  |
 |
  |
 |
 |
|
Streaming Magazine Column - November 2002 | |
|
|||||||||||||||||||||||||||||||
|
|
Video Production - How to Avoid the Worst Video Artifacts.We've all seen more than our share of video artifacts. Artifacts are a by-product of the encoding process, which amplifies imperfections in the source video. They come in a variety of flavors - blocking artifacts, stair-steps on diagonals, reduced contrast, and overall fuzziness to name a few. It's therefore critical to maintain the highest quality throughout your video production chain. We'll find out more about encoding in a future column; for now let's talk about capturing, processing and rendering video.Capture at the Highest Possible QualityThe process of transferring video from a camera or VTR (video tape recorder) to a computer is known as digitizing or capturing. This is done by connecting the camera to a video capture card, which converts the video signal to digital information. Video quality can be compromised by cheap capture cards, which don't do a very good job of digitizing video. Use a good quality capture card and capture at the highest possible quality you can, using no compression.Another issue to consider when capturing video is frame size. Video is often digitized at a reduced frame size to save disk space. However, resizing video is a tricky business and not always implemented well on capture cards. A better approach is to capture full screen and resize using high-quality resizing algorithms in your video editing or encoding software. If you must capture a reduced frame size, capture at quarter-screen resolution (320x240 or 352x240, depending on your capture card). The mathematics behind doing a quarter-screen capture are far less complex than an odd frame size, and consequently the capture card will do a better job. Using IEEE 1394 (FireWire or iLink)If your camera has an IEEE 1394 output (commonly known as FireWire or iLink), you can transfer the video digitally using a FireWire card. This method is infinitely preferable because the transfer is lossless - you don't have to worry about the quality of the capture card. Additionally, FireWire includes device control, which allows your editing software to control the camera or VTR.Using device control has another benefit - batch capturing. Batch capturing allows you to identify specific segments of your videotape that you want to capture. Once identified, these segments can be captured automatically. This capability is a huge time saver and a must for any serious streaming video producer. If you can't use FireWire, spend a little extra to get a better quality capture card. The better the capture card, the higher quality video you'll capture, and consequently your encoded streams will contain fewer artifacts. Good quality capture cards are available from ViewCast and Winnov. Broadcast quality video capture cards are available from a number of vendors, including DPS, Matrox, Aurora, and Pinnacle. Processing (Contrast, Brightness and Black Levels)When video is displayed on a computer monitor, it tends to look a bit dark and murky, because television and computer monitors use different display standards. If your video will only be viewed on computer monitors, it's a good idea to boost the brightness of your video to match the brightness of a television monitor. If you're producing for broadcast and the Internet, make sure you add brightness only to the Internet streams, or your broadcast video will appear too bright.Another danger of adding brightness is that black areas of the video end up being more of a charcoal color. To combat this, you must adjust your black level to make sure your blacks stay black. When it comes to contrast, it is often tempting to bump up the contrast to make things look sharper. However, this usually isn't a good idea. Adding contrast makes your picture grainier and artificially sharpens edges. Grainy images and sharp edges are tough to encode. If you're not sure about adding contrast, leave it alone. Render the Highest Possible Quality MastersAfter you've captured, edited and processed your video, you have to render it. Rendering means writing the video, including all the processing and edits, to a new file. Some video editing packages allow you to render directly to a streaming media format. If your editing system does not provide this functionality, you must render a high quality master before you can encode it.Render the highest quality master you can. In general, you want to render uncompressed video, using the frame size you want for your streaming video. The video editing application can take as much time as it needs to process and resize a high-quality master. If you take care to maintain the integrity of your video throughout your production chain, you'll see the results in the resulting streaming media file. Of course, you still have to choose the proper encoding settings - which is the topic of an upcoming column. Next Month: Encoding - Codecs Don't Have Eyes or Ears
|
||||||||||||||||||||||||||||||||||
|
Order Now | Email | Site by Smacktastic/GrDesign | All Content ©2002 Streaming Media Bible.com |