nts dn
ny
ljust-
same
cases
hoto-
uted.
. The
thin
ffer-
or of
hs on
ilters
The
rated
Table 13. Standard error of unit weight of image coordinate measurements in
microns.
Camera: R. C. 5 Aviogon.
Image Filter nis Radii
| eno Le | 8/1 215 a 13515
5 -€ 541 2x | ael sol. 5| ss] oo] si] —
20 — 827 191207 41 31! 5: 4.91 6831 —
9 520 my, P21 271 414 371 16! a3 8&9] 7.0! 95
10 520 46 | 221 151 8.7 |. 30, 68] *1! 96| 129
11 = 51 / 18) 32 | 382| 49[ 54l 48[| Di
1 520 12/26/37! 50! 40/ 90! 80/ 140 —
2 5 12/19} 23 / G1 48/49) 541 977119
3.161 : 25|.56| 12| 60| 19] 82] 88| s1| s»
17 | orto 417 2015.21 48 | 108. 100) - 864 — —
18 08 (28 1.26.4 18 | 5 | 4.0] 11.0! 11.0) 104
[ pipe m | |
Averages | [8584| 26161] 40[ 50| 57) as] SI! ug
|
The average of the averages is 5.3 microns. This figure can conse-
quently be used as expression for the accuracy of the image coordinates
in films which are photographed under present circumstances and when
the image coordinate measurements are performed in a projector of the
autograph A7 no. 310. Evidently there is a clear increase of the standard
errors with the radii. This means that the weight of image coordinates
under the circumstances mentioned decreases with increasing radii. The
same tendency was also found for plates according to the correspon-
ding Table 8 above.
Some photographs were also taken with a filter made of simple glass.
The accuracy of these photographs was found to be much lower than of
those which were demonstrated above. The standard errors of unit
weight of one of the photographs taken with this filter are demonstrat-
ed below. The standard errors in the upper row were computed with
seven and in the lower row with six parameters.
Table 14. Standard error in microns.
Camera: R. C. 5 Aviogon.
Image | gn Sh Del Radii c
| Bo cL ook 2$. 1 23:4: 4 115 | 617 | 5
| |
Turc m e | 110 | 13.0 | 190 | 290 | 420 | 650 | 65.0 |
| 4 | 926 | 100 | 100 | 180! 290 | 460 | 700 | 710 |
The disastrous effect upon the accuracy of the image coordinates of
a bad filter is quite evident.
41
