This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM572x VIP inline scaling

Other Parts Discussed in Thread: AM5728

Hi,

I had earlier posted a query on VIP inline scalar: https://e2e.ti.com/support/arm/sitara_arm/f/791/t/529037

In addition to the above queries, I have few more which are listed below:

1 ) The AM5728 TRM doesn't clearly mention the limitations of VIP inline scalar. Can you please help me figure out the capabilities of VIP inline scalar? I am interested in knowing what is the maximum scaling ratio of the vertical and horizontal scalar.

2) Can VIP inline scalar down scale an incoming video to any lower custom resolution? i.e for example can VIP inline scalar down scale a 1280x720p60 fps input video to 168x96p60 fps along with Chroma down sampling.

3) I have noticed the below comment in the VPE driver (TI linux PSDK 02.00.00.00)

/*
* the line stride should 16 byte aligned for VPDMA to work, based on
* the bytes per pixel, figure out how much the width should be aligned
* to make sure line stride is 16 byte aligned
*/

 Is this a limitation of VPDMA? Is there any work around if the line stride is not 16 byte aligned?

Regards,
Prasad.

  • I will forward this to the software team.
  • Prasad said:
    1 ) The AM5728 TRM doesn't clearly mention the limitations of VIP inline scalar. Can you please help me figure out the capabilities of VIP inline scalar? I am interested in knowing what is the maximum scaling ratio of the vertical and horizontal scalar.

    This has been responded here - 

    https://e2e.ti.com/support/arm/sitara_arm/f/791/p/529037/1941421#1941421

    Prasad said:
    Can VIP inline scalar down scale an incoming video to any lower custom resolution? i.e for example can VIP inline scalar down scale a 1280x720p60 fps input video to 168x96p60 fps along with Chroma down sampling.

    No. Check the response in above e2e post. 

    Prasad said:
     Is this a limitation of VPDMA? Is there any work around if the line stride is not 16 byte aligned?

    You can allocate buffer with stride size as nearest ceil multiple of 16, set the buffer stride as that and access the buffer as 2D with the actual non-16 byte aligned resolution. Try if this works.