Support Center Email Support Knowledge Base Videos
Geoscience > Main Launcher > Seismic Menu > Wavelet Analysis

Wavelet Analysis

Wavelet Analysis is a tool that will perform a comparison between multiple wavelets and the seismic data. There are two methods of comparison, cross correlation of the data or a Manhattan distance calculation. The tool can generate many forms of output, but the ones that are of most interest are the best and worst CC results. Each wavelet is compared against the seismic data, then the correlation value for the wavelet that best (or worst) matched will be saved. Users may also output the ID of the wavelet that best (or worst) matched, as this output may be easier to understand. Individual results are the same as the results generated in Seismic Zone Attributes.

ClosedHow do I perform a wavelet analysis?

  1. Select the Seismic Lines that you will perform the analysis on. You can select them from the list by clicking on them, holding Ctrl and selecting multiple lines, or selecting an area on the map to use all lines within that region. Note: If using a time window only ensure the data has been tied.

  2. Specify either the Correlation coefficient or Manhattan distance algorithm to be used. Correlation will take wavelet shape into account but not amplitude. Manhattan distance will take into account both wavelet shape and amplitudes.

  3. Define your Seismic Window by using one of the windowing methods. This is the zone that all wavelets will be compared over, and will also be the default when selecting wavelets.

  4. Give your output a Base Horizon Name that will be used for display on the map. 

  5. Select the Output Results you would like generated. As a minimum, Output best result should be selected with custom color palettes and this will give you a colored composite map with all wavelets. You can review the other options here. Note: Do not select Use absolute value if you would like to take phase into account. Without this option checked wavelets that are 180° out of phase will correlate perfectly. 

  6. (Optional) Select an Optional Constraining Polygon to limit the geographical extents of the analysis.

  7. (Optional) Set Cut-off values and select a  Cut-off Color which will be applied to any results outside of the cut-off range.

  8. Click  to select your wavelets. These steps can be repeated as many times as necessary, but make sure you select a different color for each wavelet.

    1. Click and select your wavelet from either the Basemap or Seismic Viewer. 

    2. Select your windowing method. Note: Typically the time window of the selected wavelets should match your seismic window for better results, so select Use Seismic Data Window.

    3. Click .

  9. Click   to ensure the wavelets selected are correct and do not correlate too closely. Note: The more wavelets you have, there is a greater chance that some of the wavelets will correlate well with each other.  If this is the case you may end up with random looking results. 

  10. Click  to generate the results.  All results are output in a horizon format.  The results will be located under the horizon drop down on your Basemap but is not visible on the seismic.



General Information

Seismic Lines

Select the lines you want included in the wavelet analysis. These are the lines that your selected wavelets will be compared to. The working set version is listed, but alternate datasets can be selected by clicking on the down arrow in the Dataset cell. Left click to select a seismic line, or use the Ctrl key on your keyboard to select multiple seismic lines. Alternately, you can select a group of lines from the Basemap.

To help you make your line selections,  you can sort any column by left clicking on the column header, you can apply a filter (), or you can type a search string into the search bar to limit the list using Line, and Dataset information. 

For additional selection and navigation options right click on the Seismic Lines list to access the shortcut menu.

Algorithm To Use

Correlation coefficient: This compares the wavelets picked in the frequency domain to output a cross-correlation value. The results are shape based and do no take amplitude effects into account. Note: Do not select Use absolute value if you would like to take phase into account. Correlation coefficient is calculated using the formula:

cc-formula.png

Manhattan distance:This is a statistical comparison performed using the amplitudes of the selected wavelets showing similarity. It's the ratio of the sum of the differences divided by the maximum possible difference and then scaling the results to lie between 0: 100. A perfect match is 100% which means the wavelet have the same shape and same values along the waveform. Amplitude differences affect the output. Manhattan distance is calculated using the formula:

mhd-formula.png

Seismic Data Window

Select a windowing method to determine the range of data against which the selected wavelets will be correlated . Note: Offset values can be positive or negative. The value is added to the reference so a negative time will move above and a positive value will move below the reference.

