Surface Roughness - Filter Wavelength not equal to cutoff length.

[Jacob Cheverie]

Everywhere I've seen states that the wavelength of the filter is equal to the sampling length. Is this a firm requirement? We are measuring spherical surfaces and we can't quite get the amount of surface measured that we would need to use the filter that matches the roughness requirement. My idea was to measure as much of the surface that we can, and then apply the filter at whatever wavelength we want to for roughness evaluation.

For example, 5 sampling intervals of width .010 are measured with two .010 intervals at the ends to account for end-effects. The evaluation length is .050. Would it be wrong to apply a filter of wavelength .003 (or anything other than .010)?