TEST of my new Singh-Ray Vari-ND (variable neutral density) filter. Advertised to let you "continuously control the amount of light passing through your lens by as much as 8 exposure stops - with no loss of color fidelity or sharpness".

"EXECUTIVE SUMMARY" - My results show that 7 stops (or maybe 7 1/3) is the filter's maximum - not 8. More importantly, a noticable color cast appears at about 4 stops, is unacceptable by 6, and by 7 the true colors of the subject are lost TOTALLY - see sample images. I can live with 7 instead of 8 stops but NOT with the color cast. A final observation - my D70 (& D70s) metering is almost useless with the filter attached; manual exposure control is the only way I can can proper exposure.

TEST METHOD-
1. Constant light level and constant color (closed room using daylight color temperature incandescent bulbs as the sole light source).
1a. Shot each shot with a WhiBal gray card in the scene in order to precisely measure & control white balance
1b. Made each shot (RAW) using manual exposure to precisely control the camera's exposure (stops)

2. Made five shots:
2a. One with no filter - base reference shot
2b. Four filtered shots with the camera exposure changed by 2 stops, 4 stops, 6 stops and 7 stops (filter could not hold back enough light for 8 stops)
2c. For each of the 4 shots using the filter the camera exposure was set appropriately for the specified number of stops and then the filter was adjusted to keep the resulting exposure constant from shot to shot (as measured by the camera's histogram and the viewfinder's over/under exposure meter - that is, each shot was made so that the histograms and over/under meter from shot to shot were the same - see image of 5 histograms below).

3. The RAW shots were processed in Nikon Capture NX. The only processing was to adjust the white balance of each shot to match the base reference gray point setting as determined by the WhiBal card. WB was set in NX by using the measured gray point of the base reference shot (save setting). Each filter shot was set to this same gray point setting (apply saved setting).

RESULTS - Click on each thumbnail to see embedded exif data for f/stop & shutter speed - as well as the amount of filter ring rotation needed to achieve the specified exposure stop setting. I don't know if there is variablility from one filter sample to another. If there is, I've got a bad one. If not, then I don't understand the basis of the advertised performance of this filter. I am a retired as a Chief Engineer for a large technical organization and an award winning advanced amateur photographer. I could have made an error in testing, but it's not obvious to me how/where. These tests were repeated with four different lens/body/lighting combinations - all with the same results. Looking for insight from Singh-Ray, please.