Open Ephys GUI v0.5.5.4 and v0.6.0-rc3

[Josh Siegle]

Hi everyone,

We’d like to announce two new versions of the Open Ephys GUI:

===================

v0.5.5.4 (Windows, Mac & Ubuntu installers; Windows, Mac & Linux .zip files)

This is a patch release that addresses three bugs:

- Hidden channel info when using 64-channel headstages with the Rhythm FPGA module. In the latest version all channels are visible.

- GUI hangs when starting acquisition with certain signal chains that include two splitters in a row. We have now updated the processor graph connection algorithm to handle this condition. See this issue for more details.

- Dropped events on Linux for high TTL event rates (>100 Hz). v0.5.5.4 includes a much larger event buffer that prevents this from occurring. See this issue for more details.

===================

v0.6.0-rc3 (Windows, Mac & Linux installers; Windows, Mac & Linux .zip files)

This is the 3rd (and hopefully final) set of test binaries for the next major release of the GUI. This is the biggest update in the GUI’s history, so we want to make sure get it right. We have polished every plugin, added a variety of helpful new features, and simplified the process of creating new plugins. We have been extensively testing v0.6.0 at the Allen Institute, and we think it has a much better overall user experience than previous versions. We are excited for others to try it as well!

The configurations and plugins for 0.6.0 will be stored separately from those for previous versions of the GUI, so it’s fine to run them side-by-side.

There are a few things we plan to finish up before the official release:

(1) Convert additional plugins to use the new API. Currently the only plugins available through the Installer are Rhythm Plugins (includes separate plugins for the Open Ephys Acquisition Board and the Intan RHD USB Interface), the Open Ephys Data Format, Network Events, NIDAQ Plugin (Windows only), and Neuropixels PXI (Windows only). In the next few weeks we plan to add the Spike Sorter, NWB Format, Event Broadcaster, Pulse Pal Output, Ephys Socket, Phase Calculator, Multiband Integrator, and Crossing Detector, among others.

(2) Update the relevant documentation.

(3) Compute performance benchmarks across all three platforms, to allow us to provide more detailed system requirements for common recording configurations.

One important thing to note is that in v0.6.0 we have attempted to make a distinction between “timestamps” (globally consistent floating-point-valued times in seconds) and “sample numbers” (integer-valued counts of the number of samples from a particular data source since the start of acquisition). Previously, both of these things were called timestamps, leading to the potential for confusion.

If you have questions or feedback, or encounter any bugs, don’t hesitate to reach out to sup...@open-ephys.org.

Best,

Josh, Pavel, and Anjal

[Hisayuki Osanai]

Hi Josh,

It looks header information in .events file is lost when I recorded using v0.6.0-rc3.

When I tried to open .events file using load_open_ephys_data.m, there was an error message:

Loading events file...
Error using eval
Unrecognized function or variable 'Events'.

Error in load_open_ephys_data (line 101)
    eval(char(hdr'));

Error in EventAve_220606 (line 9)
[data, timestamps, info] = load_open_ephys_data(efile);

Actually, event data and its timestamps were still stored in the .events file, but it looks it can't separate which input channel the external event triggers was sent. I could only obtain a single array of the event timestamps, but it looks I cannot separate DAC ch1 input and DAC ch2 input.

Thank you,

Hisa

2022年4月27日水曜日 15:59:48 UTC-5 josh....@gmail.com:

[Pavel Kulik]

Hi Hisa,

It appears the .events file header is missing quotations around the value for the header.channel field. This is a bug in the OpenEphysFormat plugin I am fixing now. If you manually open the .events file and change "header.channel = Events;" to "header.channel = 'Events';" are you able to run the script and see the corresponding channels in the "data" field?

Thanks,

Pavel 

[Hisayuki Osanai]

Hi Pavel,

Thank you! I could read the header information after editing. However, when I edited the .events file, the variable of 'data' when I use " [data, timestamps, info] = load_open_ephys_data(filename);  " in Matlab becomes array of '100' instead of event channel index.

Instead, when I commented out the following lines of load_open_ephys_data.m and loaded the original .events file before editing, I could see which event channels were used in data variable.

Although info.recordingNumber contains the corresponding information but it is a bit difficult to interpret them.

line101-108

% eval(char(hdr'));

% info.header = header;

%

% if (isfield(info.header, 'version'))

% version = info.header.version;

% else

% version = 0.0;

% end

line630-634

% if (isfield(info.header,'sampleRate'))

% if ~ischar(info.header.sampleRate)

% timestamps = timestamps./info.header.sampleRate; % convert to seconds

% end

% end

Thank you,

Hisa

2022年6月7日火曜日 15:49:29 UTC-5 Pavel Kulik: