3D Data Processing
(New with Version 990725)
We have tested this only on a single 3D data set from a Bruker spectrometer. Users with data from other systems are encouraged to supply sample data sets for us to test.
Processing of 3D data is done in a new 3D Arrayed Mode. Many of the commands necessary for 3D processing are non-2-letter commands, so the user must enter this mode with the 2F command. In this mode, commands are not executed until <Enter> is typed.
Because the import routines do not recognize 3D data (yet), the data is imported as though it were 2D data. Thus it is necessary to change parameters to define the number of data points in each dimension. This is done with the new dims command. Typing just dims <enter> will print out the number of data points in each dimension. To change the values, type
dims i j k l
where i=number of (complex) points in the first dimension, j=number of (complex) points in the second dimension, k=number of (complex) points in the third dimension and l=number of (complex) points in the fourth dimension (which for now is 1).
The total number of points (i x j x j x l) must remain unchanged, or the process will abort.
Acquisition parameters can be viewed for all 3 dimensions by selecting Spectral parameters from the View menu.
Processing is done in a manner very similar to 2D. The main difference is that when data are transposed for processing and display along a different dimension, it is now necessary to specify which dimension will be transposed. A general procedure is described below, and specific steps for one particular data set are described here.
Processing normally begins by opening the data file and examining the first slice, setting parameters for window functions, doing FT and phasing. Then the 3D arrayed mode can be entered. To do this, first execute 2F to change to non-2-letter command mode, then type
AR 3D <enter>
This will load the entire 3D data set, so will take several seconds. Now, each command will be executed on the entire data set, not just a single slice. For example,
will apply sine apodization, FT and phase with previously determined phase correction. The appropriate steps needed will depend on the particular experiment, as it does with 2D.
Next, we need to transpose the data to process in the second dimension. The TD command now requires an argument to specify which of the other 2 dimensions will be transposed with the current dimension. So
will display interferograms along the second dimension, and processing proceeds in essentially the same manner as 2D data.
As the different dimensions are transposed, it is necessary to keep track of which is which. View/Spectral Parameters will display 3 columns of parameters, one for each dimension. The user can change the label for each to something more meaningful. The first (left-hand) column always corresponds to the currently “active” dimension, whose points are displayed from left to right. The second (middle) column is dimension 2, which are the different slices. A contour plot or stacked plot will display dimension 1 horizontally and dimension 2 vertically on the screen. The third column of parameters is for dimension 3, which can be visualized as a stack of 2D planes.
At the moment, Nuts has no way to display the third dimension. Instead, a particular plane is chosen and a 2D plot (contour or stacked plot) of dimensions 1 and 2 is displayed. While in the contour or stacked plot display mode, the sequence of planes can be stepped through using the [ and ] commands, which (like all commands in subroutines) are executed immediately, not requiring <enter>. A particular plane can also be specified with the plane command, for example:
SUM — Sum planes
The sum command can be used for summing of planes in a 3D data set in a manner analogous to summing slices of 2D data. When the second argument of the “sum” command is “planes”, all planes of a 3D data set are summed into a single 2D data set. The command also takes additional optional arguments allowing the user to specify the starting and ending planes to sum. The original 3D data set is destroyed and replaced by a new 3D data set which has only one plane. The command must be used in the 3D arrayed mode.
sum planes <start> <end>
A step-by-step example of processing a 3D data set is given here.
Last updated: 12/22/04