4: SDIF matrix representation
Columns of an SDIF matrix are generally called the fields, and rows the elements of that matrix. Typically, fields represent the different parameters of a description (e.g. frequency, amplitude, phase for an additive type description).
SDIF-Edit allows the visualization of a data field deployed according to the elements of the matrix which compose it (in one dimension) and in time (in the other dimension). The height (vertical axis) of the points, as well as its color, determine the value of this field at the moment and at the position given by each point of the matrix flow. The data fields section on the interface allows you to select this field, using the mouse or the home and end keys of the keyboard. It is this field that is then also considered in the cross-sectional views, the sonogram, and especially the data editing. In addition, the values of the point selected in the data for each of the other fields are displayed in this part.
With the X-Parameter button, it is possible to observe this selected field as a function of another field in the matrix (e.g. amplitude as a function of frequency). If this second field does not follow a linear evolution, the data grid may be more or less distorted. In some cases, however, this may provide a more accurate representation of the data. It may then be appropriate to use the grid/lines button to hide this distortion.
On the figures below the matrix is displayed according to a non-linear parameter as grid and as line.
For an even more complete representation, it is also possible to choose a field different from the selected field for the color gradient (see Colors).
The SDIF format being an open format, it is likely to contain data of very diverse nature and form. Thus the matrices can be of variable dimensions, or null, for example. These particular cases will be of more or less interest, but some will remain exploitable.
A restriction of SDIF-Edit is that all matrices of the same flow must have the same number of fields (columns). The case of matrices with a variable number of lines is supported.
In case there is no time dimension, the viewer behaves in the same way but represents a 2D curve in a 3D space. The various functions are sometimes restricted but remain usable in the majority of cases. Matrices of dimension 1 work in a similar way, following the evolution of an element value over time (see figure below).
