Media (Intel® Video Processing Library, Intel Media SDK)
Access community support with transcoding, decoding, and encoding in applications using media tools like Intel® oneAPI Video Processing Library and Intel® Media SDK
Announcements
The Intel Media SDK project is no longer active. For continued support and access to new features, Intel Media SDK users are encouraged to read the transition guide on upgrading from Intel® Media SDK to Intel® Video Processing Library (VPL), and to move to VPL as soon as possible.
For more information, see the VPL website.

DecodeFrameAsync is taking more than 5ms

das__monotosh
Beginner
411 Views

Hi,

In Windows platform I am using Media SDK.

API version is 1.17 and DecodeFrameAsync is taking more than 5ms time. Can you please suggest what could be the possible reason?

sts = m_pmfxDEC->DecodeFrameAsync(pBitstream, &(m_pCurrentFreeSurface->frame), &pOutSurface, &(m_pCurrentFreeOutputSurface->syncp));

Regards,

Monotosh

0 Kudos
3 Replies
Jeffrey_M_Intel1
Employee
411 Views

There could be many possibilities to consider.  To narrow things down could you send more info?

  • Processor, driver version, and OS (Version of Windows)
  • Which codec?
  • What resolution?
  • How many simultaneous decodes?

Also, how are you measuring?  Decode only or decode->render, decode->write to file, etc?  

Can this behavior be replicated with sample_decode or one of the simple_decode tutorials?

Regards, Jeff

 

0 Kudos
das__monotosh
Beginner
411 Views

Hi Jeff,

Thank you for your reply. I am trying to reproduce the same with the sample codes.

I am having intel i5 4210U and media sdk API version is 1.17 running Windows 10.

Can you please suggest me some document which will give me some idea about how many simultaneous decoding of 1920*1080 @25fps is possible in the above mentioned processor.

Best regards,

Monotosh

0 Kudos
Jeffrey_M_Intel1
Employee
411 Views

Sorry for the delayed reply.  I don't know of a document to point you to, but the best way to start exploring performance on your hardware is with the sample applications.  For decode, you can run sample_decode with no output to measure decode FPS by itself.  You should see standalone decode performance in the high hundreds of FPS.  You can also render from that sample.  Output file I/O is unoptimized -- if you write out to file nearly all of the execution time is in NV12->I420 color conversion and fwrite.   

0 Kudos
Reply