[M4IF Technotes] Fw: Why not bring forward another recommenda tory standards for the real-time multimedia system?

Fri Jul 18 12:33:07 EDT 2003

Would you mean more like another profile and/or put more levels into SP?
I am seeing this is a big issue for DivXNetworks. I followed a discussion on the doom9 forum and they pretty much have given up on ASP. Their claim is that chip manufacturers are not willing to add Qpel interpolation and GMC to their decoder ICs. You hit the nail in the head regarding MPEG-2 reuse. I think that to support MPEG-4 SP at SDTV resolutions it is only a matter of firmware update on MPEG-2 chips. Chip manufacturers maximize their ROI this way. B-frames seems the only ASP tool that DivXNetworks is requiring.
 I think there are a lot of implications to the way DivXNetworks is going. One of them is that  any of these "DivX certified" chips does not conform to any profile/level in the standard, so I would wonder what are the implications regarding patent licensing.
Also, what would be then the overall improvement over MPEG-2. I'll say minimal...So I wonder if the MPEG is sensitive to this issue.
Regards,
Gerardo
-----Original Message-----
From: ahan [mailto:caogd   irercn.com] 
Sent: Thursday, July 17, 2023 2:52 AM
To: technotes   lists.m4if.org
Subject: [M4IF Technotes] Fw: Why not bring forward another recommendatory standards for the real-time multimedia system?
>Recently, I came to an exhibition and found a type of MPEG-4 encoder 
>products, which had implemented Simple Profile   level 3, along with G.729 audio and UDP transport. "It could transmit DVD images within the speed of 500kbits/s." He told me. And I asked how coule it implemented so high compression ratio? The answer is that all the coding algorithms came from Standard MPEG-2, and only one rectangular VO. Of course, the image quality was worse than the DVD source.
>
>So I have one question:
>It seems hard to implement the advanced MPEG-4 algorithms in 
>real-time,embedded MPEG-4 encoder products unless the performance of 
>chips will enhance greatly. Why not bring forward another 
>recommendatory standards for the real-time multimedia system?
>
>How do you think about this problem?
>
>Thanks a lot.
>
>guodong

_______________________________________________
Technotes mailing list
Technotes   lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes