Lpc2000

>> One possible solution is to use DSP. Supersample the data (insert >> 
0's) until you have the proper resolution then run it though a low
 >> pass (interpolation) filter and pull out the value you need. Might be
 >> more compuationally intensive though.
 >>
 >> I think regression would be a bit of an over kill.

Actually it doesn't have to be computationally intensive. If you use a 
FIR filter (eg a windowed Sin(x)/x function) you will only have to 
calculate the output for the single point that you want. In additon , 
most of the input samples are zero - so there is no need to include them 
in the computation. - instead you just resample the windowed FIR 
coefficients to match the origional sample rate of the data. The 
computation is not to bad at all.

So you see that some of us with grey hair have some uses after all.


OLD FARTS RULE !!!!!    :-)

Owen Mooney