[Mp4-tech] H.264--Questions about POC

Ye-Kui.Wang nokia.com Ye-Kui.Wang nokia.com
Tue Feb 15 16:42:07 ESTEDT 2005


See inline, please.
-----Original Message-----
From: mp4-tech-bounces lists.mpegif.org [mailto:mp4-tech-bounces lists.mpegif.org]On Behalf Of ext Zhao Ran
Sent: 15 February, 2005 13:51
To: mp4-tech lists.mpegif.org
Subject: [Mp4-tech] H.264--Questions about POC
Hi, all
I have two questions about poc. Is there anybody could help me make them clear.
As we know, in the bitstream created by jm refenrence encoder, the poc increments 2 every complete frame, just like 0, 2, 4, 6,...... 
Then, is standard permit such a bitstream which has poc incremented by 1 for every complete frame, just like 0, 1, 2, 3,......
[YK]: No. But it is allowed to have POC values of 0, 1, 2, 3,... for a series of coded pictures.  
Another problem confused me is that is a negative poc allowed. I suppose that for IDR picture, the poc should be equal to 0, and followed pictures should have poc in ascending order. In other words, poc should be positive or zero.
However, for pic_order_cnt_type==0, poc is derived from pic_order_cnt_lsb and delta_pic_order_cnt_bottom. The former is in the range of 0 to MaxPicOrderCntLsb-1 and the latter is range of -231 to 231-1. There might be a bitstream with a bottom field picture having pic_order_cnt_lsb==2 and delta_pic_order_cnt_bottom==-4, well then the poc would be negative. Is such a bitstream valid? 
[YK]: In a coded video sequence, pictues preceding the IDR picture in output order, if present, have negative POC values.
BR, YK  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/mp4-tech/attachments/20050215/ffba31e0/attachment.html


More information about the Mp4-tech mailing list