shiva-x
11th May 2003, 12:42
First sorry if this method has already been discussed many times :confused:
I encode using a custom method much inspired from the RoBa method. To set the bias in CCE after the first pass, in advanced settings, I first set it to (AverageBitrate/100) and watch both avg and max Q factors. Then I decrease the bias by one unit, watching carefully how the Q factors evolve. If one Q factor value decreases while the other isn't modified or if they both decrease or don't change, I continue decreasing the bias by one unit and repeat the process. If one of the Q factor values increases, I was wrong trying to decrease the bias so I increase it. I repeat the same process, increasing the bias instead of decreasing it.
I continue modifying the bias by one unit until avg and max Q factors are both minimal.
To summarize, I try to find when avg and max Q factors are minimal, using a bias as low as possible.
This method seems to give excellent results to me, no "mosquito noise" because of a too low bias, or distorted pictures because of a too high bias.
Sorry my English is not very good... Hope I was clear enough...
What do you think of this method ? It could quite easily be implemented in programs such as DOCCE4u. I only do movie-only rips so much room on a DVD-R and my average bitrate is already quite high (about 3900 kbps at least).
Thanks
I encode using a custom method much inspired from the RoBa method. To set the bias in CCE after the first pass, in advanced settings, I first set it to (AverageBitrate/100) and watch both avg and max Q factors. Then I decrease the bias by one unit, watching carefully how the Q factors evolve. If one Q factor value decreases while the other isn't modified or if they both decrease or don't change, I continue decreasing the bias by one unit and repeat the process. If one of the Q factor values increases, I was wrong trying to decrease the bias so I increase it. I repeat the same process, increasing the bias instead of decreasing it.
I continue modifying the bias by one unit until avg and max Q factors are both minimal.
To summarize, I try to find when avg and max Q factors are minimal, using a bias as low as possible.
This method seems to give excellent results to me, no "mosquito noise" because of a too low bias, or distorted pictures because of a too high bias.
Sorry my English is not very good... Hope I was clear enough...
What do you think of this method ? It could quite easily be implemented in programs such as DOCCE4u. I only do movie-only rips so much room on a DVD-R and my average bitrate is already quite high (about 3900 kbps at least).
Thanks