Capture Frame with PXIe-1491 and videoMaster

[DanielID]

I actually have a 6545 in the same chassis, I'm going to give that a try.

[DanielID]

It looks like you are correct.  The code works just fine on my old PXIe-6545...

Why doesn't it work for my pxie-1491?

Daniel

[DanielID]

I took a look at https://decibel.ni.com/content/docs/DOC-28699 and it kinda works.  The coloring was all off and the resolution was wrong.  I'm acquiring a 1920 by 1080 signal and the application created a 2750x1125 image.  I guess the good news is that I was able to extract the data from the hardware buffer but how to reconstruct it into an image is still the problem.  

Seems to me that you guys need a simple VMS frame grabber API function for that PXIe-1491...

[Leah-L]

This is very strange. I am wondering if there is a difference in the way that the 1491 acquires the video and saves it to a VBF file. 

It doesn't really surprise me that the image was being saved strangely. I spoke with the guy who wrote the program, and he had an application where he needed to manually reconstuct the image in LabVIEW, which probably means that this example will only work with signals of the same size. 

I want to try using the VBF to image file VI with a VBF file acquired with a 1491 to see if I am getting the same results with it saving the picture as a binary file. 

[DanielID]

The VBF to file vi works when you give it a vbf file, I have already tested that out.  It just seems that the function can not get the data from the vms buffer and write to a file when using a pxie-1491. 

Is this going to be one day added? At the moment we really need it.  

Using the example that reconstructed the image; would you know how to reconstruct that image for a 1080p signal?

Daniel

[Tarek]

Hello Daniel,

I just got an update from R&D that we have identified a bug in the LabVIEW VMS API that we are currently looking into. The previously mentioned community example probably won't work in your case unless you're using unencrypted content. I have to point out that once we address the bug, the output image resolution will be reduced due to standards with handling encrypted media.

I apologize for the inconvenience.

[Tarek]

Daniel,

After some discussion, a workaround for your issue was proposed.  We can leverage the fact that PQA is made for dynamic content and save our stream as an AVI file or user processor.  Probably an AVI file will be best.  We can then use the Vision Development Module to save each frame invididually or save the stream as a VBF and reopen it and use a user processor to save each frame individually.  We are suggesting this method because if we already know we want to capture multiple frames, it will be better to capture them together, instead of one at a time to avoid the processor intensive process of opening and closing our resources.

[DanielID]

Capturing a stream to an avi file then post processing it to get a single image is a bit heavy...  At the moment I already have a function that uses the vms capture to file function that works with the pxie-6545, ideally I would like to have this function work with the pxie-1491 by simply changing an input parameter to select either the pxie-1491 or the 6545 device.

I'm currently using the vms capture to image function to navigate a menu system on a box that is like a dvd player, I then read the items from the UUT's menu using NI Vision's OCR functions.  It will be important to go through the menu system fairly quick because this will also be used on the manufacturing.

Daniel

[RGSchmidt]

Do we have a new version of VMS that addresses the problem of saving the captured image when using a 1491?

Thanks

Randy

[Matt_L]

Hi Randy,

NI VMS 3.2 is still the latest version, I don't have an official timeline for when the new version will be released or if this feature will be included.  If the workaround above isn't a good solution for you, I would start a new topic and provide more details in order to get more attention to your issue.