Fix time blackrock

[luiztauffer]

a quick fix for the miliseconds / microseconds part of rec_datetime

[luiztauffer]

@zm711 @h-mayorquin found this other error, small but important. We were getting the wrong start recording time (in the miliseconds scale)

[zm711]

@luiztauffer could you post an image just of the times before and after? Since I don't know the format it is hard for me to know when to do the factor of 1000 or not?

Or a link to some documentation :)

[h-mayorquin]

@zm711

The corresponding description of the NEV header:

Here is the spec for the timestamps where you can see that the last two bytes are milliseconds as per the windows convention and we read here:

python-neo/neo/rawio/blackrockrawio.py

Lines 1131 to 1161 in 5df3403

	# basic header
	dt0 = [
	# Set to "NEURALEV"
	("file_type_id", "S8"),
	("ver_major", "uint8"),
	("ver_minor", "uint8"),
	# Flags
	("additionnal_flags", "uint16"),
	# File index of first data sample
	("bytes_in_headers", "uint32"),
	# Number of bytes per data packet (sample)
	("bytes_in_data_packets", "uint32"),
	# Time resolution of time stamps in Hz
	("timestamp_resolution", "uint32"),
	# Sampling frequency of waveforms in Hz
	("sample_resolution", "uint32"),
	("year", "uint16"),
	("month", "uint16"),
	("weekday", "uint16"),
	("day", "uint16"),
	("hour", "uint16"),
	("minute", "uint16"),
	("second", "uint16"),
	("millisecond", "uint16"),
	("application_to_create_file", "S32"),
	("comment_field", "S256"),
	# Number of extended headers
	("nb_ext_headers", "uint32"),
	]
	nev_basic_header = np.fromfile(filename, count=1, dtype=dt0)[0]

And you can see that the datetime object requieres microseconds here:

https://docs.python.org/3/library/datetime.html#datetime.datetime

[h-mayorquin]

LGTM

[zm711]

Cool thanks I didn't realize datetime needed the microseconds. Sorry read too fast :)