Baseline correcting digitally filtered data

An artifact of digitally filtered data is an abrupt curve at the ends of the baseline.  

A tilted baseline, as seen above, is normally easily removed with the BC command, which removes DC offset and linear tilt by assuming that the ends of the spectrum are zero.  However, for spectra such as this, the ends are not zero and using BC results in a spectrum whose baseline is not truly at zero, as shown below:

The BR command performs an operation similar to BC, but disregards the ends of the spectrum, so that the correction is based on points that are really zero.  The result, seen below, is a much better correction.

By default, the BR command (or equivalent non-2-letter command Bruker) ignores 2% of the data points at each end.  A similar command, JE (for Jeol) ignores 5% of the data points at each end. 

The command has some additional options that can be invoked in the non-2-letter command mode. 

If the number of distorted points doesn't match the 2% or 5% defaults for the BR or JE command, it is possible to specify the number of points on each end to be ignored.   This requires 2 arguments: the first is "pts" and the second is the number of points to be ignored.  For example, to ignore 350 data points at each end, the command would be

br pts 350

It is also possible to replace the distorted end points with zeroes.  This requires 2 arguments: the first is "ZeroPts" and the second is either 1 (to turn ON zeroing of the end points) or 0 (to turn OFF zeroing of the end points).

To choose to zero the ends points,

br zeropts 1

This command does not change the spectrum, it simply selects the zeroing option for subsequent BR commands.  The result of the BR command with zeroing turned on is shown below:

Once zeroing of the end points is turned on, it remains on for subsequent BR (or JE) commands until turned off.

It is also possible to discard a specified number of data points at each end of the spectrum, using the /2 command.  

Last updated: 3/9/04