
In [ ]: