If you didn’t know, I have a background in broadcast television.  For many years I worked in R&D for a television studio integrator/VAR and have followed the progress of HD for about 10 years now.  In the broadcast world there are many HD formats: DVCPRO HD, DV100, D5, Uncompressed HD, XDCam, Etc.  These formats are all very expensive and the general consumer or even prosumer could not hope to afford them (a DVCPRO HD camera goes for about $80,000 without a lens). So what will be done about HDTV in the cosumer realm?  A couple of years ago JVC introduced a format they were calling HDV and I hoped it would never catch on.   Unfortunately it did.

So why do I dislike HDV so much?  To explain that, I need to explain how DV works and what a CODEC is and does.  DV video, such as the DV camcorder you would buy at Amazon.com or Best Buy, records a digital video signal on magnetic tape at 29.97 frames per second, with a datarate of 25Mbps.  The CODEC, which stands for COmpressor/DECompressor, is an all i-frame format.  What the heck does all this mean?  It’s actually very simple:  Your camera will record almost 30 frames of video per second to the magnetic tape.  These frames are compressed into a format called DV.  If we didn’t compress the frames, which is the process of taking data out of the picture and thus lowering its quality, we would not be able to fit it on a small tape such as we do.  Compression is both good and bad as it allows us to make the video file smaller, but the more compression we add, the more quality we loose.  The thing about DV is that it uses something called i-frames on every frame.  This means that every single frame is an entire picture rather than a partial picture.  This comes in very handy for editing and error correction.  Lets say you bump your camera and 10 pixels get corrupted.  The chances of you being able to see 10 pixels for 1/30th of 1 second is quite slim, since the next frame is an entire new picture.  That’s DV.

Enter HDV.  Make no mistake; HDV has absolutely nothing to do with DV.  The only similarity is that HDV can be pushed down a firewire cable (that’s the preferred transport, actually).  HDV is a long GOP (GOP is a Group Of Pictures) format based on MPEG 2.  This means that rather than recording full frames for every picture, we record 1 full frame, then the next frame only records the changed pixels, the frame after that is the changed pixels, and this continues until we hit a full frame (AKA i-frame) again.  Bang your camera here and get 10 corrupt pixels, now instead of being on the screen for 1/30th of a second, they could be there several seconds possibly.  This also creates a problem with editing.  Since DV is an all i-frame format, we can make an edit on any frame we like.  Since HDV has virtual frames, it becomes much more complex to edit.  How do you start on a frame that is missing 2/3 of its information?  In addition to the long GOP problem, HDV is 25Mbps, just like DV.  So a standard definition video takes 25Mbps and a high definition video also takes 25Mbps… They are throwing something away to make the HD video fit on that tape (HDV uses the same tape as DV).

These are the disadvantages of HDV:  The fact that it’s not all i-frames (long GOP) and the amount of compression.  So what are the advantages and why would I buy HDV over DV (which I would)?  The advantage of HDV is the type of CODEC and the resolution.  HDV uses MPEG 2, while not the newest CODEC on the market, still a very good one.  Since MPEG 2 is long GOP rather than all i-frames, the virtual frames take a fraction of the space as a real frame.  This means it’s more efficient with the data that it is using, allowing us to put the added resolution on tape.  HDV runs at 720p or 1080i, quite a bit better than the 480i that DV runs at.  720p would be my format of choice, which is 720 lines of resolution displayed progressively.  1080i is a decent format, but keep in mind that if you froze time and looked at your monitor, only 540 lines would actually be showing since it’s interlaced.  So 1080i has better resolution, but all the problems of the interlacing found in DV.

Interlaced video	Progressive Video

There are issues editing HDV, but companies such as Apple and Pinnacle Systems have done a good job of creating the tools needed to edit HDV just like you would any other format.  Due to the fact that this is an MPEG 2 CODEC and your editing software needs to re-create the GOP structure on every edit point, these editing systems are painfully slow.  In general if you have 1 minute of HDV video it will take 4 minutes to export that same video to tape.  1 hour of video would be 4 hours.  This number shifts depending on the number of edits in your timeline, the speed of your processor and the amount of RAM your system has, but it’s a good general number.

The added resolution and general quality of HDV makes it a hands down better format than DV.  My hope was that we were going to come up with a HDDV format that’s HD and all i-frames, but that’s not what happened, so we’re stuck with HDV.  If you’re in the market for a digital camera today, take a good look at the HDV cameras from Sony and JVC.  In the consumer realm there are two options: DV and HDV. You can wait for HDV to come down in price, or you can buy today, but pay a premium for being on the bleeding edge of technology. I personally would go HDV today to get the better picture quality and progressive frames. Which would you do and why?