From: Siyuan X. <xia...@gm...> - 2011-03-17 23:20:28
|
Dear Médéric, Thank you very much! It works! I download the latest version and use PC.c instead of H264_plg.c, I do not know how to use H264_plg.c. Is this used for TCPMP? Anyway, every layer can be decoded. :) Regards, Siyuan Xiang --- On Thu, Mar 17, 2011 at 02:19, Médéric Blestel < Med...@in...> wrote: > Dear Siyuan, > > It seems that the stream is correctly decoded with the last version of the > decoder. > Can you confirm it? > > I join you the last version of the h264_plg.c file. > > Kind regards, > Médéric > > > Le 14/03/2011 20:51, Siyuan Xiang a écrit : > > I see. Thank you Mickael and Mederic. > > I get an access violation when decoding a bitstream, which has 6 layers. > The configuration is 320x180 (2 quality layers), 640x360 (2 quality > layers), 1280x720 (2 quality layers). Each layer has the same frame rate > 24. > > The lower resolutions can be decoded without problem. But when I set the > DqID to 32 or 33, decoding the 1280x720 resolution, I will get access > violation at mb4x4_mode [mode](ptr_img, PicWidthInPix, residu -> AvailMask, > locx, locy); in file 264_baseline_decoder\lib_baseline\decode_MB_I.c. > > It seems that mode is larger than the array size. > > I divide the source video into segment of 17 frames and encode them > separately, then concatenate them together. The bitstream has 17x5 = 85 > frames. > > > > > Regards, > Siyuan > --- > > > > On Mon, Mar 14, 2011 at 02:04, Médéric Blestel < > Med...@in...> wrote: > >> Dear Siyuan, >> >> To clarify the ghost picture discussion. >> *ImageToDisplay is equals to 2 when a temporal layer has been detected in >> another layer. >> >> This mechanism helps to keep the frame rate between all layers. >> In fact, when temporal scalability is present, each layer has got >> different number of frames. >> >> For instance, with a 2 temporal layer (and 1 enhancement layer): >> - Base layer: 150 frames at 15 fps >> - Top layer: 300 frames at 30 fps. >> >> Each layer has to be decoded at different frame rate, between the stream >> duration is the same (10 s). >> Without this mechanism, the base layer would be decoded at 30 fps instead >> of 15 fps. So to keep the same video duration, >> *ImageToDisplay is set to 2, and a frame is displayed, and will be >> displayed again with *ImageToDisplay = 1. >> >> Kind regards, >> Médéric >> >> >> Le 14/03/2011 09:26, Mickaël Raulet a écrit : >> >> With a temporal layer with a lower frame rate, we are inserting ghost >>> pictures to give the same framerate of the enhancement layers. >>> Mickaël >>> >>> Le 14 mars 2011 à 09:23, Siyuan Xiang a écrit : >>> >>> Hi, all, >>>> >>>> I downloaded the latest source from svn, now all the frames can be >>>> flushed out. >>>> I tried a 17-picture bitstream, but found 18 pictures are displayed. The >>>> second displayed picture is a ghost picture (I checked *ImageToDisplay which >>>> is equal to 2). >>>> What is a ghost picture? Can I just ignore the ghost picture? >>>> >>>> By the way, I found the decoding is faster than before :) >>>> >>>> Regards, >>>> Siyuan >>>> --- >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> Colocation vs. Managed Hosting >>>> A question and answer guide to determining the best fit >>>> for your organization - today and in the future. >>>> >>>> http://p.sf.net/sfu/internap-sfd2d_______________________________________________ >>>> Opensvcdecoder-support mailing list >>>> Ope...@li... >>>> https://lists.sourceforge.net/lists/listinfo/opensvcdecoder-support >>>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Colocation vs. Managed Hosting >>> A question and answer guide to determining the best fit >>> for your organization - today and in the future. >>> http://p.sf.net/sfu/internap-sfd2d >>> _______________________________________________ >>> Opensvcdecoder-support mailing list >>> Ope...@li... >>> https://lists.sourceforge.net/lists/listinfo/opensvcdecoder-support >>> >> >> > > ------------------------------------------------------------------------------ > Colocation vs. Managed Hosting > A question and answer guide to determining the best fit > for your organization - today and in the future.http://p.sf.net/sfu/internap-sfd2d > > > _______________________________________________ > Opensvcdecoder-support mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/opensvcdecoder-support > > > |