Stack imaging
Stack imaging software
One interactive application for everything between conditioned gathers and a finished stack: stacking velocity analysis, mute picking, NMO and CMP stacking, and a processing sub-sequence that runs inside the module — on a 2D line, a set of sail lines or a full 3D cube.
One module
Pick, condition, stack, compare
4 insertion points
Processing sub-sequence inside
Partial restack
Rebuilt only where picks changed
In view
Stack imaging session
The stack is where velocity work becomes a picture. Stack imaging takes pre-stack gathers in any sorting, without NMO applied, and gives you the semblance, the mini-stacks, the gather before and after correction, the velocity inline, the stack inline and the location map as views of one live session — so a pick made in one window is visible in the others straight away. It writes 3D stack cubes for the 3D viewer, a fast stack for a quick look, velocity picks to an ASCII file or to the project database, and it will read a legacy velocity field and build a section from it. Migration velocity analysis is deliberately somewhere else: that work belongs to the migration velocity selector and the imaging selectors on the time-imaging side.
Inside the module
Everything the stack needs, in one window set.
Picking, conditioning and stacking are usually three flows in three tools. Here they are parameters, sub-sequences and action items of a single module — which is what makes iterating on a velocity field quick enough to actually do.
Semblance and mini-stacks
Pick on the velocity spectrum, or on constant-velocity mini-stacks built at a chosen trace radius and velocity increment. A corridor brackets the trend, the current pick line sits between its limits and the interval velocity is drawn as you go.
Before and after NMO, together
The current seismogram and the NMO-corrected seismogram sit side by side with the travel-time curves overlaid, so an over- or under-corrected event is visible at the moment the pick moves. The stretch factor you set is the one the stack uses.
Mute picked in the same session
Pick the mute on the NMO gather with a tapered edge; optionally let it act on the semblance too, so the spectrum reflects what will really be stacked. Symmetric-from-positive and symmetric-from-negative application handle split-spread data.
A sub-sequence at four points
Insert AGC, deconvolution, filtering, denoise or spectral balance for velocity analysis, before NMO, after NMO or after stack. The data driving the semblance can be conditioned differently from the data being stacked — no second workflow needed.
Pick anywhere on the map
Work through a picking grid at a defined inline and crossline increment, or switch to all bins and pick at any location you click on the location map — useful where the section changes character between grid nodes.
Auto-picking inside a corridor
Pick a corridor skeleton at a few locations and let it interpolate between them, then run the auto-picker with a semblance threshold below which nothing is picked. Semi-automatic mode snaps your own picks to the local semblance maximum.
Sorting and offset masks
Sort by offset, absolute offset, offset-vector tile or as-is, and mask minimum or maximum offsets by bin grid so the near or far traces you do not want never reach the stack.
Velocities live at topography
Picking happens on the topography, not on an assumed flat datum, and source and receiver elevation statics are applied to the CMP elevation automatically during NMO. Shift the result to a final datum inside the module when you want it there.
Variants
The same session, shaped to the survey.
A single 2D line, a set of sail lines, a horizon-guided update or a first-look volume are different problems. Each has its own module, and they share the picking interface.
Interactive
Stack imaging
The main application: velocity and mute picking, CMP stacking, the internal sub-sequence, 3D stack cubes viewed in the 3D window, and a fast stack for QC without leaving the module.
Many lines
Stack imaging multi
For several 2D lines or marine sail lines at once, with two velocity-interpolation modes — from one line outward, or from every pick on every line. The difference between them shows up at the line intersections, which is exactly where mis-ties live. Progressive stack then merges the individually stacked sail lines, gaps and all, into a single volume.
Horizon-guided
HVA stack imaging
Picks velocity along a picked or imported horizon on a reference stack, then rebuilds the stack from the updated field. It can read a precomputed velocity analysis instead of recomputing semblance.
Quick look
Fast stack
Accumulates each CMP bin as the data streams past and writes straight to a .gsd volume, with NMO applied on the fly if you want it. No sorting required and no need to hold the survey in memory.
Optimization & output
Built for the second pass, not just the first.
Velocity work is iterative, and the cost of an iteration is what decides how many you do. Two features carry most of that weight: a stack that is rebuilt only where the picks moved, and a semblance that is computed once and stored.
Making iterations cheap
Stack optimization and the storage that feeds it
Partial restack
Recreate stack in places where picking changed: the edited areas are rebuilt and the rest of the stack is left exactly as it was.
Precomputed VA
Compute the velocity analysis for the whole volume once and store it on disk, so later picking sessions — and horizon-guided runs — read it instead of recalculating.
