Can the filter change the signal shape?

Dear MEA team,
I'm using a 64 filter amplifier (FA64) with a home made system. The system works very well; however, I was wondering if the filter (10 Hz -10000 Hz) can influence the shape of the signal that I'm recording. I am working with cardiac myocytes from embryos of rat and/or mouse. I mean, this could influence the parameters that I'm getting afterwards in the post-processing?
What should be the recommended configuration (filter bandwidth and amplification) if I want to evaluate the parameters of the signal (Fpmax, Fpmin, etc.) by using a filter amplifier like in this case?

Thank you.

The filter settings can massively modulate the shape of the signal.
While the depolarisation peak (Na component) is very rapid and not significantly affected by the highpass filter the Plateau (Ca) and Repolarization (K) components are slow and filters will act as differentiators close to the cutoff or simply remove components if they are substantially subthreshold (that´s what filters are made for...)

In our lab we use 0,1 Hz - 3kHz or 1Hz-3kHz filterband for cardiomyocytes. The 0,1 Hz is mostly used if the focus is also on Ca component. The downside of those settings are slow drifts of the baseline...

If you send us a quick e-mail to support@multichannelsystems.com I would be glad to share some slides what filters do to the signals...

Greetings

Thomas

Hello,
I'm using the Slope-Method in MC-Rack to detect spikes in brain-slice with 60pMEA100/30iR-Ti. Until I read this post, I always applied the spike-sorting to a 300Hz highpass signal. Is this possibly a bad idea? Especially because i'm particularly interested in the waveform-shape and -kinetics.

What are, in your experience, good settings for the slope detection (I know, it probably depends on the data)? In the moment I'm going with
dV 25 µV Min -0.5 and Max 0.8
Pre- and Post-Trigger and Dead-time are all 1ms

Hi,
for spikes filtering is less critical, as there are no slow signal components which would be changed by a high pass filter so easily. You could always display filtered and raw data in overlay, to see what the filter is doing to your signal shape. The detection parameters look O.K. but in the end each user has to decide whether the detection covers everything you consider a signal, but keeps false positives to a minimum. To be more conservative, increase the dV.
As a general advice, I would always record the unfiltered data, to be able to go back to the raw data.

Best regards Frank