Excuses for being a little stupid with regards to compression, can someone help to enlighten me.
I have a DC30+ which will compress down to 3:1 (if my system's good enough),
I have just shelled out for a Sony DRV900E in my first move to go completely digital. After reading the frightening events on this board regarding the current crop of digital capture cards. I think I'll sit on my hands and wait a while before diving in the mirky pool, the Pinnacle DV200 may be better? So I intend to carry on using the DC30+ to capture from the Sony's analogue out sockets.
My confusion? I understand the Sony camera compresses video during recording at 5:1. So when I output from the camera through the analogue ports, I'm outputting "already compressed footage at 5:1" yes? My DC30+ wants to compress this again at - lets say for example 4:1.- Does this mean I compress down even further, footage that's already compressed to 5:1?. If it doesn't, what would be the best compression rate to set the DC30+ at? 5:1 perhaps? I'm confused!
If I output from the DC30+, back to the Sony. onto digital tape via the Sony analogue inputs, how noticable would the quality drop be, compared to using a digital capture card? - Barely perceptable? - Perceptable? - Strikingly perceptable ?
Thanks in advance for anyone taking the time out to reply
Ken W
Hopefully I can explain things a bit because there is a simple answer and the more complex one.
DV does infact compress approximately 5:1. I say approximately because the compression system is designed for a constant data rate of 25mbits/second and so the compression can in fact vary. BUT for arguments sake and in practice 5:1 is pretty good assumption.
This means that 80% of your original data is discarded, and the sophisticated compression system on decompress recreates the image for you. This recreated image (decompressed) is what is fed out of your analog outputs.
SO it is not a compressed image that you are feeding into the DC30 but in fact an uncompressed image but one that has been reconstructed from 20% of the data.
DV uses a specific codec which allows the picture to be broken up into a number of blocks and each block is individually compressed. The system is very efficient and very good in quality.
The DC30 uses an older compression system M-jpeg which forces the entire image to be treated as a single image to be compressed.
Here in lies the potential problem>
DV does its compression throwing away 80 percent of the image.
The decompression system then uses the 20 percent to create a decompressed image.
This is transfered to you DC30. The DC30 at 4:1 throws away 75% of the Image. BUT IT MAY NOT BE THE SAME PORTIONS THAT THE DV CODEC WOULD CHOOSE SINCE THIS CODEC USES A DIFFERENT SYSTEM AND HAS TO WORK ON THE IMAGE AS A WHOLE UNIT.
Now you do your edit etc. IF you only write back to ANALOG then you probably will not see these issues BUT if you write back to DV and then DV has to once again take the image
that was originally reduced by 80% then regenerated, then reduced by 75% using a different system, then regenerated and then Reduced again by 80% for the final output to DV and then regenerated for the last time to be copied to ANALOG or Viewed you will usually see degredation.
This is called concatonation of multiple codecs. USUALLY what is done is that the M-Jpeg system needs to be compressing at only 2 to 1 and in this mode the compression is almost lossless. Even in this case there can be problems but it is minimized.
SO:
Short answer, IF you only go from DV to DC30 and then out to Analog, you probably will not see any problems.
IF you add a step back to DV and then to Analog you probably will unless you compress at a level that the DC30 does not support. IT is this 2:1 Mjpeg level that Avid DIgiSUite etc can work at that minimizes these problems.
As an aside this whole issue of multiple codecs has caused many of the Digital Sat and digital terrestrial suppliers to now recommend that the programs they use NOT be COMPRESSED during editing.
John Ferrick