If I do need to edit the video,such as moving video pieces and orders around ,cutting etc.
Then mpeg2 is not good for this?
What would you suggest then.
DV?
Thanx for all your help.I am newbie video man at this point.
Thank you.;-)

If u want to edit the vedio,
u must capture the avi file,
i rememder dazzle snazzi|| have 1394 i/o for DV,
That can do all you want.

Consider DV a working format, and MPEG a delivery format and you'll have good results. MPEG always gets recompressed no matter what you do to it. Not so with DV.

Not entirely so,
DV is a member of the JPEG family and also gets compressed like MPEG.
The difference is that JPEG compresses intraframe (i.e. within one frame) so you still have your 25 frames (PAL/SECAM) or 30 frames (NTSC).
JPEG treats each frame as a picture (i.e. like a photo in photoshop for example.
It uses tools such as DCT scans, witch scans the image in 8 by 8 pixel blocks and trows away the irrelevant and frame resolution reduction, etc...
As it is compessed the full uncompressed frame is compressed by an algorithm and what is left over is stored.
If you want to see or edit it, it has to be decompressed to the full frame again then edited or applied an effect on and the resulting image has to be recompressed.
The difference to MPEG is that MPEG compresses in interframe (i.e. multiple frames at a time) by using GOP'S (Group Of Pictures).
The quality is determined by frame resolution after a DCT scan (how many pixels you trow away in many variants) and how long the GOP is.
A GOP, and this is essebtially the MPEG part of it, consists of an I-frame (the image compessed by DCT and reducing, or not, it's resolution as described above), an P-frame or PREDICTED frame and a B-frame or bidirectionally PREDICTED frame.
B and P frames aren't whole images it's just some low bit rate info of what has changed sinse the last I-frame.
For example a Betacam SX (witch delivers better image quality than DV but can only be edited 2 by 2 frames) consits of an I-B GOP (2 frames long).
Another posibility more used for internet could be I-P-P-P-B-P-P-P-B-P-P-P witch is a 12 frame GOP.
Like you see this can only be edited per 12 frames and is far less in quality 'coz there's only one usable frame, stored in a RAM buffer until the next GOP starts, witch is changed only by the difference info contained in the P and B frames.
App's like Adobe Premiere can edit within a gop by means of recreating a frame with the P and B info, but quality suffers as much data is already lost by the DCT scan (and other MPEG tools used), once you trow away something you can't just recreate it perfectly.
As to "Consider DV a working format, and MPEG a delivery format and you'll have good results. MPEG always gets recompressed no matter what you do to it. Not so with DV", that is true.
Just keep in mind that the more editing or effects you apply to either format, the quality WILL suffer.

I hope this helped,

Cheers,

Dapierdon

As a further clarification, if you have straight-cut DV edits and output to the same DV settings as the source, there will be no quality loss since there is no recompression. Add fx or otherwise change the video and recompression will of course be needed.

With MPEG, we always recompress, even with stright cut mpeg source material.

Is DV the only codec which only recompresses the changes?
I am now rendering my MJPEG captures to HUFFYUV and doing my editing with that. It is supposed to be a 'lossless' codec. Does it get recompressed even in a straight transfer? Thanks.

If your codec is a non-temporally compressed codec (all frame data is in every frame) you should be able to render cuts-only video with no quality loss.

Thanx for all the ongoing info.
Thanx to all.