During hollidays in 2006 july, a few tests have been conducted to get further information on some filter's transmission. The measured curves are indispensable but direct photometric measurments bring operational values taking into account every single elements, especially the CCD quantum efficiency.

The idea of these tests emerged from P. Rousselle's evaluation of the true transmission of the W47 violet filter, which proves that the filter is highly transparent in near-UV light, and quickly blind in blue light. The filter is therefore extremely interesting for Mars imaging, but requires another filter to block the strong infrared leak that is also its propriety ! Although an IR-blocking glass looks to be the most evident solutions, some others can be considerated. Thanks also to Arnaud van Kranenburg who sent me some very stimulating analisis about UV wavelenghts !

Two series of tests have been realized. Each time, the star Vega was used as a target. Its brightness, high altitude in the sky, and its white color designed it as a valuable tool.

1° UV transmission of various filters (11 july 2006)

The idea of this first series is to evaluate how much UV light each filter is able to transmit.

Method :
A movie file is taken with the Lumenera LU075M equiped with filter various filters. 50 to 60 raw frames are aligned and the 10 best are stacked. The settings of the camera are kept identical through the process.
The first image is taken with the Lumenera and the Schuler UV filter on. Then, the other images are realized through each filter plus the U filter still in place. As a result, a coefficient of UV exctinction is recorded in comparison with the first UV only shot.
Finally, the intensity of recorded light is measured using IRIS's photometric tools.

Results :

Comments :

As expected the W47 and the Fuji SP-4 present high transmission in near-UV, but the Schuler P filter gives a sligthy disappointed value. It should have given the same result than the W47, according to the theorical transmition curve.
The Astronomik L and Schuler Bu have acceptable loss of transmission of UV wavelenghts. The Astronomik B filter gave the lowest UV transmission. This is also surprising because its transmission characteristics before 400 nm have been found to be almost identical to the Astronomik L filter, with a weak first peak near 340 nm.

The firsts conclusions propose the Astronomik L, Schuler P and Bu filters as the most interesting IR-blocking filter to be used along with the W47. This must be verified with a second series of tests.

2° Real transmission of light with the W47 equiped with various filters (13 july 2006)

This test was conceived to measure how much light is obtained through the W47 with various IR-blocking filters.

Method :
Same as above, but with the W47 instead of the UV filter as main directing glass.

Results :

Comments :
The use of the L filter as the IR-blocking glass gives the brighter image with the W47. The hierachy of filters obtained with the first test is not confirmed as the Astronomik B gives a comparably bright image. In the meantime the P filter which had a good UV transmission present a slightly noticeable loss of light. The Bu filter gives a relatively too dim image to be valuable.
This test shows how reduced is the UV part of the image obtained under real observing conditions. It's so unimportant comparatively to violet and deep blue light (390-440 nm) that the blindest filter in UV, the B filter, is almost the one that now gives the brightest image.

CONCLUSIONS

These two tests design the Astronomik L filter to be the best IR-blocking glass to use the W47 as a violet filter ; the 2nd test proving that differences in UV transmission are not relevant (one would wish to have the best UV transmission possible in violet light). A quick evaluation of how much UV light (choosing wavelenghts <390 nm where the U filter is most sensitive) is present when imaging with the W47 would probably gives a result of around 10 % or little more.

The application of the W47+L filter is above all Mars imaging ; it gives high contrast for white cloud while providing a strong blocking of surface details (except during opposition period when they're more easily visible even in UV). The weak UV part however confirms that the filter is much less efficient than U filters for Venus imaging, although it does reveal the same details.

The test also suggests to me that the SP470 is less transparent than the constructor claims. A peak of 70 to 75 % instead of 85 % would probably be more realistic.

RETURN TO THE FILTER PAGE