Magnification calibration

Comparison

linear calibration for a 512×512 pixel image, nm/pixel
magnification displayed on the microscope and in Digital Micrograph	calibration displayed in Digital Micrograph	extrapolated calibration
450 000	0.4103	0.40
340 000	0.543	0.53
245 000	0.7535	0.74
180 000	1	1.00
130 000	1.4	1.38
92 000	2	1.95
64 000	2.9	2.80
46 000	4	3.88
34 000	5.4	5.24
25 000	7.4	7.12
19 000	9.7	9.35
13 500	13.7	13.12
10 500	17.6	16.84
7900	23.4	22.34
5800	31.8	30.37
4600	40.1	38.24
3400	54.3	51.63
2600	71	67.40
1950	94.7	89.69
1100	167.8	158.39
800	230.8	217.33
620	297.8	279.95

calculate a scale bar using either calibration at 130 000 ×
image size (pixel)	scale bar (nm)	with calibration displayed in Digital Micrograph (pixel)	with extrapolated calibration (pixel)
512×512	50	50 ÷ 1.4 = 35.7	50 ÷ 1.38 = 36.2
1024×1024	25	(1024 ÷ 512) × 25 ÷ 1.4 = 35.7	(1024 ÷ 512) × 25 ÷ 1.38 = 36.2

specimen: a grating replica with 2160 line/mm or a "d" spacing of 462.9 nm, example on the right is taken at 10 500 ×
eucentric height was adjusted at 10 500 ×
512×512 pixel images were acquired between 130K and 1100 ×, 100 KV and emission 3
use log-log plot of magnification vs object size (physical dimension of the image as calculated using the "d" spacing) to obtain the slope and intercept of the relationship

dimension of a 512×512 pixel image at various magnifications, nm
displayed magnification¹	displayed dimension²	actual dimension³
130 000	725.68	694.05
92 000	1025.42	991.74
64 000	1474.04	1427.87
46 000	2050.84	1994.85
46 000	2050.84	1960.73
34 000	2774.66	2794.12
19 000	4965.18	4946.30
10 500	8984.62	8693.11
5800	16 265.25	15 174.07
3400	27 746.61	26 237.66
1950	48 378.70	44 898.22
1950	48 378.70	45 025.48
1100	85 762.24	82 003.15
1100	85 762.24	82 899.21

¹magnification displayed by the microscope, Digital Micrograph call this the indicated magnification

²values displayed by Digital Micrograph i.e., from the previous calibration; dimension = 10^(7.9747+(log(magnification)*-1))

³actuals were calculated from the grating replica images, dimension = 10^(7.9302+(log(magnification)*-0.9934))

A log-log plot of the image dimension vs displayed magnification

log-log plot of the image dimension vs displayed magnification showing the displayed dimension are very close to the actual ones

the actual values calculated using the linear regression of the log-log plot are at least 95 % of the value displayed by Digital Micrograph
the linear calibration is dimension ⁄ 512 nm/pixel and is half for a 1024×1024 pixel image