Cinepak's motion estimation was best at camera panning, or linear motion. Things like camera zooming do not fair well with Cinepak at low data rates.
Tomlee, you have gotten me curious about this strange phenomenon. I don't understand why Cinepak is performing poorly compared to Motion JPEG (MJPEG). Cinepak is ideally suited for 320x240 at 15fps and roughly 100-300 KB/sec. (video only), although it performs fairly well with some slight variations. For instance, I like to use 320x180, 10fps, 80 KB/sec for my 16x9 web videos.
MPEG-1 is well suited for 352x240 at 30fps, and will produce a better picture than Cinepak, but it is much more processor intensive.
Does your Cinepak test have keyframe strobing, where the keyframes look good, but the frames inbetween look bad? Are your Spatial Quality and Temporal Quality sliders all the way at "Best"? I'm stumped. I'd love to see some samples posted.
Tomlee, you have gotten me curious about this strange phenomenon. I don't understand why Cinepak is performing poorly compared to Motion JPEG (MJPEG). Cinepak is ideally suited for 320x240 at 15fps and roughly 100-300 KB/sec. (video only), although it performs fairly well with some slight variations. For instance, I like to use 320x180, 10fps, 80 KB/sec for my 16x9 web videos.
MPEG-1 is well suited for 352x240 at 30fps, and will produce a better picture than Cinepak, but it is much more processor intensive.
Does your Cinepak test have keyframe strobing, where the keyframes look good, but the frames inbetween look bad? Are your Spatial Quality and Temporal Quality sliders all the way at "Best"? I'm stumped. I'd love to see some samples posted.