StarNet Software

AI-powered software for star removal and noise reduction in astrophotography.

PixInsight Modules

This page describes the StarNet and DeepSNR PixInsight modules: which repository to add, current releases and system requirements, and where the shared installation instructions live.

PixInsight repository setup

Supported PixInsight versions: 1.9.4 and 1.9.5 only. Intel Mac modules remain on PixInsight 1.9.4.

Use this shared flow to add the StarNet or DeepSNR repository in PixInsight. Current releases use a single repository for each tool.

Do not add the TensorFlow/runtime repositories for current StarNet or DeepSNR installs. They are deprecated for the current PixInsight release path.

On macOS, PixInsight module architecture follows the installed PixInsight application, not hardware architecture. PixInsight x64 installs receive the x64 ONNX/ORT packages, even on Apple Silicon hardware, and will not get Apple GPU acceleration. For maximum StarNet2 and DeepSNR performance on Apple Silicon, install the PixInsight 1.9.4 ARM64 build so PixInsight receives the ARM64 CoreML packages.

Add repositories in PixInsight

If you need more detail on adding PixInsight repositories, follow the sequence below.

Open Resources → Updates → Manage Repositories.

PixInsight Resources Updates Manage Repositories menu path.
Step 1: navigate to the PixInsight repository manager.

In the opened window, click Add to add a new repository URL.

PixInsight repository manager window with Add button.
Step 2: add a new repository to PixInsight.

Enter the StarNet or DeepSNR repository URL. After the repository is added, apply updates, restart PixInsight, and verify the module appears and runs on a representative image.

StarNet: https://pixinsight.starnetastro.com/

DeepSNR: https://pixinsight.deepsnrastro.com/

PixInsight repository manager with StarNet repositories added.
Final repository list example from the original StarNet setup flow.

Windows GPU acceleration

Current Windows ONNX Runtime packages support DirectML acceleration with CPU fallback. Advanced users experimenting with manual NVIDIA CUDA setup can see the Windows GPU acceleration instructions.