Shandra
10th May 2003, 13:51
Hi Folks,
I now meddled/explored my SAA7134 capture device (with a 2.6GHz P4 running XPHome) for a few weeks (first think to seek was how to enable 768x576 in VDubSynch/Mod - wich was just a matter to include the EnablceCapPreview=0 key to the registry) and yesterday I came to find the following registry entry (in a different branch of the reg) ->
[HKEY_LOCAL_MACHINE\SOFTWARE\Philips Semiconductors\SAA7134\VidCap]
"UseOverlaySurface"=dword:00000001
"In_Height_50Hz"=dword:00000120
"In_Height_60Hz"=dword:000000f0
"In_StartX_50Hz"=dword:0000000a
"In_StartY_50Hz"=dword:00000018
"In_Width_50Hz"=dword:000002be
"In_StartX_60Hz"=dword:0000000a
"In_StartY_60Hz"=dword:00000018
"In_Width_60Hz"=dword:000002be
Does anyone know what this entry is good for, or what apps are referencing it (the drivers?) - 1st what is puzzling me is that UseOverlaySurface=1 seems to contratict the EnableCapPreview Setting, but the Height&Width for 50Hz (704x288) is what I am curious about...
Hope that someone has some experience on this topic to share with me ;)
now to three minor questions (maybe the should belong to the VDub Forum):
I am now used to capture first to Huff, then do the filtering (deinterlace, cleaning up the picture, crop+resize) and then saving it to MJpeg(due to lack of that much HD-Space i registered PicVideos codec) to encode that with the final codec i want to use...
a) Now I noticed that if I load the first captured avi into VDub
and look at the frame information and the framerate displayed in []change so that av match is slightly off from 25 (ie. 24.997, or 25.001 or such minor deviation) - but if I capture using MJpeg it is 25 -> should that worry me (in using Huff)?
b) As it seem to me that the filtering process I apply is done using the rgb-colourspace, but most of the time for capturing you are given the information to go with yuv2, i am a bit confused - shouldn't i capture using rgb24 so that no double conversion is done while filtering? (or have I got this one wrong)
c) As MJpeg can store the interlaced information in seperated fields, should I still leave the box (two fields if y>288) checked if I am feeding the codec with progressiv-material (as descriped above in my process)? Or better asked - is there any negative effect in leaving that box checked (because I surely going to forget to enable it again if I capture directly to MJpeg) in such a case?
Thanks,
SHAnDRA
I now meddled/explored my SAA7134 capture device (with a 2.6GHz P4 running XPHome) for a few weeks (first think to seek was how to enable 768x576 in VDubSynch/Mod - wich was just a matter to include the EnablceCapPreview=0 key to the registry) and yesterday I came to find the following registry entry (in a different branch of the reg) ->
[HKEY_LOCAL_MACHINE\SOFTWARE\Philips Semiconductors\SAA7134\VidCap]
"UseOverlaySurface"=dword:00000001
"In_Height_50Hz"=dword:00000120
"In_Height_60Hz"=dword:000000f0
"In_StartX_50Hz"=dword:0000000a
"In_StartY_50Hz"=dword:00000018
"In_Width_50Hz"=dword:000002be
"In_StartX_60Hz"=dword:0000000a
"In_StartY_60Hz"=dword:00000018
"In_Width_60Hz"=dword:000002be
Does anyone know what this entry is good for, or what apps are referencing it (the drivers?) - 1st what is puzzling me is that UseOverlaySurface=1 seems to contratict the EnableCapPreview Setting, but the Height&Width for 50Hz (704x288) is what I am curious about...
Hope that someone has some experience on this topic to share with me ;)
now to three minor questions (maybe the should belong to the VDub Forum):
I am now used to capture first to Huff, then do the filtering (deinterlace, cleaning up the picture, crop+resize) and then saving it to MJpeg(due to lack of that much HD-Space i registered PicVideos codec) to encode that with the final codec i want to use...
a) Now I noticed that if I load the first captured avi into VDub
and look at the frame information and the framerate displayed in []change so that av match is slightly off from 25 (ie. 24.997, or 25.001 or such minor deviation) - but if I capture using MJpeg it is 25 -> should that worry me (in using Huff)?
b) As it seem to me that the filtering process I apply is done using the rgb-colourspace, but most of the time for capturing you are given the information to go with yuv2, i am a bit confused - shouldn't i capture using rgb24 so that no double conversion is done while filtering? (or have I got this one wrong)
c) As MJpeg can store the interlaced information in seperated fields, should I still leave the box (two fields if y>288) checked if I am feeding the codec with progressiv-material (as descriped above in my process)? Or better asked - is there any negative effect in leaving that box checked (because I surely going to forget to enable it again if I capture directly to MJpeg) in such a case?
Thanks,
SHAnDRA