Double Precision Format

WaveBPS "wbps" Double Precision Floating Point Type

After the number of channels described in the parent topic. The floating point format includes a 4 byte integer describing the offset to the data (Table 1). You can use the offset to set the file position to the start of the data. Using the offset, the number of raw data samples in the file can be calculated with this equation. (You probably want to cast each item in the equation to an __int64 considering the files sizes can be very large)

NumberOfDataValues = (SizeOfFile() - m_iOffsetToTheData) / (sizeof(double) * (1+m_iNumberOfChannels))

Table 1 - Offset to the data

Type	Header
Unsigned int (32 bits)	m_iOffsetToTheData

At the offset there is one double precision float (8 byte value) describing the time, and then one double precision float for each channel at that time. This repeats for every data sample. The time is in microsecond units while the channel data is in volts.

For reading the wbps format the above information is the minimum needed to use the format. For writing the format you must have the following information as well.

After the offset integer are two parameters that describe the trigger (Table 2). First is a unsigned integer that indicates whether there is a trigger location present in the file. This will be either 0 meaning no trigger or 1 meaning a trigger is present. Following the trigger is a double precision parameter indicating where in microseconds the trigger exists (this must be present if a trigger is valid or not).

Table 2 - Trigger Information in File

Type	Header
Unsigned int (32 bits)	m_iHasTriggerLocation
Double Precision Floating Point (8 bytes)	m_dTriggerLocationInMicroseconds

Following the trigger is information about the decoders used when the file is saved. The decoder is stored as a convenience in the wbps file so when it is loaded it will be properly displayed (very useful when sharing waveforms).

The file format for the decoders will not be described here. For readers of the file format this data can simply be ignored. For writers of this format, you can write a signed integer equal to -1 for each channel and WaveBPS will simply ignore the decoder section when it reads the file (Table 3).

Table 3 - Decoder Ignore Sequence (repeat for each channel)

Type	Header
Signed int (32 bits)	iDecoderType (set to -1)