Aux channels Auto Range not working properly

[ChucklesOnGitHub]

Describe the bug
Applying Auto range shows waveforms smaller than optimal. There seems to be a scaling mismatch. In the attached example, a Range of 2000mV set manually was good to display the entire signal. The Auto setting displayed a signal of significant less amplitude.

This might be related to the fact that the values on the Range drop-down do not match the values of the cursor property units (obtained by clicking on the lower bound of the graph):
When I used 2000 mV in the dropdown, the cursor reported -1 on the y axis lower bound, instead of -1000 mV as the label indicates.
For the Auto range, the value of the cursor on the lower bound was -100.

To Reproduce
Acquisition Board > LFP Viewer. Switch between ranges and inspect the value under the cursor and what would be appropriate as an Auto range.

Screenshots
Video here.

Versions:

• OS: Windows 10
• GUI version: 1.0.2
• Plugin version: LFP Viewer 1.0.2

[anjaldoshi]

Hi @ChucklesOnGitHub,

The Auto setting uses the range defined by the source processor that owns the AUX channel. In this case, the Acquisition Board’s AUX channels are currently set to a maximum range of 100 m/s², which is probably too large for this signal. That’s why the waveform amplitude looks much smaller than expected when using Auto range.

So I think this is something that should be fixed in the Acquisition Board plugin. A maximum range of 10 m/s² would likely make a lot more sense here.

open-ephys_acq-board_aux_scaling.mp4

As for the mismatch between the mV range dropdown and the cursor readout, that part is expected behavior: the display is converting the dropdown value from millivolts to volts for the cursor readout in AUX channels. So if the selected range is 2000 mV total (that is, -1000 to 1000 mV), then a lower bound of -1 shown by the cursor is correct, because that value is being reported in volts.

[ChucklesOnGitHub]

Hi! Thanks for explaining what the Auto range actually is, that was were my confusion started.
Will you link this to making the change in scale acq board directly or should I? Regarding units, though, I think originally the viewer was only the displaying the traces in volts and not m/s2 because the accelerometers are not calibrated. It would only make sense to display the IMU data with actual units, right? It seems the IMU channels have also been called "Aux", but should be treated differently than actual Aux channels coming from the Intan chip which could have an accelerometer or other sensor?

My only remaining qualm here would be that the lower bound of -1 is displayed with a unit of mV in the label under the channel name, when it should be V.