As I have explained in other threads I need to use source MPG files, first convert to AVI in order to apply the HSL filter in order to avoid transcode errors which severely distort the result if I try to render directly from MPEG to MPEG.
Doing this in version one using the uncompressed AVI template the default values were: Render 360x240x32, 29.970 fps video with 44,100 Hz, 16-bit, Stereo audio. Using a fast AMD 1200 Mhz CPU based PC the rending time is roughly two to one, so a small file of just under 4 minutes takes just a little over 9 minutes to render and the results are excellent.
Repeating using the same file in version two of VF the default AVI template was changed to 720x480, so I changed the frame size to the Video CD 352x240 template which matches the source file frame size of 352 x 240. The result of trying to render a AVI this way is much slower rendering. The same file now takes over an hour compared to just over nine minutes in version one. Worse, the quality even after setting output to best is horrible with very noticeable raggy edges around all objects anytime there is motion in the video. So if someone moves their arm, there is a trailing ghost effect extending several pixels beyond where the arm actually is. Repeating, rendering to RM format selecting higher quality speed of rendering and output quality returns and is as good as in version one.
Doing this in version one using the uncompressed AVI template the default values were: Render 360x240x32, 29.970 fps video with 44,100 Hz, 16-bit, Stereo audio. Using a fast AMD 1200 Mhz CPU based PC the rending time is roughly two to one, so a small file of just under 4 minutes takes just a little over 9 minutes to render and the results are excellent.
Repeating using the same file in version two of VF the default AVI template was changed to 720x480, so I changed the frame size to the Video CD 352x240 template which matches the source file frame size of 352 x 240. The result of trying to render a AVI this way is much slower rendering. The same file now takes over an hour compared to just over nine minutes in version one. Worse, the quality even after setting output to best is horrible with very noticeable raggy edges around all objects anytime there is motion in the video. So if someone moves their arm, there is a trailing ghost effect extending several pixels beyond where the arm actually is. Repeating, rendering to RM format selecting higher quality speed of rendering and output quality returns and is as good as in version one.