
Author: michael Date: Mon Mar 3 15:12:12 2008 New Revision: 649 Log: Some additional words about seeking and keyframes spread. Modified: docs/nut.txt Modified: docs/nut.txt ============================================================================== --- docs/nut.txt (original) +++ docs/nut.txt Mon Mar 3 15:12:12 2008 @@ -1092,6 +1092,14 @@ B. forward seeking smallest one which has a back ptr >= the position of what was found in 1. 2. Follow the back pointer to the corresponding syncpoint. +Alternatively a demuxer can search for a shorter back ptr (which ensures that +keyframes in all streams are closer together) before 2. +One way to do this for backward seeking is: +if the back ptr is longer than the demuxer wants, step back by half the ptr, +search for another syncpoint and repeat until either the ptr is short enough +or the search went too far. + + Seeking with an index (non-normative): -------------------------------------- The demuxer only has to find the appropriate keyframe in the index and @@ -1101,6 +1109,10 @@ Note, more complicated seeking methods e seeking to the optimal point in the presence of an index even if only a subset of all streams is active. +Note2, A demuxer might wish to favor a syncpoint which has keyframes +in all active streams shortly afterwards instead of one where they are +distant. + A muxer SHOULD place syncpoints so that that simple low complexity seeking works with fine granularity. That is, syncpoints should be placed prior to keyframes instead of non-keyframes and with high enough frequency
participants (1)
-
michael