Fast stack for QC
A bin-by-bin stack straight to disk when you want to see the section rather than perfect it.
Selective displays
Create inline only, crossline only or both, and keep data gaps visible on the section instead of silently interpolated.
Interpolation control
Kriging between picks with a chosen model, range and point count — beyond the range, picks are treated as uncorrelated rather than smeared together.
Residual passes
Re-pick residually for the current bin or for the whole dataset once a first field exists, instead of starting the picking over.
What you get out
Products, formats and where they go next
Stacks
Stack, Vrms, NMO gathers and fold, written to SEG-Y or to the internal GSD format, individually or all at once.
Cubes
A 3D stack cube and a Vrms cube for the 3D view — and the pair HVA stack imaging needs as its reference on 3D data.
Velocity picks
Saved as an ASCII picking file or into the project database, so a field can be reloaded, compared or handed to another module.
Mute functions
Saved and reloaded like velocities, or taken from an external picking file or another module's picking item.
Legacy velocities
Import an existing velocity field — ASCII picks, a Vrms item or a velocity gather — and build a stack from it before picking anything new.
Downstream
The stack feeds post-stack time migration and MultiFocusing imaging; the picked field is the starting point for migration velocity analysis and for depth model building.
Modules in this group
The stacking family in the g-Platform module tree
Stack imaging
Stack imaging multi
HVA stack imaging
Precompute VA
Fast stack
Progressive stack
Stack
NMO
Create velocity model
Velocity editor
Workflow
Six steps from gathers to a saved stack.
1
Connect the gathers
Pre-stack data in any sorting and without NMO. On a large survey, point the module at traces on disk instead of loading them, and give the precompute storage a name.
2
Set up the analysis
Velocity range and step, semblance smoothing and resolution, stretch factor, super-gather radius, and mini-stacks if you want to pick on constant-velocity panels as well as on the spectrum.
3
Condition inside the module
Drop the processing you need into the sub-sequence — band-pass and AGC for a cleaner spectrum, statics before NMO, denoise after it, spectral balance after the stack.
4
Pick velocity and mute
Work the picking grid by hand, or pick a corridor skeleton and let the auto-picker fill it in, then pick the mute on the NMO gather. Bad picks come out with a polygon.
5
Stack and inspect
Execute the module to build the stack inline and crossline, compare it against the previous version, and revise. Only the areas whose picks changed are rebuilt.
6
Save and hand on
Write the stack, the Vrms field, the NMO gathers and the fold to SEG-Y or GSD, shift to a final datum if the next step needs it, and keep the picks in the database for the next iteration.
More g-Platform capabilities
Stacking sits between velocity and migration.
The same project, the same batch processing and the same cluster carry a survey from field tape through to depth. These are the other topics in the g-Platform workflow.
FAQ
Questions, answered.
What is inside the stack imaging module?
Stacking velocity analysis, mute picking, NMO and CMP stacking, and a processing sub-sequence, in one interactive application. Semblance panels, constant-velocity mini-stacks, the gather before and after NMO, the velocity and stack inline, the location map and the fold are all views of the same session, so a pick changes the displays around it immediately. Migration velocity analysis is deliberately not part of it — that belongs to the migration velocity selector and the PSTM imaging selector.
Do I need a separate flow to condition the data before stacking?
No. The module carries its own sub-sequence with four insertion points: for velocity analysis, before NMO, after NMO and after stack. AGC, deconvolution, filtering, denoise or spectral balance are inserted at the point where they belong, so the data feeding the semblance can be conditioned differently from the data that is stacked, without building a second workflow around the module.
What happens to a big stack when I revise a few velocity picks?
Only the edited part is rebuilt. Stack optimization has a Recreate stack in places where picking changed option: the stack is regenerated where the velocity picks were changed and left alone everywhere else. On large surveys the companion measure is a precomputed velocity analysis — the semblance for the whole volume is computed once and stored on disk, so later picking sessions read from storage instead of recomputing.
Does it handle multiple 2D lines and marine sail lines?
Yes, through two companions. Stack imaging multi picks velocities across several 2D lines or sail lines at once and offers two interpolation modes — interpolate from a single line, or from every pick point on every line — with the difference showing at the intersections. Progressive stack then combines individually stacked sail lines, gaps included, into one seismic volume.
Get started
Pick a line, stack it, compare it.
From raw gathers through an interactive velocity field to a saved stack and its Vrms volume — take g-Platform for a trial run, or talk to Geomage about your survey.