Skip to content

TODO.md

Latest commit

 

History

History
94 lines (77 loc) · 3.24 KB

TODO.md

File metadata and controls

94 lines (77 loc) · 3.24 KB

Some TODOs

This is not a comprehensive list, nor a plan for the research or development of future versions of wave_tracer.
These TODO lists are a wish list or ideas that came up. The lists are unordered, and points vary widely in scope, importance, and complexity.

Features

  • beams: rework beam extents, sourcing phase-space extents, and gracefully dealing with degenerate idealistic sensors (like point perspective).
  • beams: exact beam footprint on interaction
    • is this needed?
  • emitters: antennas
  • surface bsdfs
    • Fresnel transmission for arbitrary (complex) IOR
    • all surface BSDFs should admit some transmission, IOR dictates attenuation through the medium
    • replace dielectric with smooth bsdf, handling arbitrary IOR
      • smooth bsdf: add polarizing capability
      • smooth bsdf: add birefringence
  • quantities: read power related quantities in scene files (e.g., spectrum scales)
  • sensors: sensors associated with geometry (outer lens, etc.); forward connectable
  • sensors/films: 1d/3d sensors & non-imaging preview/output
  • area emitters: detector beam integration over emitter
  • sensors: additional camera models
    • thin lens
    • physical pinhole
    • lens systems?
  • bsdfs: diffraction gratings
  • bvh8w: create better trees (higher occupancy)

Fixes

  • double precision: broken tracing, why?
  • double precision: make intersection tests and traversal respect double-precision

Research

  • diffractions
    • polarimetric
    • edge geometry, curved edges, electric properties
    • corners
    • better sampling
    • cross-interaction diffractions
  • intersections: better clip distances and numeric tolerances
  • polarimetric BSDFs
    • surface bsdfs for long-wavelength scattering (Degli & Esposti 2007 as a starting point)
    • better diffuse model?
  • rough surface profiles:
    • is sampling correct?
    • sampling only of relevant surface frequencies.
  • acoustics
    • integration without temporal averaging
  • beams: how to handle focusing optics well?
  • medium interactions
  • fluorescence and cross-spectral effects
  • lens systems
  • GPU
    • CUDA?
  • "true" Gaussian beams: curvature, Gouy phase, anisotropic solid angle
    • is this needed?
  • polarimetric sampling of interactions
    • is this needed?

Engineering and C++ Modernization

  • XML schema for scene file
    • loader/deserializer
    • serializer
    • really missing docs (priority)
  • loader: cleanup aux_tasks mess; handle dependencies well
  • bitmap write2d/load2d: handle colourspaces, linear/sRGB, etc. Currently, support is a mess (especially openexr)..
  • distributed rendering
  • profiling: add memory tracking mechanism for scene elements
  • profiling: add loading time tracking for scene elements
  • units: complex units
  • units: matrices with units
  • units: resolve the current mix of dimensionless SI angles and strong angular system (u::ang::*)
  • c++ modules

Misc

  • tests
  • gui
    • simple debug views
    • allow configuring runtime defines
    • interactive scene editing
    • path tracing debug view, with visualizations of paths, beams and interactions
  • extend docs
  • python bindings?