All terms
Sampling Rate and Bandwidth
How often the signal is digitized — it caps the bandwidth and sets the data volume.
Sampling rate is the number of times per second an acquisition system digitizes each channel, in kHz; it caps the bandwidth that can be recorded at half the rate, so 30 kHz sampling records up to about 15 kHz. Bandwidth is the range of frequencies the chain passes, set by the analog filters and the sampling rate together. The spike band is about 300 Hz to 6 kHz and the field-potential band below 300 Hz, so 30 kHz is the working standard for spikes with margin for the anti-aliasing filter; 20 kHz is adequate for spikes and is the fixed rate of SiNAPS probes; 1 to 2 kHz is enough for field potentials alone. A wide-band recording at 30 kHz captures both, and the bands are separated by filtering afterwards.
Rate is chosen against the electrode and the storage together. A site's spike waveform is about a millisecond long and its shape is what a sorter uses, so under-sampling blunts the features that separate units; over-sampling adds nothing the site can deliver and doubles the data volume for every doubling of the rate. The SmartLink and X-Series headstages sample at up to 30 kS/s and 16 bits, in record and record-and-stimulate modes alike, and XDAQ ↗ and Smartbox ↗ acquire every headstage on one clock, so the rate is a property of the chain rather than a per-session setting.
📚 More resources
Guides, downloads, and support docs for Radiens and NeuroNexus hardware.
Open Resources Hub →