RatVega
Enthusiast
Jan 1, 2006, 8:57 PM
Post #5 of 6
(1050 views)
Shortcut
|
|
Re: [LarsCA] FCP export for internet
[In reply to]
|
Can't Post
|
|
Like Dep says, there's no single best setting but understanding what you're doing may get you closer to what you want more quickly.
First off, you're looking for something like 50:1 compression, so no matter what you do your quality is going to be mediocre at best. Here are some things that may help you get more of what you want:
Frame Rate: This is a big player in file size because you reduce the total number of frames you'll use. When image quality is important and motion smoothness is a lower priority, substantially lower frame rates will help.
Frame size: Going from 720X480 to 360X240 doesn't cut the size in half, it reduces it to a quarter, so if detail is a lower priority, big reductions can be had here.
Keyframes: This is a less understood area for many people. In most compression the data is reduced to I-frames (actual frames from the original, also refered to in compression as keyframes) and P- and B-frames which are interpolations of an I-frame and another P- or B-frame. The higher the number of I-frames, the larger the file will be, so reducing the keyframe frequency (that is, inserting fewer keyframes) will make a smaller file. The penalty is that fast-changing or "busy" frames may get very nasty looking.
There are several codecs available for "web encoding", common ones are MPEG-4 (AKA m4v), H.264 (AKA MPEG-4, Part 10 or AVC), MPEG-1 (AKA "junk" in my book  ) and the Sorenson codecs. All have adjustments for the above parameters except H.264, which I'll cover below. Play with the codec settings to see what quality hits/size benefits each provides for your footage.
H.264 is the latest rendition of the MPEG-4 encoding family, expanded to include HD amongst other things. It is a very efficient codec, but requires significant processing power to encode and decode. This makes it a lesser choice for older, slower target machines. H.264 has been touted as being twice the quality of MPEG-4 (at the same file size) or half the file size of MPEG-4 (at the same quality.)
H.264 is different from most codecs in that it uses a "GOP-less" structure. All the above codecs use a GOP (Group Of Pictures) structure consisting of a repeating pattern of I-, P-, and B-frames with fixed relationships. H.264 uses these frame types more intelligently in a structure derived from its analysis of the clip being encoded. Clearly, there's a lot of advanced math going on, which is why it demands greater processor power.
I recently did a 5-minute clip in H.264 that was 480X360, good quality, and only about 20MB. In another somewhat specialized case (Animation codec, no audio, low [but acceptable] quality), I received a clip that was 2MB for 6 minutes.
Good luck with your project.
______________________________________________________________
Currently on a loaded 2.5GHz G5 dualie/5GB/1TB internal RAID/dual 19" monitors. Final Cut Studio, Adobe Suite, Boris RED. Shooting with Canon.
VU California Crew, Inland Empire Sub-Chapter (paragraph?)
|
