Use this dialog to analyze the sample rates in any existing runs that were acquired with square wave inputs.
Go
to start analysis.This box shows analysis progress and then the old and newly analyzed results for the selected streams.
Standard errors on measurements spanning 20 minutes or longer should be less than ~0.01 Hz. If larger, there is some instability to fix before trusting the measurement.
If you are measuring a given clock for the first time, the difference from the manufacturer's uncalibrated rate should be less than ~2 Hz.
If you are remeasuring a calibrated clock, the new and old rate should be within ~0.1 Hz of each other. Otherwise, one of the calibrations may have failed.
If you don't trust the results, do not click 'Apply.' Rather, try to find the problem, and then repeat the measurement. Good places to look for issues are the reported error and warning messages in the Log
window. Also, try running with the Metrics
window (Ctrl+M) open to look for warning signs. Note that imec error flags, which signify possibly dropped samples, are visible in the Metrics window and are always recorded in probe metadata files.
Note too, that whether doing an online calibration run, or calculating rates from an existing run, 20 minutes should be considered an absolute minimum time span for calibration, but 40 minutes or longer is preferred.
All results producing (measured rate +/- std error) are applied.
Select whether to use the results to:
or both.
fin