Per https://eoranalysis.slack.com/archives/C1GS5FQ2J/p1647389854273629 in slack, digital noise tests (crossing odd and even samples) with different seeds integrates down for a minute or two, but reaches a floor that is likely to have to do with PFB window overlaps between odd and even samples. To mitigate this, we should look into slowing down the odd/even jacknifing, either in F-Engine packet transmission, or by accumulating into buffers at a lower cadence in the X-engines.
Per https://eoranalysis.slack.com/archives/C1GS5FQ2J/p1647389854273629 in slack, digital noise tests (crossing odd and even samples) with different seeds integrates down for a minute or two, but reaches a floor that is likely to have to do with PFB window overlaps between odd and even samples. To mitigate this, we should look into slowing down the odd/even jacknifing, either in F-Engine packet transmission, or by accumulating into buffers at a lower cadence in the X-engines.