IST Vector Postprocessor turns finite-element solver output into publication figures: vector PDFs that stay sharp at any zoom, H.264 animations that reproduce the on-screen motion frame for frame, and project files that restore a scene exactly. Screen, PDF, video and project all come from one scene object, so the figure in the paper is the figure seen on screen.

- Vector-true PDF. Surfaces are written as Gouraud-shaded triangles and lines as stroked paths, with no raster layer. Labels are set in Latin Modern with the font embedded, so figures match the LaTeX document around them.
- Recovery first. Six filters, from Zienkiewicz–Zhu SPR to edge-preserving Perona–Malik diffusion, move element data to the nodes and smooth out discretisation noise. The colour bar, contours, iso-surfaces, threshold and deformation all read the recovered field, and paired nodal and element fields give an a-posteriori error indicator.
- Scalars, vectors and tensors. Labelled iso-lines, iso-surfaces, streamlines and pathlines, principal-axis tensor glyphs, section planes, symmetry expansion, a black-and-white print mode, and 60 Hz transient playback whose interpolated frames stay within the values the solver produced.
- Reproducible. A project file restores the complete scene, a figure recipe applies the same camera and settings to another run, and rendering is deterministic across runs and thread counts.
- Reads what solvers write. EnSight Gold and EnSight 6, VTK legacy, XML and multiblock files, ParaView .pvd series, STL and OBJ, up to 50 million nodes. A bundled Python module with no dependencies writes your own mesh and fields as EnSight Gold.
- Its own engine. The readers, filters, renderer and PDF writer are the program's own C# code: pure Avalonia, no VTK library, compiled with .NET Native AOT into a self-contained executable that needs no runtime install.

Free for Windows 10 and 11 (x64). It needs no account or licence key and never contacts a server.