Wednesday, March 22, 2006

Perfect Digital Copies?

It is not the first time I have commented on this case because of the glaring technical falsehoods perpetrated by the RIAA, things that they proclaim as fact which can be shown not to be so. This is yet another one. In their affidavit "Plaintiffs' Memorandum of Law In Opposition to Motion" they write about "perfect digital copies". I submit that the files in question in this case are not as they state. They may be digital, but they are not perfect copies by any means. They believe their own FUD.

One thing that has puzzled me through all the mess with the RIAA lawsuits is the claim that P2P networks are distributing "perfect digital copies" of copyright works. While I know that such copies are possible using imaging methods like Toast or EZ-CD Creator are possible, most of what the RIAA is claiming to be these perfect copies are really tracks which have been ripped or converted to MP3 file format.

One of the great advances in computer technology has been the ability to make large things smaller. Indeed, it was a necessity in order to make the transfer of data even possible. It is not much problem to send text over the internet or any computer network because the amounts of data is in the 100 or thousands, but not the 10 thousand or million, billion or trillion which files for multimedia would consume.

How do you compress something? First, remove all extra empty space you can, then throw away any data that is not needed. So far we have our data completely intact. This is good for text but we haven't even started on binary type files.

Next is to find a way to remove duplicated data, so instead of printing the letter R one hundred times you might say that R occurs 150 times then provide a map of where those spots are. Its more efficient than you think and the more R's there are the more efficient it gets.

It turns out that the last step is the hardest to accomplish. Data needs to be reconstructed exactly, not approximately. The most popular forms of compression used today do not use this method. The saving in space simply is not enough.

We are stuck with the last method, throwing some data away in hopes that it will not be missed or that it can be reconstructed by "guess" work. There are some great alogrithms out there, but they still throw away data and it sounds like it!

So, in compressing binary files, lossy compression became the norm. You no longer had a perfect copy, you had something that was nearly like the original. Most image file formats are lossilly compressed and that was the way it was with MP3 files. When the MPEG 1 Multimedia formats were being defined, layer three was the audio layer hence the term MP3.

There is no way in which one can reasonably consider an MP3 file a "perfect" copy of a track from a CD. This is Home Computer and Music industry FUD which has duped consumers into believing, for instance, that black is white. It plainly is not. How long has it been since you listened to some real music played live? What you hear with your own ears live is the only bench mark for comparrison, not what some advertising agency, music industry guru or home computer salse campaign tells you.

Read Introduction to audio measurements and terms and hear what a real audio engineer has to say about this. I will agree with Christopher Scott whole heartedly when he says
"In the author's opinion, one cannot speak of high-fidelity when including lossy compression."

No comments: