Measurements of my photometric filters

American Association of Variable Star Observers (AAVSO)
Thu, 01/12/2023 - 03:55


If anyone would like the graphs and photos from this article, just email me via the contact here or attempt this URL:

(1) Measurement of Photometric Filters | LinkedIn

New photometric filters were purchased. Some photometric filters had been reported as having out of band leaks. I wanted to measure the actual response of my filters.

An inexpensive LD-721 spectrophotometer was used to obtain transmission curves for Baader Bessel BVRI and Baader Sloan g’, r’, i’ filters. The measured curves for my filter set are all nominal. They look very much like those posted earlier by Mike Wiles.

Baader also specifies the orientation of filters. The Red filter was measured both ways to look for a transmission difference. No significant difference was noted.

The LD-721 spectrophotometer wavelength range is 350 to 1020 nm. That closely matches the wavelength sensitivity range of CCD devices. Silicon CCDs have almost no response to IR leaks beyond about 1000 nm. The LD-721 is inexpensive at less than $300. It is a single path device. That means that the sample has to be slid out to calibrate and slid in to make a measurement. Wavelength selection is done with with an analog knob which rotates the grating. The grating is 1200 lines per mm. Sample positioning is done with a notched 4-position slide that is accessed from the front. The sample normally rides in a four-position holder. Samples are normally placed in 1cmx1cm cuvettes and a black cuvette is supplied for zeroing.

Standard spectrophotometers measure absorption, transmission and concentration. Online tutorials for absorption and concentration are abundant. Absorption is commonly used for chemical and biological samples and is not used here. Molecules absorb certain wavelengths and can be identified.


This measurement is transmission of optical band-pass filters and the transmission mode is used. In transmission mode a white light source is directed to a grating. Wavelength is selected by rotating the grating so that a single wavelength is focused on a slit. The selected wavelength passes through the photometric filter being tested. A photodiode measures how much light is transmitted through the filter.

The cuvette holder does not hold astronomical filters. A filter holder was cut out of display-board and the filter-in-holder was taped to the side of the cuvette holder so that two slots are covered by the filter, 

one slot is open, and one slot is blocked by the inserted black plastic cuvette. Later a 3D printer was used to make filter holder that would mount two sizes of filters. The filter holder is taped to the side of the cuvette slider. The instrument is allowed to stabilize for 30 minutes after powering on.

The white light here is a tungsten lamp. It does not have the same luminosity from 350 to 1020 nm. The photodiode does not have the same response over the 350 to 1020 nm range. The spectrophotometer has a handy way of correcting for those shortcomings.

To make a measurement, %T (percent transmission) mode is selected. A wavelength is dialed in with the grating knob and read from the dial. The cuvette slider is pushed in so that the black cuvette blocks the beam path and the 0%T button was pushed so that the display reads zero. Then the cuvette slide is slid to the open slot so that the beam path is unobstructed. The 100%T is pushed and 100% is displayed. At this point the thing is calibrated at 0 and 100% transmission. The amplifiers seem to be linear over the range. The filter is slid in and the percent transmission is read and recorded on paper. The sequence is repeated every nanometer. The result is 1020-350 or 670 calibrations and 670 measurements. In practice the zero can hold for 5 to 10 rapid measurements depending on the tungsten light curve around the point. The 100% calibration tends to be done for every wavelength The results are then typed into a spreadsheet for plotting. The spectral bandwidth is 6 nm due to the slit width, but with careful technique, 0.2% transmission peaks can repeatedly be defined over a 3 nm band with this instrument.

The manually operated LD-721 spectrophotometer is an inexpensive (about $300) way to check optical filter pass-bands and is available for purchase online. I don’t use U-band, but other units extend the measurement band further into UV and IR.