Alan Fisher asked me (greg) how hard it would be to history buffer the
chromaticity (the relationship of the energy change to the tune). If you
were interactively using correlation plots you'd 1) step the master
oscillator and fit tune to the change in freq (both of which are available
in Correlation plots, and the correlation plots software already has
facilities for locking the master osc, stepping, sampling, relocking, and
converting freq to energy and vice-versa). 2) Then you'd ask CP to do the
fit. 3) You'd write the displayed fit slope param in the log book.
Since one can't screen-scrape the fit parameter off the screen in order to
history it, this is where his request is more involved than just a button
macro.
I suggested there were 2 possible solutions:
a) Write a standalone and use the public interface API of Correlation Plots
to make the acquisition and do the fit. I believe the API includes a way to
retrieve fit parameters. Then put the fit parameter somewhere in the db
where it will be historied. The standalone would be run periodically,
probably as part of slow-feedbacks.
b) Write a button macro that does steps 1 and 2. Use the same "techniques"
as were used in the GOF overhead display for running periodically, and
communicating values from a running SCP with logical names. Again put the
value somewhere in the db.