I ran a little experiment with blocking reads (like cat or mplayer perform) to see exactly how well the cx18 driver driving my HVR-1600 74041 LF responds to constant requests for data compared to the ivtv driver driving my PVR-150MCE.
The result is that the ivtv and PVR-150MCE have one outlier delay at startup and then some consistent delay "quanta". The cx18 driver has some consistent delay "quanta" as well, but also outlier delays that occur often. Here's how you can see for yourself (/dev/video0 is ivtv, /dev/video1 is cx18 in my setup): $ strace -o vid0-trace -e trace=read -T cat /dev/video0 > /dev/null --- After 60 seconds in another window $ ps axf $ kill -9 ... $ strace -o vid1-trace -e trace=read -T cat /dev/video1 > /dev/null --- After 60 seconds in another window $ ps axf $ kill -9 ... $ wc -l vid*-trace $ sed -e 's/^.*<//' -e 's/>$//' vid0-trace >vid0-interval $ sed -e 's/^.*<//' -e 's/>$//' vid1-trace >vid1-interval $ tail -n 2 vid*-interval --- Make both files the same length omitting the last 2 lines --- from the shortest file $ head -n 11648 vid0-interval > vid0-int $ head -n 11648 vid1-interval > vid1-int $ gnuplot gnuplot> plot [0:11650] [0:*] 'vid0-int' gnuplot> replot 'vid1-int' gnuplot> plot [0:11650] [0:0.2] 'vid0-int' gnuplot> replot 'vid1-int' The ivtv data set for me has one delay "way out there", The cx18 data set for me has 37 delays "way out there", both were take over the course of 60 seconds. R, Andy _______________________________________________ ivtv-devel mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-devel
