Skip to main content
Version: 3.5 (unreleased)

Figures and Images

A screenshot of the VRGS window is rarely a good figure. It is the wrong resolution for print, it has a perspective distortion that makes anything you measure off it wrong, and it usually carries interface furniture that a journal will ask you to remove.

VRGS can do considerably better than a screenshot, and the tools are not obvious. This guide covers producing a still figure; for moving pictures see Flight Paths.

Capture above screen resolution

Home ribbon → Copy Screen is the important command, and its name undersells it. It does not capture your monitor — it re-renders the scene to a virtual screen at whatever size you ask for, so the output is not limited by your display.

The dialog offers:

ControlWhat it does
Size presetA dropdown of common output sizes.
Width / HeightThe output size in pixels, editable directly. Defaults to 1024 × 800; the range runs to 32768 in each direction.
×2 and ÷2Double or halve both dimensions in one click — the quick way to get from a preview to a print-resolution version of the same framing.
Save to fileWrite a PNG instead of putting the image on the clipboard.
Work at preview size, export at print size

Frame the view, capture at the default size to check it, then press ×2 twice and capture again. The framing is identical; only the pixel count changes. A figure destined for a journal column at 300 dpi needs roughly 2500 pixels across, which is two doublings from the default.

The saved file is always PNG, regardless of the extension you type. This is deliberate — an earlier version offered .bmp and .jpg filters that wrote PNG data under a misleading name.

Get the projection right

A perspective view is the wrong basis for any figure someone might measure off. Home ribbon → Ortho Camera switches the 3D view to orthographic projection, where distance from the camera no longer changes apparent size.

Two things follow from being in orthographic mode:

  • A scale bar becomes meaningful across the whole image rather than only at the depth it was drawn.
  • The Orthographic Grid Overlay becomes available (Display Properties → Post-Processing Effects). It draws an adaptive measurement grid whose spacing adjusts with zoom, with Detail Levels controlling how many nested scales are drawn at once and Opacity and Line Width controlling how much it intrudes. It has no effect in perspective view.

For an outcrop panel — the classic figure in this field — orthographic projection with a grid overlay is what turns a picture into a document someone can work from.

Add the furniture

Home ribbon → Overlays carries the annotation layers that make a figure self-explanatory:

OverlayUse
Scale BarA dynamic bar at bottom-left that rescales as you zoom.
CompassOrientation widget. Essential on any oblique view.
Colour BarThe colour key for the active attribute — needed whenever the figure is coloured by anything other than texture.
Strat ColumnThe stratigraphic column, for a figure that references units. See Stratigraphy.
NavigatorThe on-screen navigation buttons. Switch this off before capturing — it is interface, not content.

For scale inside the scene rather than as an overlay, Display → Add Virtual Scale Object places a synthetic object of known size — 5 m, 1 m, 10 cm, or a triaxial marker showing all three directions. This is the virtual equivalent of putting a hammer in the shot, and it works in an oblique view where a screen-space scale bar would be ambiguous. See Measuring.

Choose shading that shows the structure

The default lit-and-textured view is often the worst way to show geology in a figure, because the texture's own contrast swamps the geometry. The Post-Processing Effects section of the Display Properties has better options:

EffectWhat it reveals
Eye-Dome LightingDarkens depth edges so structure reads without shading. The standard choice for un-coloured point clouds.
Curvature ShadingConcave creases darken and convex ridges brighten, making folds and fractures stand out. Overrides eye-dome lighting while on.
Show Slope / Aspect (Dip)Colours surfaces by orientation — hue is dip-direction, saturation is dip angle. Works even on tiled models with no per-vertex attributes. Overrides texture and attribute colour.
Enable Shadows / Ambient OcclusionRay-traced depth cues. Expensive, and they need the Vulkan view on a ray-tracing-capable GPU — but for a single still capture the cost does not matter.

Vertical Scale in the Geometry & LOD section exaggerates relief. It is legitimate in a figure provided you say so in the caption — and it is one of the easiest ways to make subtle topography readable.

Exaggeration changes what you see, not what you measure

Vertical Scale does not affect measurements or exported coordinates. That is convenient, but it means a figure at 2× vertical exaggeration will not agree with numbers taken off the same view. Always state the factor.

Other things worth capturing

OutputHow
A photograph or orthoimageRight-click the item → Export Image.
A stereonet or rose diagramRight-click → Export → Stereonet Diagram / Rose Diagram, which writes SVG — a vector file that scales without loss and can be edited in Illustrator or Inkscape. See Stereonet.
An attribute histogramRight-click the attribute → Export Histogram.
A colour scale on its ownCopy Color Scale to Clipboard on the attribute menu.
A whole assembled reportThe Report View builds figures, tables and text into one document.
A link back to this exact viewCopy as HTML copies the current viewpoint as an HTML anchor you can paste into a report or wiki, so a reader can jump straight to the view you are describing.

Video

Recording motion belongs to flight paths rather than to a capture command. Build a flight path, then record it — the recording runs against the project frame rate rather than whatever the display manages, so a 60-second path is 60 seconds on every machine and narration stays in step.

Capture Movie From Screen is the OpenGL-only legacy route

The old Display → Animations → Capture Movie From Screen command is handled by the OpenGL view only. Flight-path recording works in both renderers and gives a far better result. See Why is this greyed out?.

A workflow that works

  1. Hide everything that is not part of the figure — this is also the fastest way to make a heavy scene responsive while you frame it.
  2. Switch to Ortho Camera if anything will be measured off the result.
  3. Pick the shading: texture for context, curvature or slope/aspect for structure, eye-dome for point clouds.
  4. Switch on Scale Bar and Compass; switch Navigator off.
  5. Frame the view. Save it as a scene so you can come back to exactly this.
  6. Copy Screen, check at default size, then ×2 to print resolution and save to PNG.
  7. Note the vertical exaggeration and the projection in your caption.

Tips and troubleshooting

  • The capture is blank or partly black. The scene was still streaming. Let it settle, or load the model into memory first — see Working with Large Models.
  • The capture is slow at large sizes. Expected: a 8192-pixel capture renders the whole scene again at that size. It is not hung.
  • Text and labels are tiny in the big version. Labels with Maintain Size on are sized in pixels, so they do not grow with the capture. Increase the label size before capturing large, or switch Maintain Size off so the size is in metres.
  • Colours differ from the screen. Post-processing effects are applied to the capture too — that is usually what you want, but check bloom is not blowing out highlights.
  • The image has the interface in it. You captured with a screenshot tool rather than Copy Screen. Copy Screen renders the scene only.

See also