Hi,I'm developing tvp5158's driver with linux-2.6.32 on dm6467t board,now I need to demux the captured Superframe.But I'm still a little confusedabout the data format of 4-ch D1(PAL),BT.656,Line-interleaved mode.Q1:In 4Ch-D1(PAL)mode,the Superframe size is 728x2500(625x4),if it is stored on DDR2 as YUV422Semi-planar,what it will be?Like this ?
----------------------------------------|| || 728x2500(Luma data) | (720 Y0,Y1,Y2,...,Y720,what's the other 8 bytes?)| |----------------------------------------|| || 728x2500(Chroma Data) | | |----------------------------------------|
if I want to demux the line-interleaved video,how to realize it?
Regards,xiaoyu
Xiaoyu,
There is a TVP5158 to DM6467 driver here.
http://e2e.ti.com/support/data_converters/videoconverters/b/videoconvertersblog/archive/2010/09/20/video-decoder-driver-download-website.aspx
Other info is under announcements.
Regards,
Tom
Hi Tom,
Thank you for your reply,I'm afraid I can't use mcvip_tvp5158 driver,because the kernel is Linux-2.6.32-rc2-davinci1,dvsdk version is dvsdk_3_10_00_19,are you sure this is already the latest tvp5158 driver for DM6467T board? Thanks!
Xiaoyu
XiaoYu,
The linker Tom pointed out just is our referece code,based on one version SDK,one version linux kernel. For TVP5158,its registers' settings and demux module is the same.
I suggest you to read mcvip_dumux.c/mcvip_tvp5158.c/mcvip_tvp5158_i2c.c.
Now,about your question,superframe's structure,which is:
Metedata(8 bytes) ==>Cb0,Y0,Cr0,Y1...........(total 720*2=1440 bytes)==>Metedata(8 bytes) ==>Cb0,Y0,Cr0,Y1...........(total 720*2=1440 bytes)==>......
One metedata,one line.
Attention:these lines include VBI data,dummy lines and active lines.
hi,leon
I have realized demux funciton,according to the ChId,LineId,f_bit and v_bit information from metaData.I just want to make sure my understanding is right.Thank you again