Lupissimo
9th April 2004, 16:38
It's just a thought!
Having not too much experience in encoding I wonder:
1. Is it possible by some algorithm to determine the most critical parts in a video??
2. Is the "noise" PSNR a valid criterion for optimizing encoding??
If 1. and 2. is true I could envisage a prepocessor which scans the data for critical parts ( lets assume less then 10 each 5sec. long )and then optimizes the encoding parameters until the PSNR is max.
I know that all the gurus will clobber me and tell me I should just read the readmes etc and then just do this manually, but why should I spend hours trying to do what a computer can do for me!!
Certainly the experts know how set the parameters depending on the material, so all they need to do is formulate their criteria in a way a program could use them. The rest would be a fairly straight forward optimization.
Lupissimo
Having not too much experience in encoding I wonder:
1. Is it possible by some algorithm to determine the most critical parts in a video??
2. Is the "noise" PSNR a valid criterion for optimizing encoding??
If 1. and 2. is true I could envisage a prepocessor which scans the data for critical parts ( lets assume less then 10 each 5sec. long )and then optimizes the encoding parameters until the PSNR is max.
I know that all the gurus will clobber me and tell me I should just read the readmes etc and then just do this manually, but why should I spend hours trying to do what a computer can do for me!!
Certainly the experts know how set the parameters depending on the material, so all they need to do is formulate their criteria in a way a program could use them. The rest would be a fairly straight forward optimization.
Lupissimo