Streaming data in IGOR PRO

That is a serious bummer. I would wish I knew that beforehand and wouldn't have wasted my time on this.

Would streaming work in Matlab? I could try to put something together, although I am not as proficient anymore in Matlab....

Alternatively, is there a workaround? Maybe reading single values and reading out the internal clock to get the timing?

Thanks!

It is for a behavior setup, so I would really need more than 100 samples/ second. 1000/sec might work, so I'll have a go at that in Igor and see how the timings work out.

If you have an updated version of the XOP, that would be great. I'll make sure to make my code flexible to also work with a possible feature in the XOP to accommodate streaming.

Please keep me posted on any progress on the XOP level.

Many thanks!

This level of customer service is beyond awesome!

Thank you so much, the code is working perfectly and is adaptable to my needs. I will play a bit more with some variants, but it is looking very promising for my goals.

Many thanks! -Laurens

Unfortunately I have encountered some serious hiccups in the code that I can't seem to sort out. Igor keeps crashing when I try to stream more than one channel. I can't figure out what I'm doing wrong, but I have the feeling it goes wrong somewhere with the requested number of scans from the Labjack. If someone has some insight in this, here is my code:

Function Stream(channels,length)
    string channels    //String or stringlist with channels (Use AIN and FIO, e.g. AIN1;FIO5;AIN0
    variable length    //length defined ms

    dfref olddf = GetDataFolderDFR()
    setdatafolder root:Packages:Labjack
    NVAR LJref,sampling,readspersecond            //Sampling is defined per channel!
    variable NrCh,err,t,i,j,buffertime
    string str,ch
    NrCh = itemsinlist(channels,";")
    variable numScansRequested = sampling/readspersecond*NrCh        // This is the length of one chunk
    Variable numReads = length/(1/sampling)/numScansRequested*NrCh    // We read the data in chunks, numReads is how many chunks there are

    setdatafolder olddf
    Killwaves /Z datawave, ConcatData
    buffertime = 1                                    // 1 second buffer configured
    // Check whether we can make this sampling rate theoretically:
    if(sampling*NrCh>5e4)
        print("Error! Sampling rate is too high. Maximal 50kHz combined for all channels.")
        return(0)
    endif
    // General configuration of the stream:
    // Setting the scanfrequency. This is how many times per second each channel is read.
    err = LJ_AddRequest(LJref, LJ_ioPUT_CONFIG, LJ_chSTREAM_SCAN_FREQUENCY, sampling, 0, 0)
    ShowError(err)
    // Setting the buffer in the Labjack. Samplingrate * Number of channels * buffertime
    err = LJ_AddRequest(LJref, LJ_ioPUT_CONFIG, LJ_chSTREAM_BUFFER_SIZE,sampling*NrCh*Buffertime, 0, 0)
    ShowError(err)
     // Configure reads to wait for requested amount of data (wait mode LJ_swSLEEP).
    err = LJ_AddRequest(LJref, LJ_ioPUT_CONFIG, LJ_chSTREAM_WAIT_MODE, LJ_swSLEEP, 0, 0)
    ShowError(err)
    // Clear any configuration of streaming channels that might be present:
    err = LJ_AddRequest(LJref, LJ_ioCLEAR_STREAM_CHANNELS, 0, 0, 0, 0)
    ShowError(err)

    for(i=0;i<NrCh;i+=1)    // Loop through the channels requested to configure them
        ch = stringfromlist(i,channels)
        if(stringmatch(ch,"AIN*"))
            // Setting the resolution of the streaming channel
            err = LJ_AddRequest(LJref, LJ_ioPUT_CONFIG, LJ_chAIN_RESOLUTION, 0, 0, 0)
            ShowError(err)
            // Setting the range of the streaming channel (+/- 10 V)
             err = LJ_AddRequest(LJref, LJ_ioPUT_AIN_RANGE, str2num(ch[3,4]), LJ_rgBIP10V, 0, 0)
            ShowError(err)
            // Adding the stream
            err = LJ_AddRequest(LJref, LJ_ioADD_STREAM_CHANNEL, str2num(ch[3,4]), 0, 0, 0) //Single ended
            ShowError(err)
        elseif(stringmatch(ch,"FIO*"))

        endif
    endfor
    // Execute the list of requests.
    err = LJ_GoOne(LJref);
    ShowError(err)
    // Get all the results just to check for errors.
    variable ioType=0,channel=0,value=0,x1=0,userData=0
    Variable LJE_NO_MORE_DATA_AVAILABLE = LJ_StringToConstant("LJE_NO_MORE_DATA_AVAILABLE")
    Variable LJE_MIN_GROUP_ERROR = LJ_StringToConstant("LJE_MIN_GROUP_ERROR")
    err = LJ_GetFirstResult(LJref, ioType, channel, value, x1, userData)
    ShowError(err)
    do
        err = LJ_GetNextResult(LJref, ioType, channel, value, x1, userData)
        if( err != LJE_NO_MORE_DATA_AVAILABLE )
            ShowError(err)
        endif
    while(err < LJE_MIN_GROUP_ERROR)
    // Start the Stream
    err = LJ_eGet(LJref, LJ_ioSTART_STREAM, 0, 0, value)
    ShowError(err)


    // The actual scan rate is dependent on how the desired scan rate divides into
    // the LabJack clock.  The actual scan rate is returned in the value parameter
    // from the start stream command.
    Printf "Actual Sampling Rate per channel = %.3f\r", value
    Printf "Actual Total Sampling Rate = %.3f\r", NrCh*value
    // Preallocating some waves. Two for the data and two for backlogs and one for samples read:
    make /O/D/N=(numReads) commBackLogWave=9999, UDBackLogWave=999, SampRead=0
    Make /O/D/N=(NumScansRequested) dataWave    // This is the length of one chunk
    make /O/D/N=0 allDataWave // allDataWave is where we store all samples. Will be concatenated
    variable commBacklog,udBacklog


    for( i=0;i<numReads;i+=1 )//Can be parallelized
        // Read the data.
        // Note that the array we pass must be sized to hold enough SAMPLES, and
        // the Value we pass specifies the number of SCANS to read.
        numScansRequested = sampling/readspersecond*NrCh
        err = LJ_eGetPtr(LJref, LJ_ioGET_STREAM_DATA, LJ_chALL_CHANNELS, dataWave, numScansRequested)
        ShowError(err)
        SampRead[i] = numScansRequested
        // Retrieve the current Comm backlog.  The UD driver retrieves stream data from
        // the U6 in the background, but if the computer is too slow for some reason
        // the driver might not be able to read the data as fast as the U6 is
        // acquiring it, and thus there will be data left over in the U6 buffer.
        err = LJ_eGet(LJref, LJ_ioGET_CONFIG, LJ_chSTREAM_BACKLOG_COMM, 0, commBacklog)
        ShowError(err)
        commBackLogWave[i] = commBacklog
        // Retrieve the current UD driver backlog.  If this is growing, then the application
        // software is not pulling data from the UD driver fast enough.
        err = LJ_eGet(LJref, LJ_ioGET_CONFIG, LJ_chSTREAM_BACKLOG_UD, 0, udBacklog);
        ShowError(err)
        UDBackLogWave[i]=udBacklog
        // Move data to our full data wave
        Concatenate /NP {dataWave}, allDataWave
// HERE WE PUT OUR MANIPULATIONS

    endfor
    wavestats /Q allDataWave
    SetScale /P x 0,(1/sampling), "s" allDataWave


end

I had a little bit of a look in event viewer and upon Igor crash it seems to have a problem with ntddl.dll

I have no clue what that might mean, or how it is caused, but it seems to be some core windows thing?