Why don't you link to the original preprint [0]? phys.org and its paywalled brethren are all too often touted on HN to waste time on sci-pop.

The paper has lots of pictures.

[0] https://arxiv.org/pdf/2009.08987.pdf