cortex-lab / phy

phy: interactive visualization and manual spike sorting of large-scale ephys data
BSD 3-Clause "New" or "Revised" License
311 stars 157 forks source link

ISI < 0.4 ms and Correlogram #1132

Open Idavr opened 2 years ago

Idavr commented 2 years ago

Hi there!

I kept stumbling upon units that had a huge peak at 0 in the Correlogram, but nothing showing when examining the ISI view. However, after a it I found that the default for phy is to to show the ISI < 0.4 ms, so when adjusting the ISI to true 0 I know see the huge bin also in this view. However, I keep wondering why Kilosort consistently categorizes these units as Good (if there is no other/very low contamination in the refractory period). Thus, I was wondering why phy has this cut-off as a default and if this center bin peak is true spikes or is it something else? If this bin is a huge collection of true spikes then why are these units categorized as single units?

Example unit below:

image

Any input would be greatly appreciated!