​Resources»​Reveal Newsletter»​July 2019

Welcome to our periodic newsletter that aims to highlight recent changes and additions to Shearwater Reveal. In this issue we discuss new tools we have added and updates to existing tools.

For a more extensive list of updates to the software, view the Software Changelog found under the Help menu within Reveal.

User Group Meetings & Conferences

  • 2019 European User Group Meeting

    Save the Date for this year’s European Reveal User Group Meeting, central London on October 23rd.

    Further information to follow

  • 2019 EAGE Conference

    Reveal had a large presence at the 2019 EAGE Conference in London. 

    Along with giving one-on-one demos to interested parties, there were also 4 on-booth demos each day.

    Photo: Nate giving “Intro to Reveal” demo at 2019 EAGE


  • 2019 Houston User Group Meeting

    This year’s 2019 Houston UGM took place at the Westin Memorial City in April. The Reveal team demonstrated new tools and enhancements to the software to a group of users, followed by a cocktail hour and one-on-ones.

    Photo: Roy giving update at 2019 UGM

Tips, Tricks & Hints

MatplotLib has been added to the standard Reveal python third party. 

New Analysis Tools


CoherentNoisePredict provides a robust way to interactively model and QC coherent noise in the FK domain for regularly and irregularly sampled input gathers. The noise models generated by CoherentNoisePredict can be filtered from the original dataset with the SRMESubtract tool.

  • Figure 1. A synthetic shot gather with coherent noise.

  • Figure 2. The regularized input data in the FK domain with low-velocity and high-velocity picktables bounding the noise.

  • Figure 3. The noise model generated from the synthetic shot gather.

  • Figure 4. The filtered synthetic shot gather obtained with SRMESubtract.

TimeLapse Binning

TimeLapseBinning is a new tool that bins the best matching seismic traces from multiple repeated surveys into common-offset bins based on similarity criterion. Currently, our tool allows users to match traces based on the following list of similarity criteria:  source distance, receiver distance, source and receiver distance, midpoint distance, amplitude difference, azimuth difference, NRMS,  correlation, and predictability. In addition, users have the option to apply thresholds based on an additional similarity criterion. Currently, the following list of threshold criteria are available: amplitude difference, correlation, predictability, and NRMS.


A new tool that computes refraction tomography. The tool comes with many options, including the ability to store the L matrix if so desired. TomoStatics is available in the advanced land package. 

Figure 5. Comparison of TomoStatics velocity before (left) and final (right)


Resample is a new tool that allows users to resample to any floating point sample rate. This removes the required to down sample prior to resampling.


Horizon2Gather is a new tool that projects a horizon to the matching event on a gather. The projection is done by taking the cross-correlation between the current trace and a model stack in a sliding window along the gather. Many attributes are stored in headers e.g. the RMS of the samples in a window around the projected horizon. 

Figure 6. Horizon overlaid onto a set of flattened gathers. The horizon time value is plotted below each gather.

New Groundroll Workflow

We have developed a new workflow for the velocity analysis and suppression of groundroll noise.  This workflow is targeted specifically at noise that is highly dispersive and spatially heterogeneous. The workflow consists of three tools:

GRAutoPicker – performs dispersive velocity analysis and auto-picking of groundroll records

GRFilterCurves – applies constraints to raw picks from GRAutoPicker, high-grading the velocity analysis to output a gridded dispersion volume (x, y, velocity, frequency)

GR3DApply – uses the dispersion volume to compute a unique dispersion correction for each trace in a record (depending on the dispersion velocities along the path from source to receiver).  After dispersion correction, which separates the noise from signal, the tool filters the noise from the data.

The following displays illustrate the workflow on a dataset with synthetic dispersion noise.

GRAutopicker – Left panel is original shot record with two reflectors and strong groundroll.  2nd panel - Mute lines and windows are specified in the tool to isolate the part of the noise local to the source and consisting of one single mode of groundroll. Third panel is slowness vs frequency plot of semblance, with automatically picked line in black.  Note the dispersion (slowness increasing with increasing frequency). Fourth panel is the ground roll with dispersion removed to focus the groundroll to a very narrow time window.  The fourth panel is a QC of the auto-picking:  If the picked dispersion curve is correct, the groundroll mode should be flat.  Note that the primaries are dispersed to other parts of the record, minimally overlapping the dispersion-corrected groundroll.

GRFilterCurves – Data record and (right panel) auto-picked curves for analysis windows in the vicinity of the location of the record.  Note that some of the curves have high variability, especially at high frequency.  Constraints are applied to the curves to eliminate those that would not lead to a good model (when averaged).  Suitable constraints are applied – such as limiting the variance or the maximum offset, or by requiring monotinicity of curves.  The constraints can be QC’d by looking at the curves for several locations or (middle panel) the map view of the average of all the curves after constraints are applied.  Once proper constraints are determined, the average frequency-velocity volume is gridded for use in the GR3DApply tool.

GR3DApply uses the gridded dispersion volume to dispersion-correct the data, remove the noise by applying narrow windows, and reverse the dispersion correction to restore primaries to their original location in the record.  Note that this is a true 3D dispersion correction – the method accesses the specific velocities on the source-receiver path for each trace, integrates the slowness along that path, and applies the dispersion correction for that trace based on that integrated dispersion curve unique to that path.  Because the dispersion-corrected groundroll minimally overlaps the dispersed primary signal, noise can be removed with very small impact on primaries.  It is also possible to output the noise-only model from the tool, which can be adaptively subtracted from the data.

Technology licensed from ExxonMobil Upstream Research Company.

Additional Functionality Updates


The SurveyMatchCompute tool was updated to allow the user to output a survey-wide match filter. The output is a wavelet that can be applied to the non-reference survey with the ApplyFilter tool.

