Geopsy: H/V and Spectrum Toolboxes: Time Tab: General sub-tab

Time windows section in the Time tab of he H/V and Spectrum toolboxes

The general sub-tab is used to define four criteria of the window selection process:

• Window length
• Window overlap
• Bad sample definition
• Anti-triggering definition

Window length is the only criterion that is not optional and is mandatory to be set-up.

Window length

Window length pop-up menu

There are three ways to define window length (duration, in seconds) using the Length pop-up menu

• Exactly: all selected windows will have the same exact duration.

Choose Exactly in the Length pop-up menu and enter the window length (in seconds) in the spin box to the right of the pop-up menu.

• At least: selected windows will have a duration comprised between two values.

Choose At least in the Length pop-up menu and enter the window lengths (in seconds) in the two spin boxes to the right of the pop-up menu.

All windows which duration is comprised between these two values will be selected, thus possibly resulting in windows with different sizes.

This may be useful so as to keep the most possible windows of a very perturbed signal.

• Frequency dependant

Window Overlap

It is possible for successive windows to be contiguous or to overlap, using the Overlap box and the associated spin box.

Resulting windowing when Overlap box is not checked

• When the Overlap box is not checked, contiguous selected signal windows will not overlap.

• When the Overlap box is checked, contiguous selected signal windows will overlap by a chosen percentage entered in the neighboring spin box.

Resulting windowing when Overlap box is checked

Bad Sample Definition

The "bad sample" concept has been introduced for two purposes:

• eliminate clipped parts of recordings (which imply spectrum distorsion)
• introduce some "tolerance" on the sometimes too severe window selection criteria

A "bad sample" is therefore:

• either a sample with too high an amplitude (above a given threshold)
• or a sample which falls outside the specified STA/LTA range.

Bad Sample Use

Bad sample parameters

The use of the bad sample concept implies the definition of two parameters:

• Bad Sample Threshold
• Bad Sample Tolerance

Bad Sample Tolerance

A maximum number of bad samples may be accepted (tolerated) inside windows during the selection process. Check the Bad sample tolerance button, and enter the maximum cumulative time (in seconds) reached by all the bad sample that is accepted in each selected window.

Warning : It is recommended to use this option with care, and only when the plain application of default values lead to a very small number of automatically selected windows.

Bad Sample Threshold

This box allows to define the amplitude level above which a sample is considered as bad, and should be rejected when looking for usable windows.

To do so, check the Bad sample threshold box and enter the desired threshold in the spin box.

The threshold can be defined in two different ways:

Relative bad sample threshold

• the percentage of the maximum amplitude observed over the entire signal. (Thus, setting a 99% threshold will automatically eliminate all clipped parts of the recording)

Absolute bad sample threshold

• a sample above this value is considered as a bad sample.

Anti-triggering

The goal of the anti-triggering is to keep the most stationary parts of ambient vibrations, and to avoid transients. User's defined "bad samples" - if any - are also considered as to be avoided.

Two types of anti-triggers may be applied to he signal, both based on STA/LTA detection.

Anti-triggering on raw signal

When the Anti-triggering on raw signal box is checked, the anti-trigger is applied sample per sample all along the entire signal, using the STA/LTA parameters entered in the Raw signal sub-tab.

Anti-triggering on filtered signal

When the Anti-triggering on filtered signal box is checked, the signal is first filtered using the parameters entered in the Filter sub-tab, and then the anti-trigger is applied sample per sample all along the entire signal, using the STA/LTA parameters entered in the Filtered signal sub-tab.

IMPORTANT NOTICE: when using the anti-triggering on the filtered signal, the signal is filtered ONLY for the window selection process. Later processing (Spectrun, H/V, etc...) is performed on the original RAW signal.

Anti-triggering use

All box checking combinations are possible:

• one of the boxes checked: only the checked criteria is applied to the select windows
• both boxes checked: only samples that fulfill the criteria of both anti-triggers are considered as part of a "quasi-stationary" signal
• both boxes unchecked: no anti-trigger is applied in the window selection. It is thus possible to select windows based only on the Bad Sample criteria, or even no criteria at all.