Can you tell us more about your application in which you are planning to use OpenGL and MSDK to help you better that if that is feasible and how it can be achieved.
There is an article written a while back on texture sharing from Intel Media SDK to openGL, which you might wanna look at https://software.intel.com/en-us/articles/texture-sharing-from-intel-media-sdk-to-opengl.
Hope it helps!
Thanks for the link.
Yes we plan to start a project with the following:
media sdk server , CentOs, which shall encode muldiple HDMI channels (5) and decode HDMI channel.
The graphic additional might be some symbolic (simple text, buttons, etc).
I am interested to know if OpenGL is supported with media sdk.
Thanks for providing the information. What I am understanding from the description is that it looks like that you would like to set up 1:N transcode pipeline (1 decode followed by 5 encode sessions). To do so you don't need openGL.
If the above understanding is not correct, do you mind clarifying the pipeline and where the OpenGL fits in.
Along with that, can you please tell us the resolution of your input to decode and if you will encode at a different resolution?
Are you planning to work on CentOS 7.1 or an old version of CentOS?
Thanks for the comments,
As I wrote above, aside from the encoding/decoding we might add symbolics on the image, such as buttons or other text/symbol. The symbols might change according to externak command.
How can we achieve this symbolic graphic ? Should we use OpenGL for this capability ?
Hi Ran, OpenGL is a good place to start to add objects as you mention. For that, the article referred above (MSDK->OCL->OGL) is a good place to start - https://software.intel.com/en-us/articles/texture-sharing-from-intel-media-sdk-to-opengl. If your requirement is simpler than that, for example adding watermark or logo etc -> you can use the composition filter in VPP as well (https://software.intel.com/en-us/articles/video-composition-using-intel-media-sdk).
If your requirement is more specific that the resources above, please send us the details on your pipeline and what you layers you plan to add to your video.
I would like to use media sdk sevrer (on linux) with opengl capability.
something like the following:
<app over openGL > -------> ----- media sdk server -----> opengl framebuffer
It seems that the above link is for Windows (D3D), but I do not plan to work with windows, or windows display card, but just need the opengl surface/frames from the GPU with media sdk server (on linux).
We will need to have framwbuffer /dev/fbX which can be used for the OpenGL surface input and output buffer that can be send to display
Is that possible ?