Windowing Method: Select a windowing method.

Output Parameters

Base Horizon Name: The name to use as the base when generating output horizons. Thus, if the output base name is set to "channel", the output horizons will potentially be "channel_000", "channel_001", "channel_best", "channel_bestid", ...

Output individual results: This is the result of individual wavelets correlated against the Seismic window selected.  Therefore each wavelet select will have its own Individual Output.  This is very useful when checking to ensure the results are valid, and to understand the distribution of correlation. 

Output IDs:  The output is created by comparing the correlation results from all wavelets at a specific trace and then displaying the ID Color of the Wavelet that correlates Best/Worst to that trace. This option will generate an output horizon if "Output Best CC Result" or "Output Worst CC Result" is enabled.

Custom color palettes: A color palette will be created for each horizon. To use this feature, you should ensure that the map has "Automatically Save Color and Scale Information For Each Horizon" enabled. Then, when you switch to the output horizons, they will scale and color correctly. Note:  Often it is necessary to create these in order to properly view the results.

Output best result: Similar to the CC IDs but  will use the Output Colors option as well as the Fade Color To options to give a smoother appearance of your results. Note:  The results are displayed between 0-100.  To display a proper colorbar each consecutive wavelet will have 100 added to it.  Thus, for wavelets that best match wavelet ID 0, the values will be 0: 100. For wavelet ID 1, the values will be 100: 200. For wavelet ID 2, the values will be 200: 300....

Output worst result: Opposite of the Best results.

Use absolute value: Enabling this will cause advanced CC to take the absolute value of the results of the cross correlation.  This means that phase will not be taken into account in the correlation. This has been provided for backward compatibility. Prior to SeisWare 7.0, the absolute value was always used. Note: Typically phase is important and this option will be left unchecked.

Optional Constraining Polygons: Open the Select Polygons dialog to limit the Wavelet Analysis to a certain geographical area.

Output Colors

Cut off for Best Results: Normally, custom color palette will be scaled between 0 and 100 for each wavelet. However, the results for the best match is likely to have higher values. In fact, one could expect that the values outputted would always be 90: 100 if enough wavelets were specified. To allow for increased dynamic range, users may specify a cut off value of something other than zero (0). If you specify 50, then all colors will be scaled to 50: 100, and 0: 50 will be set to the cut off color. Note: This does not affect the values of the horizon, only the color palette.

Cut off for Worst Results: Normally, custom color palette will be scaled between 0 and 100 for each wavelet. However, the results for the worst match is likely to have lower values. In fact, one could expect that the values outputted would always be 0: 10 if enough wavelets were specified. To allow for increased dynamic range, users may specify a cut off value of something other than zero (0). If you specify 50, then all colors will be scaled to 0: 50, and 50: 100 will be set to the cut off color. Note: This does not affect the values of the horizon, only the color palette.

Cut off Color: The color to use for values outside the cut off ranges.

Fade Color To

Color palettes will be scaled between the fade color for the worst match and the wavelet color for the best match. Therefore if displaying your best results a 100 percent correlation will display as the ID color and a 0 percent match would display as the Fade to color.  Thus, if a wavelet is set to red, the palette will fade from black (or white) to red.

Parameters

Save: Saves the current configuration to an .xml parameter file.

Load: Loads a previously saved configuration from an xml parameter file.

Wavelets

The wavelets tables displays a list of the wavelets selected for analysis, as well as their parameters including color, name, range, and dataset. The wavelets can be enabled and disabled allowing you to quickly run an analysis with different combinations of wavelets to get the best results.

Add:Open the Add Wavelet dialog to add a new wavelet to the list.

Edit: Open the Add Wavelet dialog to edit the parameters for a selected wavelet. 

Remove: Remove a selected wavelet from the table. 

Remove All: Remove all wavelets. 

View Wavelets: Open theQC Wavelet dialog.


Also See

— MORE INFORMATION

Copyright © 2020 | SeisWare International Inc. | All rights reserved