DV is "lossless" ??? - this sounds imp
Someone please help satisfy my curiosity.
OK, I understand digital "ones and zeros". For example you have a Mini DV tape and you make a copy of it to another Mini DV tape. You have created an exact bit for bit copy. That is DV = DV
Now can someone explain to me how DV captured to a computer that gets COMPRESSED 5:1 maintains 100% integrity. It sounds impossible. If 80% of the data is being lost in the process of capture process, how can the captured footage be exactly what is on the tape? In my way of thinking that is more like DV = DV/5.
Capturing DV to computer is the same as copying to another DV tape. Using capture in this sense is a hang over from the days of analogue capture to the computer. For DV material it is actually a copy to computer. It is nominally exactly the same. The 5 to 1 you are refering to is when the original analogue signal was compressed to DV in the camera in the first place. Once in DV form it is only copied unless after editing you wish to convert to another form like mpeg2 for DVD creation.
You're transcoding the video via an analog/s-video source into the computer so you're already losing some quality in the transfer. Secondly, depending on the device you use for your source, you could use a poor quality DVD player or such and lose further quality via the interpolated video from the source. So you've already entirely change the video, and it's never LOSSLESS.
So you answered your own question. It's not possible. However, it's possible to perfectly capture the already interpolated video through the analog/s-vid INs, but it's already not the same video. The DV500 DVD doesn't record LOSSLESS video but what is captured looks very clean to the naked untrained eye. The output back to DVD isn't the same as the source, but again, it's close to the untrained eye. It's close enough where you could copy a full DVD and give it to a friend, and still have them come back and tell you it looked awesome. Really not noticable.
But there is no such thing as LOSSLESS capture. There's only the bit by bit transfer of modern media formats as you say, which is of course, lossless.
DV is not lossless. Nope no way no how. As soon as you compress it that is out the door. It is like saving a photoshop file to JPG then copying and copying it. Eventually you will notice the artifacts in doing so. JPEG guesses. DV guesses. Very good guesses and acceptably so at the DV level. Now it only manages to copy from file to file IF the capture/edit codec is the same as on the tape. Panasonic's pro DV tape line (DVCPRO25)will import into their own edit boxes at 4X transfer time, basically a file to file transfer. Sony, I understand maybe working on this and I am not sure how the individual camera makers fair with the Microsoft DV codec but any thing in the DV500 (which I have) is recompressed and is not lossless. Lossless is a buzzword which is thrown too casually away. You pay out the nose for that level of quality. Thanks to FCP and some serious add on boards like Pinnacle's Cinewave that is coming down to a reasonable level. (think 20k and not 200k).
"Living in 1D, Working in 2D, Dreaming of 3D"
-Me, March, 1998
An example of a "lossless" data compression algorithm is the LZW code used in winzip and the unix 'zcat' command. These have to be lossless or we could not use programs like winzip to package up software, right?
I think you are a little confused in thinking that the only way to reduce the size of a file is to throw away data. Redundent data patterns can be reduced down to one instance and other repeating occurances in the data stream can be reproduced perfectly. An example is the I-Pand B frames in MPEG streams. The P and B frames only contain the difference between them and the seed I-frame.
"Lossy" data compression algorithms like MJPEG work sorta like you describe. Each time thru the compression loop is sorta like pushing around a pile of sand with a broom. You pickup some dirt and lose alittle bit of sand each time you move the pile.
I'm not sure DV is a true lossless compression format but a 5 to 1 compression ratio isn't very radical. You can hammer a 30 MByte raw RGB AVI down to 300 Kbytes with MpEG4/DiVX/264. Yeah, it looks crappy at 100 to 1 compression ratio but you can still see pretty much what was in the original content. If you have low motion video with talking heads and no MTV style edits, getting a video compressed down 5 to 1 is a no-brainer.