Adjust timestamp for lost time in filtering (get_heartrate)

[itismeghasyam]

In get_heartrate, we lose a few seconds(this again depends on sampling_rate) in the signal while filtering it. This is not adjusted in the timestamp. i.e what we assume as HR at 0s after filtering is actually at 3s before filtering.

This is not a huge problem while measuring resting HR, but while measuring the drop off of HR post exercise it might be an issue.

This is not high priority, but better if fixed.