The flux variation gradient technique has enabled the determination of the nuclear colours of many well-known AGN by comparing the total galaxy fluxes at different epochs. It is now established that the intrinsic AGN nuclear colours in the optical are relatively constant throughout the class. This means that measured deviations from these colours are an indicator of partial obscuration and the resulting reddening of the nucleus. The degree of nuclear reddening determined in this manner therefore enables the estimation of the light loss due to circumnuclear matter. From this the intrinsic nuclear luminosity may be calculated. The author is engaged in a long-term observational campaign that seeks to secure photometric magnitudes of a large sample of comparatively rarely studied Seyfert galaxies during two epochs spaced apart by a period of typically 20 years. Initial results confirm that most objects varied sufficiently between the two epochs to compute the flux variation gradients, and hence to determine their intrinsic optical luminosity. This paper discusses the analytical processes necessary to maximise the accuracy of the nuclear luminosity determination. This includes the description of more sophisticated procedures to improve the nuclear component's deconvolution and reddening, and the observational tests performed for this purpose.