Machine Vision

cancel
Showing results for 
Search instead for 
Did you mean: 

PCI-1426 losing first pixel of first frame being acquired

Highlighted

I am using a PCI-1426 board with latest NI-IMAQ software to acquire a sequence of unsigned short (uint16) frames. The first frame acquired is missing the first pixel (2 bytes) of the frame.  All the rest of the pixels in the frame are shifted by the 2 bytes. All the other frames after the first frame in the sequence are correct. Using a PCIe-1427 board and the same camera and software all the frames are acquired correctly. Is there a setting in the PCI-1426 configuration file that is missing or needs changing? It seems that I have a timing problem.

 

Any suggestions would be appreciated.

John A. Lively
NDE Engineer
Pratt & Whitney
0 Kudos
Message 1 of 6
(1,634 Views)
Highlighted

Hey John,

 

Can you tell me what version of IMAQ you are using? Also, what is your frame rate, and doesn't decreasing it affect this issue? 

 

I found some suggestion that it could be related to exposure time? Does changing that resolve the problem?

 

Regards,

A. Zaatari

National Instruments
Applications Engineer
0 Kudos
Message 2 of 6
(1,627 Views)
Highlighted

I am running NI-IMAQ Version 2010-8. We have tried frame rates from 60Hz - 125Hz (normal operating range) with the same results. The camera is a Flir SC6701. The camera resolution is 640 x 512 with 16 bit data.

John A. Lively
NDE Engineer
Pratt & Whitney
0 Kudos
Message 3 of 6
(1,621 Views)
Highlighted

Hey John,

 

The NI PCI-1426 comes with either 16 or 32 MB of onboard SDRAM. However, the PCIe-1427 uses the PCI Express interface to stream data at high speed directly to PC memory. Are you creating a buffered acquisition, can you take a snapshot of your code?

 

Regards,

A. Zaatari

National Instruments
Applications Engineer
0 Kudos
Message 4 of 6
(1,608 Views)
Highlighted

 

Yes, I am using buffered acquisition. Again after the first frame the remaining 10 - 60 frames are all acquired correctly. Only the first frame is corrupt. Yes, I know the PCIe-1427 is a faster card. On new systems this is the board we are using but I must also support older systems. Enclosed is the code segment being used to acquire frames with both cards. Could this have anything to do with the camera configuration file?

John A. Lively
NDE Engineer
Pratt & Whitney
0 Kudos
Message 5 of 6
(1,606 Views)
Highlighted

Hey Pratt,

 

If you try grabbing in Measurement & Automation Explorer does it show the same behavior? The camera file seems to be working fine on the PCIe card so I am not sure if that our culprit. If you change the PCI slot of the 1426, does that affect anything?

 

Regards,

A. Zaatari

National Instruments
Applications Engineer
0 Kudos
Message 6 of 6
(1,597 Views)