Data

Data tab

Either load existing data, load test data, or simulate your own. Simulation and testing is important to get a better feeling of the limitations of the method and algorithms. (Experimental data or simulations are piped through the same processing and analysis machinery. The machinery does not know the expected result). A text box shows the vital parameters of the currently selected data (ns/pt, duration, phase estimate, etc.) displayed on the preview tab.

By default, the magnitude and the RE data are shown, but any combination can be selected.

By default, the phase is corrected automatically and RE and Magnitude should look nearly identical. Turning phase correction off allows inspecting the raw data unchanged. When phase correction is done manually (not recommended), make sure that the imaginary part (red) is also selected for display, it might contain the bulk of the data if the phase is way off.

Experimental Data

To work with experimental data, you first need to browse to a folder containing the files. You can select a folder or a file. All files matching the selected pattern from the folder will be loaded into the list and one file selected. There is a preview window to see the data before analysis. Select any file in the list to see the preview. The file name might also have a check mark next to it, meaning it has been analyzed at least once already. In that case, the small window below shows all previous analysis results of the selected data file. Double-click any file in the list to load it.

By default, the program reads Bruker ELEXSYS files containing quadrature data and requires both files present (*.DTA, *.DSC). The DTA file contains the binary data and is required. The DSC file contains information about the point spacing and data format. If the data is 2D, it is assumed that the second dimension is due to pulse variations (e.g. ESEEM suppression) and all traces are averaged automatically. The following fields are read from the DSC file (XPTS, YPTS (if 2D data), XMIN, XWID). If the DSC file is missing, the user is prompted for the point spacing and all bets are off if unreasonable values are entered. (not recommended!).

Optionally, you can also load plain ASCII files (e.g. results of an external simulation or home-made instrument). In this case select ASCII in the lower right. Columns can be tab delimited (ASCII) or comma delimited (ASCII CSV) with period as decimal delimiter in either case. Two columns are expected. First column is the time in microseconds and the second column is the real part of the signal (quadrature ASCII data is currently not supported. The ASCII data is expected to be phase corrected and the imaginary part is assumed to be zero). If your data does not match this format, use an external program to convert it first. To selectively display only ASCII files with a matching pattern in the list, enter the correct pattern in the field provided. (For example, a pattern of "LZM*.dat" would only show .dat files starting with LZM).

Simulated Data

For simulated data, we have two options ("test grid" or "distributions") as selected in the model tab. A test grid is a set of equally spaced peaks of the same shape and width as defined on the test grid tab. A distribution is a sum of up to six peaks, each with defined shape, position, width, and area. Only components activated with the green LED are used. There are general options such as the number of dipolar evolution (DE) points, zero shift position, point spacing, as well as noise. If you simulate very sharp peaks, you might want to increase the number of distance points and adjust the limits for the simulation (Note that this is only used to simulate the dipolar data and is independent of any such setting in the later analysis. The simulation settings don't need to agree with the later analysis setting and you actually might want to make them different to reduce coherence).

Sometimes you might want to use simulation settings that reflect real data. Once you have loaded real data, you can go to the settings tab and use the "simulation setup" to copy and use the real settings (e.g. number of DE points, point spacing, etc.) for the simulation.

Test Data

An extensive set of synthetic data has been published by Edwards & Stoll in 2018. All 5622 datasets are are available for selection. This data contains no background and t0=0. The noise seeds from the paper are not used, instead noise is generated randomly with each load. A reasonable selection of Gaussian noise values can be selected. There is also a secondary control for alpha to quickly inspect the preliminary analysis. For very noisy data, results look better with a higher alpha.

LabVIEW™ Is a trademark of National Instruments Corporation.