ViSL3D (pronounced "visled") is a Python package that creates X3D and HTML files to visualize datacubes from astrophysics in 3D in an interactive way. We use X3D and x3dom to represent figures in 3D and to integrate them into an HTML. The models have been made taking the x3d-pathway (Vogt et al. 2016) as a starting point. The current code was made for radio data, although it can be used with other types of 3D data.
This is an example of a 3D visualisation produced with ViSL3D. It shows the HI emission of the Hickson Compact Group 16 along an optical image from DSS2 and the positions of the galaxies from NED.
pip install visl3d
Examples of how to use the package are provided in the Jupyter Notebooks inside this folder.
The produced HTML file can be visualised in any standard browser or in notebooks themselves.
If you use the function createVis, which creates integrates the X3D model in the HTML, there are NO prerequisites. You can directly open the produced HTML file in a browser or run in a Jupyter Notebook.
If you create the X3D and the HTML files separately, using createX3D and createHTML, follow these instructions:
-
Install a local HTTP Server. Visualisations produced by ViSL3D must be opened through a local server in order to be displayed. There are a few options:
- Apache is a popular HTTP server that can be installed in most operating systems.
- Python has a built-in HTTP server that can be used by running
python -m http.serverin the directory where the HTML and X3D files are located. - VS Code has a built-in HTTP server that can be used by installing the Live Preview extension.
-
Move the created X3D and HTML files into the Apache (or other) DocumentRoot directory. By default, in Apache, it is
/var/www/htmlin Linux,C:\Apache24\htdocsin Windows, and/usr/local/var/www(or similar) in Mac. The DocumentRoot can be found and modified in thehttpd.configfile. Finally, open the visualisation in a browser (most common browsers are supported) by typing the URLlocalhost\example_file.html.
- Plot any number of contour surfaces,
- Plot galaxies with labels,
- Add a 2D image in any wavelength in background,
- Change the scale of the velocity axis,
- Shift 2D image along velocity axis and display the value,
- Rotate, zoom and pan complete figure,
- Hide/Show different components of the figure,
- Change ax labels,
- Change viewpoints,
- Change the colormap,
- Add markers,
- Represent multiple spectral lines in the same visualisation,
- Overlay spectral lines.
This package is in an early stage of development. It is not intended for production use and may contain bugs or incomplete features. These are some issues that you may encounter:
-
Long loading time. If the 3D model is large the visualisation might take a long time to render or not render at all. In this case, try to make a cutout of the cube (can be done directly in
ViSL3D), decrease the resolution of the 3D model (also inViSL3D) or decrease the number of iso-surfaces or increase the S/N of the lowest surface. -
When visualising in Jupyter Notebooks, having two visualisations in the same notebook produces errors.
-
The animation stops after completing at least one loop. Do not click the button many times if you don't see it stopping.
-
The opacity feature does not work properly along the 2D image.
-
The calculation of the RMS, which can be chosen as the unit to represent the cube, can be inaccurate. It uses negative voxels to create half a Gaussian and calculates the RMS from that. Therefore, if there is absorption or negative artifacts this calculation will be wrong. Another option is to calculate the RMS yourself and give it in the isosurfaces parameter, e.g.,
[3*rms, 5*rms...].
Please report any other issue in the GitHub issues.
If you use this package in your work, please cite Labadie-García et al. (2025).
Vogt, Owen, Verdes-Montenegro & Borthakur, Advanced Data Visualization in Astrophysics: the X3D Pathway, ApJ 818, 115 (2016) (arxiv; ADS)