This repository contains C++ classes and programs for working with the NSx and NEV data files produced by Ripple Neuro's Grapevine Neural Interface Processor.
We use the following programs to convert those data into formats that are more amenable to analysis:
-
NSxToFlac: Extract continuous wideband signals from an .NSx file to a collection of losslessly-compressed FLAC files. Metadata, in the form of Matlab .MAT (HDF5) file and a human-readable text file, is also extracted, as are analog channels. This reduces the size of the data by 2-5x, and speeds up I/O for analyses that consider only a few channels at a time. It is also easier to share a set of 50-50Mb files than one giant 300 Gb monstronsity.
-
NEVExtract: Extract digital events, spike snippets, and/or microstimulation trains from NEV files. These can be exported as Matlab .MAT (HDF5), human-readable text, or comma-separated value files. This is particularly useful if spike snippets were saved during data acquisition, because the resulting files can be annoyingly large.
This code is written in C++14, and uses
Matlab files are currently written via the Matlab C API, via a wrapper class (MatFile.cpp). This requires building the code with mex and its C++ compiler. Doing so may require that you match the Boost and LibFLAC versions with those included in your matlab install and/or build them using the same compiler that mex uses (which may not be your system compiler!).\
The class organization matches the NEV/NSx spec fairly closely. See NEVspec_2_2_vNN.pdf in the Trellis documentation.