jptheripper
27th April 2004, 18:09
Not sure if this is possible, however:
For this question I am assuming that the dynamic bitrate assignment is based on a target cell size versus # frames in that cell. Once passing through cce with the assigned high/low/avg bitrate you produce an end size m2v that is pretty close to target. Also, I am unsure of how close this typically is.
1. Would it be possible to, after each encode, determine the size of the just encoded m2v, compare it to target (which should still be a function of stored variables from prepare) and determine how much cce missed by? I.e. cell should be 143mb, was 142mb, therefore 1mb leftover?
2. Could this leftover space (in this case 1 mb) be reassigned to the next cell to be encoded (i.e. recalculate bitrate in .ecl) So space isnt wasted?
3. I believe this would be most optimal if the cells were encoded largest to smallest (most opportunity to miss) instead of in film order, I imagine this isnt much of an issue other than reordering the ecl.
Just a thought to get every ounce of potential out since this program is so awesome as it is.
-jp
For this question I am assuming that the dynamic bitrate assignment is based on a target cell size versus # frames in that cell. Once passing through cce with the assigned high/low/avg bitrate you produce an end size m2v that is pretty close to target. Also, I am unsure of how close this typically is.
1. Would it be possible to, after each encode, determine the size of the just encoded m2v, compare it to target (which should still be a function of stored variables from prepare) and determine how much cce missed by? I.e. cell should be 143mb, was 142mb, therefore 1mb leftover?
2. Could this leftover space (in this case 1 mb) be reassigned to the next cell to be encoded (i.e. recalculate bitrate in .ecl) So space isnt wasted?
3. I believe this would be most optimal if the cells were encoded largest to smallest (most opportunity to miss) instead of in film order, I imagine this isnt much of an issue other than reordering the ecl.
Just a thought to get every ounce of potential out since this program is so awesome as it is.
-jp