PCI1424 - invalid buffer with triggered onboard memory aqcuisition

[Rolboes]

Hello,

I need to capture 2 images at a very short interval using a PCI-1424 and a Kodak/Redlake camera. I have to use the onboard memory to grab the images. Using the example

\National Instruments\CVI\samples\IMAQ\Ring\ring onboard mem.prj, I added the following code in 'int CVICALLBACK Ring'

// Open an interface and a session
   imgInterfaceOpen (intfName, &Iid);
   imgSessionOpen (Iid, &Sid);
   

// ADDITIONAL CODE fOR TRIGGERING

res = imgSessionTriggerConfigure2 (Sid, IMG_SIGNAL_EXTERNAL, 0, IMG_TRIG_POLAR_ACTIVEH, 5000, IMG_TRIG_ACTION_BUFFER);
  

When the program is executed, a timer checks if a valid buffer is available:

int CVICALLBACK Imaq_loop (int panel, int control, int event,
  void *callbackData, int eventData1, int eventData2)
{
 /* definitions here */

 static unsigned int lastBufNum = 0xFFFFFFFF; 
 
 switch (event) {
  case EVENT_TIMER_TICK:

   
   imgGetAttribute (Sid, IMG_ATTR_LAST_VALID_BUFFER, &currBufNum);
   if((currBufNum == lastBufNum) || (currBufNum == 0xFFFFFFFF))
    return 0;

On my old NT system with NI-IMAQ 2.2, the value of &currBuffNum form imgGetAttribute was always 0xFFFFFFFF until the first trigger was received, so basically, return 0; was executed up to the first trigger.

On XP with NI-IMAQ 4.0 the first value of currBufNum is always 0 and the call

imgSessionCopyBuffer (Sid, currBufNum, CopyBuffer, FALSE);

waits until the trigger occurs... with the side effect, that the image is not really in buffer 0.

Is this difference between IMAQ2.2. and 4.0 intentional or am I missing something?\

Regards

Roland

[The_Chancenator]

Hi Roland,

Are you able to get the example to work with triggering?  Or is the buffer number difference causing problems?  As to how the IMAQ driver allocates buffers, NI-IMAQ Memory Allocation is a helpful article. In terms of what has changed with the IMAQ driver specifically, What's New with NI-IMAQ details everything from IMAQ 2.5 to the current version (4.0).  

Regards,

[Rolboes · Accepted solution]

I found a workaround. I use the function

ImgErrCode = imgSessionStatus(sessionID,&status, &currBufNum);

instead of

ImgErrCode = imgGetAttribute (sessionID, IMG_ATTR_LAST_VALID_BUFFER,&currBufNum);

In this case, currBufNum is 0xFFFFFFFF until the first image is recorded.

Roland