why HDV to DV? at DVinfo.net

Juni Zhao · July 2nd, 2007, 08:55 PM

alot posts here are about how to downsize HDV to DV. The only reason to me would be to preserve the HDV footage for future use. Other than that, why don't we shoot DV rather than HDV then DV? Is downsized DV quality better than original DV?

David Rice · July 2nd, 2007, 09:08 PM

Is downsized DV quality better than original DV?

Yes.

Robert Morane · July 3rd, 2007, 03:23 PM

Why? and what about the "long Gop' drama?

Joe Busch · July 3rd, 2007, 03:32 PM

long GOP has to do with the compression in HDV, after you convert it to DV I don't think GOP applys anymore... once you render it to a new DV Mpeg file...

I don't see any reason not to shoot HDV over DV honestly... atleast no reason that would make any significant impact on the final product...

Robert Morane · July 3rd, 2007, 05:54 PM

If long gop create problem when you capture image, this problem should still exist even if you downconvert?

Dick Nelson · July 3rd, 2007, 06:47 PM

HDV uses MPEG-2 to record more pixels per frame, but at the cost of not really recording every single pixel for every single frame (Long GOP).

DV uses a different compression technique, DCT (Discrete Cosine Transform), recording fewer pixels overall but recording ALL of them for every frame (sort of). The compression for each frame does not involve prior or past frames; you get a whole frame each time.

When you convert an HDV recording to DV, the software recreates the 'missing' frames of MPEG 2 and converts them one at a time into the DCT- based DV format. It has to do similar magic with the audio, since that is recorded differently as well.

If you shoot 16:9 HDV and use only the center part of your frame when you convert to DV, you throw out a lot of pixels in the end result, but who cares? The results are fantastic, and often better than if you had used the DV format for the original recording. Plus, you now have extra image area that can be used to correct for unplanned defects in the original footage.

For example, the boom mic dipped into the shot. No problem, just move that 'larger than frame' HDV image up a litle and it's gone. Caught a gawker or light or C-stand in frame? Bump it left or right and it's gone. Great for stabilizing shaky shots too without softening the image by blowing it up.

So, downconverted HDV looks good, doesn't suffer from Long GOP problems, gives you a high def image for the future, a bigger image frame to use for problem solving in standard def today. The only thing you lose is some time to do the conversion, and that's what coffee breaks were made for, eh?

If you want to learn more about compression, adamwilt.com is a great resource.

July 2nd, 2007, 08:55 PM	#1
Juni Zhao Regular Crew Join Date: May 2007 Location: North Andover, Ma Posts: 102	why HDV to DV? alot posts here are about how to downsize HDV to DV. The only reason to me would be to preserve the HDV footage for future use. Other than that, why don't we shoot DV rather than HDV then DV? Is downsized DV quality better than original DV?

July 2nd, 2007, 09:08 PM	#2
David Rice Major Player Join Date: Mar 2007 Location: Sitka Alaska Posts: 470	Is downsized DV quality better than original DV? Is downsized DV quality better than original DV? Yes.

July 3rd, 2007, 03:23 PM	#3
Robert Morane Major Player Join Date: Apr 2007 Location: Tucson, AZ Posts: 267	Why? and what about the "long Gop' drama?

July 3rd, 2007, 03:32 PM	#4
Joe Busch Major Player Join Date: Feb 2007 Location: Sammamish, WA Posts: 398	long GOP has to do with the compression in HDV, after you convert it to DV I don't think GOP applys anymore... once you render it to a new DV Mpeg file... I don't see any reason not to shoot HDV over DV honestly... atleast no reason that would make any significant impact on the final product...

July 3rd, 2007, 05:54 PM	#5
Robert Morane Major Player Join Date: Apr 2007 Location: Tucson, AZ Posts: 267	If long gop create problem when you capture image, this problem should still exist even if you downconvert?

July 3rd, 2007, 06:47 PM	#6
Dick Nelson Regular Crew Join Date: Aug 2006 Location: Silverthorne, CO Posts: 36	HDV uses MPEG-2 to record more pixels per frame, but at the cost of not really recording every single pixel for every single frame (Long GOP). DV uses a different compression technique, DCT (Discrete Cosine Transform), recording fewer pixels overall but recording ALL of them for every frame (sort of). The compression for each frame does not involve prior or past frames; you get a whole frame each time. When you convert an HDV recording to DV, the software recreates the 'missing' frames of MPEG 2 and converts them one at a time into the DCT- based DV format. It has to do similar magic with the audio, since that is recorded differently as well. If you shoot 16:9 HDV and use only the center part of your frame when you convert to DV, you throw out a lot of pixels in the end result, but who cares? The results are fantastic, and often better than if you had used the DV format for the original recording. Plus, you now have extra image area that can be used to correct for unplanned defects in the original footage. For example, the boom mic dipped into the shot. No problem, just move that 'larger than frame' HDV image up a litle and it's gone. Caught a gawker or light or C-stand in frame? Bump it left or right and it's gone. Great for stabilizing shaky shots too without softening the image by blowing it up. So, downconverted HDV looks good, doesn't suffer from Long GOP problems, gives you a high def image for the future, a bigger image frame to use for problem solving in standard def today. The only thing you lose is some time to do the conversion, and that's what coffee breaks were made for, eh? If you want to learn more about compression, adamwilt.com is a great resource.