Charles is quite right, there is often a lot of power in analogue filters that the new generation forgets - you can do things with a filter and a '52 that a high end DSP would struggle with - and I have done a commercial project where I did just that.

Then again, when your data is received entirely in the digital domain, analogue techniques are not appropriate, and digital is the only way to go. I have an instrument design that requires the highly filtered result of dividing two frequencies together, where digital filters extract only the signal of interest in horrendous noise. The nearest analogue equivalent would be the lock-in amplifier or Phase sensitive detector.

Then too, there are highly useful filter functions that are impossible to duplicate in the analogue world, however useful they would be, like the IIR filter.

And while there are some wonderfully elegant fourier analysis techniques in the analogue domain, I am thinking of the optical methods , sometimes nothing can beat the FFT and the inverse FFT.

Its "horses for courses". There is no rule that says "thou must use digital filters", nor one that says you must use analogue ones !


Steve