Covariance in Apriori in BatchLSEstimator vs. SequentialBatchLSEstimator

Hi all!
I’m currently generating synthetic ‘measurements’ in Orekit and then using them to perform OD. I would like to use the BatchLSEstimator for the OD but it seems like there isn’t an option for adding covariance in the apriori covariance or the measurements. To generate a measurement, it appears to take a ‘noise’ matrix, the sigma bound of the noise/uncertainty, and the weight. Are all of those used when the measurements are added to the estimator or are they only used to generate noise in the measurement? Or is it just the weight that is used. Is there something similar for the apriori state?
I would prefer to use the Batch estimator but is it possible to add these covariance matrices in the Sequential version (SequentialBatchLSEstimator)?

Edit: As an addition, if you add a propagatorBuilder (in my case numerical propagator) which is initialized using an orbit, how does the estimation process use that differently than giving the SequentialBatchLSE an apriori?