Re: [Deinterlace-discuss] Problems fixed

[Rob M. <rob...@gm...>]

At 15:55 22-06-05, Atsushi Nakagawa wrote:
>Isn't it only necessary that the code in question measure out a delay of
>*at least* 20 us, and not necessarily an exact 50 kHz event?

I was under the impression the latter was true. I am not sure though.
It would make things very simple when just a certain delay is needed.

When looking at the way the delay is currently handled it seems 
unlikely that just a minimum delay is needed.

I have done some testing and it seems the delay does not really 
matter, at least on my machine. Multiplying or dividing the delay by 
1000 does not result in a noticeable difference.
Maybe this means we should just switch to the high resolution timer. 
Unfortunately I can't be sure if other capture cards are working in a 
similar way.

>(Aside: What criteria determines whether 200,000 times per second is
>enough?  Since it's four times 50 kHz, isn't it sufficient?)

What is the slowest machine DScaler should run on? On a p2 200MHz 
roughly 100,000 measurements can be made. Without knowing anything 
about the requirements I guess 2 times 50 kHz is not fast enough to be safe.

Rob