Can’t you compress any data with a compression program? Why don’t I hear on mp3 video, but only video with mpeg2 compression, a la DVD?
You’re going to have to re-write the question in English if you want to get a decent response to this.
I think that MP4 is going to do for video what MP3 did for music.
My WAG (I’ve seen other people use this acronym, I still don’t know what it stands for…) is that since you know you’re dealing with sound, you can take certain shortcuts during the compression of the data and approximate. When it is then uncompressed and played, the data isn’t EXACTLY the same, but close enough so it sounds right. This gives a smaller source range, which allows for better compression.
Doing the same thing with video would probably cause a messed up picture when it is uncompressed and played back.
Mpeg compression is used for video because it is designed for video, thus not messing up the picture TOO bad when it’s played back.
Closest thing to a “MP3 Video” is exactly the MPeg2 used on DVDs. But only the names are alike; the type of compression used is completely different.
Mp3 (MPEG 1 audio layer 3) is a compression scheme designed specifically for audio. The compression is based on a psychoacoustic model–sounds that the human ear cannot perceive are thrown out (to make a very long story very short). To get good compression for video you have to use a completely different algorithm.
To expand further on this, MPEG2, MP3, and MPEG4 are what’s known as “lossy” compression alorithms, in which the developers take advantage of deficiencies in human sight/hearing to remove data that would not be seen/heard (among other things). If you expand a file compressed with MPEG2/3/4, you won’t get the same as the original. If you then compressed the expanded data again, you’d get a poorer copy. Compare this with a Lempel-Ziv compression program (like Winzip) which is lossless, and you can go through an infinite number of compress/expand cycles without changing the original data. More suited to computer data.
BTW, c_goat, WAG=Wild Assed Guess
The fact that you are asking the question shows you do not understand what is involved. Compression techniques work with different efficiency on different kinds of data. General compression schemes which may be optimized for English text will probably be pretty bad for other types of data like sound and images.
If you consider no loss compression, you can easily prove that some files will be larger after processing.
In lossy compression you also have a whole set of circumstances. JPEG takes the raw color bitmap and separates luminance from color. It then compresses the luminance to a certain degree and it compresses the color even more since the eye is more sensitive to luminance than color. In the process the picture is degraded. You can easily notice this if you enlarge a JPEG compressed picture.
In the case of movie you can compress ot only using adjacent pixels of the same frame but also subsequent frames which are very similar.
All these techniques are very different and unapplicable to audio compression which uses totally different techniques.
Compression is a process necessarily optimized for a certain type of data. Suppose my organization uses quotes from the bible all the time. Then my compression scheme could devise a system where the quotes were replaced by certain keys. But this scheme would be useless for other types of data.
Thanks for the explanations. I think I get it enough for my needs (very meager). But I understand that MPEG 4 will be for video, and that Quicktime has put into practice some things that may be included?
MPEG 4 will be for video, just as MPEG 1 and 2 were. (MPEG 3 was declared obsolete before development was complete).
Again, the misleadingly named MP3 audio compression scheme actually refers to MPEG-1, Layer 3.
About.com has lots of links to MPEG stuff:
http://desktopvideo.about.com/compute/desktopvideo/cs/mpeg/index.htm
In short, though:
MPEG-1 was for early video, including video cds, etc. Decent quality (some have said, comparable to VHS). MP3 is the audio component from this scheme.
MPEG-2 is the scheme used for DVDs. Better Quality.
MPEG-3 was going to be used for HDTV, then they found out that MPEG-2 could handle most of what they planned, and it was scrapped.
MPEG-4 was initially conceived for small-format video images, like e-mail video, internet video newsclips, and videophones. Pretty much how Quicktime is used today. So the Motion Picture Experts Group decided to use Quicktime as the basis for a lot of what MPEG-4 is/will be. I’m not certain if it has been/is still going to be implemented or not. Some of the links suggested that MPEG-4 was, at last check, going to be expanded to handle transmission of everything from those teeny video clips over 28.8 modems all the way up to DVD-quality streams via T3 or better. All the links I found are a year or two (or more) old.
Oh, and don’t mind sailor, Dave. He just gets gruff every now and then.
Sorry if I sounded that way. I didn’t mean to. I was just trying to point out that he was assuming the efficiency of compression algorithms is independent of the data which is not the case.