[M4IF Technotes] Reference software Error Resilient

[Rob Koenen]

Exactly. 
It should be clear that the objective of providing reference
code is not to help everyone in business with highly optimized software
at no cost. While setting an interoperable standard, MPEG must leave
room for competition between vendors, and must also see to it that
contributors to the standard are not asked to give away their
competitive advantages. 
Only the elements needed for interoperability are standardized.  
Hence:
- there is no standardized encoder, and don't expect the reference 
  encoder to be optimized (Although there is another part of the
  MPEG-4 Standard, part 7, which *will* provide a video encoder that is
  optimized. This is going to be published early next year.)
- Do not expect reference decoders to be optimized beyond what it normative.
  Error resilience tools are there to help a decoder cope with non-compliant
  bitstreams that can occur as a result of errors. The response to such 
  non-compliant bitstreams is not normatively prescribed in the standard. 
  The tools have been shown to allow implementations to do a pretty decent 
  job at coping with errors, but these implementations are private to the 
  companies that built them.
Rob
> -----Original Message-----
> From: Tomo [mailto:tomotohara   yahoo.com]
> Sent: Friday, December 21, 2023 9:09
> To: Kris Huber; 'Chang Yoong Choon'
> Cc: technotes   lists.m4if.org
> Subject: RE: [M4IF Technotes] Reference software Error Resilient
> 
> 
> Hello.
> 
>  I think this is one of the portions left open to "decoder system"
> designer and it is one of the technologies that each indivisual
> or corporation can add proprietary algorithm.  There are several
> technical papers.  
> 
>  Even MPEG-2, I noticed certian DVD players are very stable against
> stream errors, but some not-well-designed DVD players hang up 
> very miserablly. 
> 
>