Scientists have developed a way to accurately estimate covariance matrices in cosmology without needing to rely on mock catalogues. By using a specific type of model covariance matrix, they can achieve high precision with minimal computational resources. Instead of using numerous mock catalogues, they found that the parameters for these models can be determined accurately by analyzing a single survey volume. This means that covariance matrices can now be calibrated using real data alone, without the need for reference to mock catalogues.