Spoke V1.0 has been released! Please download the latest version of Spoke using the links to the left.
IMEC probe support is possible in the future, please let us know if you have an interest in using Spoke with IMEC probes!
What is Spoke?
Spoke is a MATLAB app for live display of neural data from high-channel-count recording systems. It offers multiple experimenter-optimized display modes based on data reduction of high-channel-count data streams.
Spoke is designed around the Whisper extracellular recording system developed at HHMI/Janelia Research Campus in the APIG lab of Tim Harris and its accompanying [SpikeGLX] (https://github.com/billkarsh/SpikeGLX) neural recording software.
Spoke is developed & maintained by the APIG Lab at HHMI/Janelia Research Campus.
MATLAB 2015b Note: If you already have a license for MATLAB, you can install MATLAB 2015b and simply register using your existing license. This is important since many users of MATLAB may have initially purchased newer versions of MATLAB. This will not prevent you from installing MATLAB 2015b and using Spoke. Spoke may work with later versions of MATLAB, but it's not guaranteed.
National Instruments Measurement & Automation Explorer (NI MAX) (This is now a component of the full NI System Configuration (aka NI DAQmx) software suite.)
SpikeGLX, written and maintained by Bill Karsh. Spoke has been tested with the latest version of SpikeGLX (Release 20161201). Note:You will need to start SpikeGLX and have an acquisition running prior to starting Spoke. Spoke requires a running SpikeGLX acquisition in order to operate properly.
Getting Started
Step 1: Connect the Whisper box to your NI Multifunction DAQ via the shielded cable.
Step 2: Connect the Chronic Headstage to the Whisper box.
Step 3: Start up SpikeGLX
3a. Click on the SpikeGLX executable
3b. In the "File" dropdown, select "New Acquisition".
3c. In the "Devices" tab, click on the "Detect" button.
The lower box should list all of your NI devices,
including the NI-6366 you have connected to the WHISPER box.
3d. In the NI Setup tab, you can change the select the
muxes assigned to neural channels and analog channels.
If you want to operate in Triggered Waveform or Raster mode,
you must define MA mux analog channels.
3e. The Gates tab is typically unused for Spoke.
3f. The Triggers tab is typically unused for Spoke.
3g. The See N' Save tab is important. You can perform
channel remapping (see SpikeGLX documentation), and you
can also define the Channels to Save to disk during an
acquisition. Spoke will only see those channels you list
in the "IM" section. Important note: If you are using
triggered modes, you must add the trigger line here.
Step 4: Start up Matlab
Step 5: Start up Spoke
Important: Make sure that the directory that spoke is located in is added to your path!
Warning: Connection refused (10061)
> In SpikeGL/ChkConn (line 13)
In SpikeGL (line 34)
In SpokeModel (line 242)
In spoke (line 25)
Warning: The following error was caught while executing 'SpokeModel' class destructor:
Undefined function 'isvalid' for input arguments of type 'double'.
> In spoke (line 25)
Error using SpikeGL/ChkConn (line 14)
Unable to connect to server. Please make sure SpikeGLX is running, and an acquisition has been started. If SpikeGLX is running on a separate machine, please specify the IP address.
Error in SpikeGL (line 34)
s = ChkConn( s );
Error in SpokeModel (line 242)
obj.hSGL = SpikeGL(sglIPAddress);
Error in spoke (line 25)
hSpoke = SpokeModel(ipAddress);
SpikeGLX is not currently running an acquisition. If SpikeGLX is running on a different host from the one running Matlab/Spoke, then you must specify the IP Address of the host running SpikeGLX when starting spoke:
spoke('192.168.1.10')
WARNING. A large number of queued-up samples to read detected: 1587. If intermittent, this should not cause a problem.
WARNING. A large number of queued-up samples to read detected: 16312. If intermittent, this should not cause a problem.
WARNING. A large number of queued-up samples to read detected: 14344. If intermittent, this should not cause a problem.
WARNING. A large number of queued-up samples to read detected: 14518. If intermittent, this should not cause a problem.