This is predominantly written for those that are just getting into video, and for the people that e-mail me from the various forums because they think that their question is just too stupid to ask in a public space. It isn't, believe me! (But didn't your parents ever teach you that there are no stupid questions, just stupid answers?)

Understanding the Basics
Just what is the difference between DV over 1394 OHCI, video captured using proprietary 1394 cards, and video captured using analog format cards?

Analog
Most of the older format high-resolution capture cards use MJPEG compression, with the video compressed at the card. (We're going to ignore SDI (Serial Digital) cards, as this article is for the masses.) The images are generally 640x480 or 720x486 in size.

Note: An easy way of understanding this, is that analog cards (Media 100, Truevision RTX, Matrox DigiSuite, etc.) use mostly proprietary MJPEG "codecs" (compression/decompression) to take uncompressed video from devices such as video tape recorders like S-VHS, Hi-8 and BetaSP sources and bring this into your computer through the card. All analog compression is done at the computer.

Digital
This is quite different from the DV technology of today. DV is compressed at the camera, on the tape itself. The camera has the DV "codec" built in. This is a revolutionary idea that allows the sharpest picture quality possible with your camera, and to guarantee that same level of quality as your computer's DV "codec" recognizes the format and "Imports" it to your computer without recompressing the data. (We'll look at the importance of the different DV codecs later.) When the video goes through the 1394 cable, into the card, and onto the hard drive, nothing is done to the video. It's merely a channel for transfer of data from tape to hard drive, much like copying a CD-ROM or a floppy disk to your hard drive. This allows the transfer to be done without losing important picture information in the process.

Artifacting
Regardless of the method and format, the less you compress video, the better it appears to the human eye -- but it also requires higher sustained data-rate speeds from your hard drives. At compression levels of less than 6:1, (over 3000 kilobytes per second) most viewers would agree that the images appear as good as the original media, when in truth, it's a demonstrable lower quality. The video will have visible artifacts and smearing. These artifacts become especially noticeable in areas of "hot" reds and oranges (which "bleed") or in areas with subtle gradients, like skies, etc. These subtle gradients will get "banding" rather than the soft, subtle gradients in a real sky.

The DV Spec
In contrast, the DV spec is a 720x480 image size with a 5:1 compression at the camera source. More accurately, it is compressed to a constant throughput of 3,600 kilobytes per second which averages out to 5:1 compression.

Quality in, quality out. What makes DV so great is that the imagery captured on the hard drive is an exact duplicate of the image captured on tape. There is no loss due to the fact that the camera and your computer share the same "codec" and each recognizes the data stream the same as any other computer data -- as a binary stream of ones and zeroes. This is a benefit of the format, not of any specific piece of hardware.

The Importance of the Codec
OHCI cards, 1394 capture cards, i-link' cards, Firewire' cards, they are all the same in terms of input quality. Where various brands come into play, is the codec that they use when recompressing the video during transitions, compositing, titling, and any filtering the user may affect the original video with. Some companies like Canopus, Pinnacle, Matrox and the like, use codecs that are software assisted/hardware based. This provides for faster throughput of recompressed media, allowing several streams of video to play back in "real-time." (Being able to see transitions, compositing, titling, filtering as it happens, without rendering the media.)

The quality of the codec is critical! Each time the video is recompressed, it loses quality of image. Think of it like being on the Starship Enterprise, and every time Captain Kirk is transported, he comes back with a few pixels rearranged. That might get scary after 3-4 trips to Ambrosialand. So, if a transition occurs between the A line and B line of the video timeline, and the section of video is then used as part of an overlay, or a title sequence, the video that appears during the transition has already been compressed twice. A loss in quality has occurred. The significance of the loss is relevant to the quality of codec used to recompress those frames of video.

So, back to 1394. A few years back, the Institute of Electrical and Electronic Engineers (IEEE) got together and determined a specification for transferring digital data. This became known as IEEE-1394. Apple didn't like the numbers and so called it 'Firewire.' Sony didn't like it either, and so they called it "i-Link." And it's all the same thing, so quit arguing with your brother-in-law about whether your Apple FireWire is faster than his Sony i-Link, and that both of you are faster than your cousin's OHCI. They are the same. They all include the same specifications and controller capability as any 1394 interface device.

Well, sort of...

Various card, software, and camera manufacturers have implemented -- and in some cases "interpreted" -- the architecture of some of the specifications. This may mean that your camera may not be seamlessly controlled by some of the cards out there.

Next Page: OHCI -- USB On Steroids!