>
>
>  I want to do some analysis on several data sets onsisting of 32768 data
> points per file.
>

It would probably be extremely useful if you could define the limits of the
step signal as tightly as possible. This may not be very tight at all, but
knowing the worst case that is being looked for may help choose a best
solution.

I did something similar long long long ago with good results. That was to
detect telephone tones (or not) in the presence of noise and speech.

In your case aspects such as the minimum and maximum step sizes, min an max
step duration, extent or normal variations during internal part of step,
reteptition rate, mim/max period between occurrences, ... anything else
relevant  ... anything else irrelevant :-) ... may help. Worst case you may
say step may be minimally small through immense and minimally long through
most of sample - but even that helps set bounds and helps suggests
solutions . (eg - go away :-) ).

As ever, actually knowing what you are actually trying to actually do would
probably actually be useful. But, most people seem reticent about actually
disclosing such matters :-). (I also sometimes).

              Russell
--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.