Likewise, the tool was updated to allow users to match 4D surveys by allowing the seismic traces to be matched along three headers.

Spectrum Mouse Mode Upgrades

The following features have been added to the Spectrum mode:

Additional display modes added to Spectrum palette:


Signal-to-noise ratio spectrum

Phase spectrum unwrapped

FX power spectrum

Addition to the “Edit Spectrum Plot Bounds” option in the 2D Viewer to easily specify the data target: a single trace, the entire ensemble, and/or the entire time range (in addition to the existing min/max edit boxes) 

Option to specify and lock spectrum plot axes

Additional display options:  % power and linear power (original Spectrum plot now called Raw Power)

Option to specify reference amplitude for power

Option to specify a time shift for calculation of phase

Figure 7: Spectrum, FK and FX view


DBInterp was given a major update, introducing new features and speed improvements. Two new interpolation options have been added to the tool on top of the already existing Delaunay triangulation: Gaussian Process and Cubic Spline interpolation.

The Gaussian process option uses Gaussian process regression and radial basis functions to create a global model for all the input points instead of interpolating data based on only nearby values. For this reason, it is able to provide high-quality results for very sparsely picked dbs in areas where the default interpolation does not have enough information. However, since the entire input must be held in-memory to compute a global solution, this also limits the size of the input dataset; if they are too big the memory will run out and the tool will fail.

Figure 8. Input db, Delaunnay interpolation, and Gaussian Process interpolation comparison.

Thus, the cubic spline interpolation was created as an alternative to the Delaunay and GP interpolation for situations where the input dataset is too large for Gaussian process to run but not regularly sampled enough for Delaunay triangulation to produce good results. This new interpolation works by first using a sparse linear interpolation to fill-in areas with poor pick coverage and then creates a piecewise cubic, curvature-minimizing polynomial surface for the final interpolation.

DTB Comparison

By dragging a .dtb file from the “Projects” tab onto the viewer for another .dtb file, new options are available that allow the new dataset to be loaded as a background, a window-pane comparison, an attribute layer with transparency, or loaded next to the other dataset. Dragging a .dtb file onto another now has a similar behavior to dragging a .seis file onto another.

Figure 9. Two .dtb datasets loaded side by side for comparison.

Welcome to the Team

  • Cameron Astill

    Cameron has over 35 years of industry experience, starting his career with GSI as a Seisomologist in Australia. In the late 1980s, Cameron had his first overseas posting as Country Manager in Egypt, followed by stints in the United Kingdom, Nigeria and Singapore – where he has been for the past 20 years. Prior to joining Shearwater, Cameron was Senior Vice President for Polarcus in APAC for 7 years and Executive Vice President of CGGVeritas Asia Pacific for over 10 years. Cameron holds a Bachelor of Science degree, completed the Advanced Management Program at Harvard Business School and served as a director of the IAGC, ASEG and ANSIR in Australia and Singapore. 

  • Aaron Lockwood

    Aaron has 14 years experience in seismic processing and geophysical operations management in addition to an academic career. Aaron holds a BSc (Hons) in Geology and Geophysics from Durham University and an MSc in Exploration Geophysics in addition to a PhD from Leeds University, UK. After completing a PhD Aaron worked as a near surface geophysical consultant before starting as a seismic processor at Veritas in Crawley, UK.

    Aaron then went on to work offshore for Fugro before moving back to a consulting role at Senergy, providing geophysical operations management out of the London office. Prior to joining Shearwater Aaron worked in Australia for DUG and as an independent; providing field support for the collection of gravity data onshore UK as well as seismic industry technical insight to financial fund managers. Aaron currently sits on the PESGB special interest group (Geophysics) committee.

    As an experienced seismic processor and project manager, Aaron is now part of the team as the Reveal Sales Manager for the EAME region and is based in the Tunbridge Wells office.

  • Enoch Huang

    Enoch was born and raised in Western Massachusetts and attended college in Upstate New York at the Rensselaer Polytechnic Institute. Enoch graduated in May of 2018 with a dual BS degree in computer science and game design. Prior to Shearwater, Enoch has worked as a software developer for a medical software development company called Velentium, as well as a software development intern for Siemens Power Technologies International. Enoch is now joining the Reveal team as a software engineer in the Houston office.

  • Warren Ho

    Warren has over 20 years experience in the geophysical oil and gas industry.  He holds a BSc in Geophysics from the University of Texas at Austin.  Warren began his career with PGS and later with ION Geophysical, where he was at for 11 years.  During this time, he held various roles in seismic time processing and depth imaging.  Warren later joined Dolphin Geophysical in 2014 as a Team Lead in processing and imaging.  Warren is now joining the Reveal software group as a Geophysical Testing Manager.

  • Aron Azaria

    As a Geophysicist, Aron has processed seismic data for 12 plus years working for Ion Geophysical, CGG and Geophysical Development Corp. As a user of many different software packages and tools over the years, Aron understands the importance of our clients desire to produce a quality product in a timely fashion. That has always been his goal and passion as a geophysicist. Moving forward Aron will apply that same passion to help make Reveal processing software reliable, robust, and of superior quality.

  • Rob Light

    Rob graduated with a 1st Class Honours Masters of Earth Science Geophysics degree from the University of Liverpool, England in 2007. Starting out as an offshore processor for CGGVeritas before some time spent processing with TGS. He joined Dolphin Geophysical in 2013 as an offshore processor for 6 months while developing skills with Reveal. In Jan 2014 he moved onshore as a Processing Supervisor for offshore processing and later as a Team Leader for processing. In 2018 he joined the Reveal team as Software Support Geophysicist in the EAME region. Rob has a wealth of experience in Reveal and Marine Processing and brings this expertise to Reveals clients through his role in Support.