@popcormix, Ok thanks for the feedback on the frame drop thing. It seems like it would pretty much have to be like that since old analog TV only provided 29.97 fps due to technicalities of broadcasting color. So to display 30 or 60 fps would require duplicating a frame about every 10 seconds. Would a duped frame show up as a skip? IOW, what exactly causes a skip increment? I'm just trying to learn something new here? I'm old school when it comes to video, so I'm used to the analog world and it's unusual requirements. I'm in the US of course where 60Hz power frequency rules the day.
I don't know any way to provide a Live TV sample without recording, and recording covers up the problem of jerks and stutters. I'm content to assume that it's stream errors. I'm sure the broadcaster could care less and does whatever is needed to keep the primary channel's bitrate up. The fact that it's a 2Mbps stream pretty much gives that away. You should see how bad it macro blocks when there is a large amount of motion. The rotating logo is good at causing that to occur. When it's a still image, it fills the picture in nice. Fire seems to be a difficult thing to compress.
I do see some artifacts when watching TV that I take to be stream errors that might be generated at the station, or due to multipath reception. I'm beginning to think that I actually have too much signal strength. I say that because I had my antenna down on the ground one day while installing a mast mounted pre-amp, and it seemed to work just fine. My one TV that actually has a signal strength indicator showed the same strength down on the ground as when up in the air. If I turn the gain up on the pre-amp, the TV starts showing a reduction in signal strength. I'm guessing that might be caused by overload or IMD. When the weather gets a little nicer, I'm going to try some things with my outdoor antenna and see if I can come up with answers. I wish had some kind of ability to analyze the live stream. Maybe it's time to get my coding hat out and start learning how this all works in the digital TV realm.
I hate to see you or Milhouse waste time over something that might be just a fact of life on this end. I find it interesting that zaphod24 is using the same TV broadcast (I don't know where he lives). We are both using the MeTV subchannel from our local OTA broadcasters. Perhaps this is originating at the uplink end of things. I don't know what city he lives in, but I don't think it's in Houston TX. IIRC, he lives in California.
I'm happy to try and provide anything you like that might help to improve error handling for everyone using DVDPlayer. I certainly don't want you to work on something that only affects me.
The following stuff is off topic.
OT EDIT: My "expertise" in video handling ends with any thing more advanced than what a BT878 chip can do. All this MPEG2, h.264 and MPEG4 (is that the same as h.264) is a new world to me. I promise to learn it though as I have time. I've written stuff that uses V4L and V4L2 interface to the kernel, but only to grab and process analog frames. I'm very interested in porting some of my stuff to the Pi using it's built in camera. I figured out how to get /dev/video0 to appear, but have a very long way to go on understanding what's taking place below that. I got my capture/motion detection stuff to compile and link, but it pukes on this:
Code:
pi@raspberrypi ~/devel/drawonimage $ ./vidcap -d /dev/video0
Searching /mnt/cap/video0 for biggest existing frame number
Largest existing frame number found = ..
cap_count set to 0
Current input Camera 0 supports:
VIDIOC_G_STD: Inappropriate ioctl for device
pi@raspberrypi ~/devel/drawonimage $
I'll figure it out though.
Thanks again.
afremont (another long winded post