TDMS C DLL returns incorrect channel data type in MATLAB

[erintomlinson90]

Hello,

Within MATLAB 2018b, I'm attempting to use the TDMS C DLL to read images from a TDMS file that are stored as an array uint16's (data type 0x6). My first attempt at this was as follows:

inds  = startInd:(startInd + numImages - 1);
pImages = libpointer('uint16Ptr', zeros(1, numel(inds)*imageHeight*imageWidth));
status = calllib(libName, 'DDC_GetDataValues', channel.ID(groupInd, channelInd), (startInd-1)*imageHeight*imageWidth, numel(inds)*imageHeight*imageWidth-1, pImages);

However, this causes MATLAB to crash. I tried to dig into the problem by requesting the datatype for this channel using:

[status, dataType] = calllib(libName, 'DDC_GetDataType', channel.ID(groupInd, channelInd), 0);

and dataType returns 'DDC_Double', which corresponds to datatype 0xA. I've confirmed with my TDMS File Viewer that 'NI_DataType' for this channel is indeed set to 0x6 (uint16):

But for some reason, the C DLL thinks that it is a double. If I modify the above snippet of code to initialize libpointer as a 'doublePtr', then my images are read correctly:

inds  = startInd:(startInd + numImages - 1);
pImages = libpointer('doublePtr', zeros(1, numel(inds)*imageHeight*imageWidth));
status = calllib(libName, 'DDC_GetDataValues', channel.ID(groupInd, channelInd), (startInd-1)*imageHeight*imageWidth, numel(inds)*imageHeight*imageWidth-1, pImages);

Can anyone help me understand what is going on here? I've noticed that within the C header file that comes with the DLL (nilibddc_m.h), it does not appear to support uint16 datatypes:

typedef enum {
    DDC_UInt8       = 5,    // unsigned char
    DDC_Int16       = 2,    // short
    DDC_Int32       = 3,    // int
    DDC_Float       = 9,    // float
    DDC_Double      = 10,   // double
    DDC_String      = 23,   // string
    DDC_Timestamp   = 30,   // timestamp (Year/Month/Day/Hour/Minute/Second/Millisecond components)
} DDCDataType;

If there is no support for uint16 data types, then I would expect it to crash or return an error, rather than declare it to be a double. Any help or insight would be much appreciated.

Thank you,

Erin