The reasons behind a bad BVP signal as the one shown below could be related to a broken optical sensor (in which case a replacement is needed) or to physical obstruction of the light that cannot illuminate the skin and measure the changes in light absorption. In order to avoid the latter situation make sure that both the red and green lights are visible under the E4, before or while wearing the wristband. Several cases have been reported because users were erroneously wearing the wristband with the charger on.
Visualization of a bad BVP signal on the Connect: a three-minute-long session example.
Example of the signal at the beginning of the session (the first 30 seconds).
A big peak is visible at the beginning of the session spanning the range of [-30, +30] and then no peaks are visible: the values stabilize around -3.5 with a much shorter range [-3, -4]) and the signal is off the scale.
Example of the signal after the peak (30 seconds time range): the signal is out of the scale range and not physiological.
As a result of a bad BVP signal, IBI values are not computed so the visualization on the Connect is as the one below. The raw IBI signal downloaded has a dimension of 0 Mb. Alternatively, just a few samples (not equally spaced in time, but at several seconds of distance between each other) are shown, with a curved line interpolating between the values.
Example of a downloaded IBI.csv file that could be totally empty or just contain a few samples that are not equally spaced in time, as the ones in the table below.
The HR.csv computation is not reliable either, even though it will contain values: E4 data - HR.csv explanation.