Re: [Deinterlace-discuss] TimeShift update

[Nick K. <ni...@ne...>]

>> Yeah, I probably read it wrong. The thing I didn't like about it 
>> aside from the fact that it didn't seem very fast was that it was 
>> affected by the choice of deinterlacer made by the user, which also 
>> makes it seem like it would be more suseptiable to lost frames in 
>> certain conditions.
>
>
> If I am not mistaken that is exactly what would one expect. If you 
> choose odd or even scanlines only, the size of the image after 
> deinterlacing is different than from other plugins that really do 
> double the data.

Well, it's not necessarily what I would expect. It just seems iffy to 
write not always consistent data to the disk. Another thing about 
capturing the output is that the data that the frame writer gets is 
being influenced by the behavior of the deinterlacer (when to display 
the output, when to skip frames), which is different for each setting.