Skip to Content
DocsGetting startedGetting started

Getting started

Welcome to the SVETlANNa documentation. This section gets you from an empty folder to a working optical simulation.

What is SVETlANNa?

SVETlANNa is a Python library for simulating light propagation through optical systems and for building diffractive optical neural networks (DONNs). It is built on PyTorch, which gives you:

  • Automatic differentiation for optimising optical systems
  • GPU acceleration via CUDA
  • Native PyTorch integration (nn.Module, DataLoader, autograd)

SVETlANNa provides four propagation methods: ASM (angular spectrum) and RSC (Rayleigh–Sommerfeld convolution), each with a zero-padded variant (zpASM, zpRSC) that suppresses wrap-around artefacts.

In this section

Virtual environmentInstallationQuickstartCore concepts

A minimal example

from svetlanna import SimulationParameters, Wavefront
from svetlanna.units import ureg
 
# define the simulation grid
params = SimulationParameters.from_ranges(
    x_range=(-1*ureg.mm, 1*ureg.mm), x_points=256,
    y_range=(-1*ureg.mm, 1*ureg.mm), y_points=256,
    wavelength=632.8*ureg.nm
)
 
# create a Gaussian beam
wf = Wavefront.gaussian_beam(params, waist_radius=0.3*ureg.mm)
 
# inspect the result
print(f"Field shape: {wf.shape}")
print(f"Max intensity: {wf.max_intensity:.3f}")

Next steps

Start with Virtual environment to set up an isolated Python environment with PyTorch and SVETlANNa, then move on to the Quickstart, where you build a complete optical system in five minutes.

SVETlANNaVirtual environment