Output file size

Question

I have been away from doing video for some time.  Been doing still photos.  In that medium I have learned the tricks of the trade on how to change the pixels per inch and compression levels to minimize the size of the final product for better performance over the internet.  My question is what games do you play with videos to minimize the end product file size?

Answer 1

Jeff
Your views on the undernoted would be appreciated.
Recapping the example in post #29
File size F = B x T
Lets say bitrate B = 3000 (kB/s) and duration T = 5min = 300 seconds
Hence file size F = 3000 x 300 = 900000 kB
So returning to the previous example, instead of adjusting the Compression Ratio from 9 to 2.25, maintain the original compression ratio based on the original 3000 kB/s bitrate, that is 9. This means that when the Frame Size is now adjusted down from 640 x 480 to 340 x 240, the bitrate can go down in the same proportions. Namely from 3000 kB/s to 750 kB/s. This then reflects in the file size reducing accordingly to 225000 kB, showing a drop in file size in direct proportion to reducing frame dimensions.
I believe this addresses the original poster's request for guidance on file size parameters and shows a direct relationship between file size and frame size and also incorporates temporal compression considerations.
John
PS
Mathematically this can be expressed as:
Rearranging the Adobe File Size equation and introducing a codec compression variable:
File size F = (Frame Size K x Frame rate (fps)) x T
                                        C
Where K = ([Pixel Width x Pixel Height x Bit Depth] / 8) / 1024
and C = compression ratio of the codec
Therefore  F = B x T = (Frame Size K x Frame rate (fps)) x T ( T not equal to 0 )
                                                       C
Therefore B = (Frame Size K x Frame rate (fps))
                                        C
or
B (is proportional to) Frame Size
This shows mathematically that B the bitrate is directly proprtional to the Frame Size assuming Frame rate and Compression Ratio are held constant. Even if the Compression Ratio is not held constant, but allowed to vary through the compression process, the average/mean/notional Bitrate used to calculate the file size in the equation F = Bmean x T can still be considered to be a function of the Frame Size. This would be achieved by:
F = The Sum of  (B1 x T1 + B2 x T2 + … Bn x Tn) = Bmean x T
Again, this ignores container overheads and audio etc