Scan-based material on non-scanning displays

PseudoRandom

Active Member
I think :laugh: I have a reasonable understanding of the principles of what a scalar and de-interlacer are up against when converting, for example, 576i video material to a 1080p stream.

There is something that has bugged me for a while though, that I've never found a good answer to:

With any image that was recorded from a scanning sensor like a video camera (let's ignore interlace which will just confuse the issue!), as opposed to a freeze-frame technique like film, there is a finite time between the top (start) of the picture and the bottom (end). Therefore, just like the root cause of the combing from a simple weave de-interlace, there are temporal effects that can cause visual distortion.

In this particular case I'm thinking of a simple panning effect being shown on an LCD, plasma or projector. If it's panning horizontally, you'd expect the recorded bottom lines to "lag" the top lines, resulting in vertical lines becoming diagonal when shown on a non-scanning display. A vertical pan would result in the scenery being compressed or expanded vertically. The effect may be slight, but probably noticeable for faster movement.

Do scalars (any price) ever try to compensate for this? One of the major problems I see with trying is that it would have to effectively "invent" parts of the image to put in the gaps it produced when it skewed the image. Or maybe this information can be found in other frames?
 

The latest video from AVForums

LG G1 OLED Evo TV and SVS SB-1000 Pro subwoofer reviews, Samsung OLED rumours and more...
Subscribe to our YouTube channel
Support AVForums with Patreon

Top Bottom