[KC-Technerd]

From what I've read about weave type deinterlacing, I have the impression that a 60i source would normally be deinterlaced to 30 frames per second with each frame containing 2 combined fields. In 60i the field sequence would be Frame 1 Field A, Frame 1 Field B, 2A, 2B, 3A, 3B, 4A, 4B, etc. In deinterlaced at 30 frames per second, the frame sequence would be 1A+1B, 2A+2B, 3A+3B, 4A+4B, etc.


Do any TVs do weave deinterlacing from 60i by creating 60 frames per second with each frame updating one field. In 60i the field sequence would be same as above. In deinterlaced at 60 frames per second, the frame sequence would be 1A+1B, 2A+1B, 2A+2B, 3A+2B, 3A+3B, 4A+3B, 4A+4B, 5A+4B, etc.


If no one does it using this method, why not? It appears to me the 60 frame per second with each frame updating one field would better replicate true interlaced scanning.

 

[walford]

AFAIK all TVs not using Bob de-interlacing use the method you descrbe to get 60 FPS content since otherwise they would have to just display each of the 30 frames received from 1080i twice which would cause unacceptable jeakiness.

 

[fastl]

If there is no motion detected in between fields, then weaving gives the best result. Motion adaptive de-interlacing selects the method to give the best result.

 

[walford]

Then apparently motion adaptive de-interlacing will then select Vector adaptive motion de-interlacing if there is motion or just weave de-interlacing if there is not. In any case it would never select Bob de-interlacing.

 

[fastl]

Let's not turn this into another semantic pissing contest. Weave de-interlacing retains the full resolution of the original image, but doesn't work very well if there is motion between adjacent fields. Adaptive de-interlacing (by whatever name you want to call it) substitutes other de-interlacing techniques when motion is detected to prevent mice teeth and the other various well known side effects.