The images contain one screen capture of
VirtualDub. The other image is a screen capture of the way VLC player renders the video, which is not a direct capture from the video. Different players render the same video differently, so a screen capture of a media player rendering a video is not so accurate.
One visible difference is that the AVT will have slightly higher gamma, which is something already known about the AVT-8710 and which can be corrected during capture with
VirtualDub's capture setup or later with any number of other filters. Other external tbc's such as the TBC-1000 have different effects, such as visibly softening the image.
Just taking one image alone instead of an actual video sample, VLC player will not deinterlace a paused image (see capture 3 and capture 4) -- another reason why screen captures and different rendering devices can lead to different and often inaccurate conclusions. The most accurate way to make frame captures is to use VirtualDub to copy a frame directly to the clipboard. Use "Video" -> "copy source frame to clipboard".
It is often the case that an internal tbc and an outboard tbc together can generate "jumpy" or unstable video. The built-in line tbc of the player is the preferred tbc if you have turn off one of them. Some of the images have what appear to be dot crawl effects, but with screen captures it's often impossible to say for certain. Some of the VLC images look blurred.
