[M4IF Technotes] Visual Object Sequence Start Code

emre.aksu nokia.com emre.aksu nokia.com
Tue Sep 3 17:45:54 EDT 2002


Hi Christopher,
> Having multiple visual_object_sequence_start_code's without question 
> should be supported by all decoders.  There are many 
> applications that 
> need to support this including live broadcasting and other 
> applications 
> involving potential channel errors.

For sure, I have no problems with that. I think you can put in as many visual_object_sequence_start_codes as you like, as long as you have a visual_object_sequence_end_code finalizing a previous strat code, I think. Otherwise, how can we identify a serialized structure of visual object sequences?
Once again, a reference from 14996-2:2001:
Section 6.1.1:
" Visual object sequence is the highest syntactic structure of the coded visual bitstream.
A visual object sequence commences with a visual_object_sequence_start_code which is followed by one or more visual objects coded concurrently. The visual object sequence is terminated by a visual_object_sequence_end_code."
How should we interpret the first and last sentences together, if we allow randomly inserted visual_object_sequence_start_codes ending a previous visual object sequence? The very same structure ( visual_object_sequence_start_code followed by another visual_object_sequence_start_code) can be interpreted as a frame loss by the decoder, which was expecting a visual_object_sequence_end_code before a start code as stated in section 6.1.1.
Best Regards,
Emre
Emre Baris Aksu
Nokia Corp.
> -----Original Message-----
> From: ext Christopher Wendt [mailto:chris   avipix.com]
> Sent: 03. September 2002 16:03
> To: Aksu Emre (NMP/Tampere)
> Cc: technotes   lists.m4if.org
> Subject: Re: [M4IF Technotes] Visual Object Sequence Start Code
> 
> 
> Hi Emre,
> 
> Having multiple visual_object_sequence_start_code's without question 
> should be supported by all decoders.  There are many 
> applications that 
> need to support this including live broadcasting and other 
> applications 
> involving potential channel errors.
> 
> -Chris
> 
> _____________________________________
> Christopher Wendt, Avipix Technology Corp
> chris   avipix.com
> 484.951.4932
> http://www.avipix.com
> 
> 



More information about the Mp4-tech mailing list