Shooting video for the Internet

Below are some general tips aimed at producing good video. Most of these tips are focused on creating video that will compress well - by carefully shooting video for internet use, you can substantially improve how well the final video will compress. Better compression results in higher-quality and/or smaller movies.

The overall goal is to produce a video signal with the least amount of noise, camera movement and fine detail possible so that the final movie will compress effectively and look good at a small screen size.

For professional results, it is critical that you shoot tests of your material and run it through your entire production process before you film the whole project. It's important to view the final results on the internet as they will appear in the final project because your image may look great when filmed and edited, but look less than optimal after resizing and compression. Early and thorough testing will help spare you painful and expensive reshoots.

Finally, if you are new to video production, having professionals light and shoot your video may be a good idea. There is a lot of expensive hardware required to do this right, and experience is critical to producing professional results.

Use a High-Quality HD Camera

The higher the quality of the original video signal, the better the final movie will look and compress. A common misconception is that because the final movie will end up small on the screen, a cheap camera won't make a difference - this is absolutely wrong. Video noise substantially degrades compression, so a "clean" video signal produced by a professional camera will compress much better than a "noisy" signal produced by a consumer model.

In addition to lower noise, professional cameras generally produce a sharper image with better colors given their superior optics and multi-chip design. Again, anything that improves the quality of the original video will help you deliver better movies. See if you can rent a HD video camera.

Below is an overview of the four common classes of cameras used for internet video:

• Professional (BetaCamSP, D1, studio/broadcast equipment) The professional formats, such as BetaCamSP, generally produce higher resolution and less noise than the other formats listed here. BetaCamSP also contains adequate color information for bluescreen work. The professional formats are not cheap - unless you are a professional videographer, you'll probably have to rent the equipment. BetaCamSP is an analog format, so you'll need to digitize the signal with a capture card to work with it on the desktop.
• DV (miniDV, DVCPro, DVCam) DV is a high-quality digital format that integrates well with desktop systems. There are currently three DV formats: miniDV, DVCPro, and DVCam. MiniDV is the most common and generally is the format used by consumer cameras. DVCPro and DVCam are professional formats which are not as widely available as miniDV. The DV format is far superior to old formats Hi8, S-VHS and other consumer formats. DV is digital, so it does not suffer from generation loss - a copy of a DV tape is identical to the original. Most miniDV cameras can be connected to your computer via Firewire (IEEE 1394). Some DV cameras offer a "progressive scan" feature. This records each frame as a single non-interlaced image, instead of two separate interlaced fields. Progressive scan source material often doesn't play as smoothly on television as interlaced material but is vastly superior for internet delivery because it contains no interlacing artifacts. You should look for this feature when buying a DV camera and use it when filming for internet delivery.
• HD Cameras come in two resolutions 720p and 1080i which are both good for internet video. Some record to MiniDV tape and some record to a memory card thus skipping a step of capturing from the tape to your computer. The quality is very high.
  
• Consumer Formats (Hi8, S-VHS and VHS) These consumer formats produce substantially noisier signals with lower resolution than the professional and DV formats. Hi8 and S-VHS are superior to VHS. As with BetaCam, these are analog formats and a capture card is required to get these formats into your computer.
• Computer-Based Cameras (video conferencing cameras, etc.) Generally, these cameras produce very noisy and low-resolution images. They often hook up directly to your computer, so a capture card isn't needed. We strongly recommend using a better camera if you are trying to deliver high-quality QuickTime movies.

Lighting for Compression

Generally speaking, video that is well-lit will compress better than under- or over -exposed material. Most codecs work best with moderate contrast material, and many codecs don't work as well with dark scenes. For example, Cinepak normally produces better compression with lighter images.

Adequate lighting is critical to producing superior movies because low-light conditions produce excessively noisy video signals lacking details in the shadows. Overexposure is less frequently a problem but should also be avoided.

You should not shoot video that you know is incorrectly exposed and plan to fix it in post-processing - detail that is missing and excessive noise can never be fully corrected after the fact. Lighting your video properly is the only way to ensure the highest-quality results.

Use a Tripod and Reduce Movement

The use of a tripod often makes a dramatic impact in the quality of the final movie. This is because keeping the camera steady reduces subtle differences between frames and therefore improves the temporal compression of the video.

Be sure to use a sufficiently heavy tripod for your camera. If you plan to pan the camera during filming, use a high-quality fluid head and keep the pan smooth and slow. Irregular or "jerky" camera motion is hard to compress. Avoid hand-held shooting if possible. If you need to film a hand-held shot, a motion stabilizer (Steadi-Cam, gyro, etc.) will improve your results. If your camera has an image-stabilization option (either optical or electronic), you should generally use this feature to reduce subtle changes between frames from camera motion.

NOTE: Some codecs, such as Sorenson Video, are able to detect moderate camera motion and compensate for pans. However, using a tripod and maintaining smooth camera motion with these codecs will usually improve the image quality as compared to hand-held camera work.

Keep Detail to a Minimum

Keeping the detail within the scene to a minimum will help the video compress better spatially. It will also make the video easier to see when the movie is reduced in size for internet delivery.

If you are shooting an interview, keep the background simple. Painted or plain backdrops are often a good choice. If you have the experience and equipment, greenscreen can work very well for interviews.

It is fairly common to film people in front of windows. If there is much detail or movement outside, you can throw the background significantly out of focus to simplify the image. Trees are often used as backdrops for interviews filmed outside. The excessive detail of the leaves poses a challenge for compression and should be avoided if possible. If you must film against a tree, using a shallow depth of field to defocus the leaves will often improve the final movie. Beware of trees moving in a breeze - the high detail and subtle changes between frames make both temporal and spatial compression difficult.

Ask your subjects to wear clothes that don't have high contrast patterns or lots of details. Plain colors are best - bold stripes or checked patterns can do very odd things when resized and compressed.

Producing Good Audio

Audio production values are often overlooked when creating multimedia, but are critical to professional results. As with video, your goal is to produce as high-quality and noise-free an audio signal as possible.

You should use high-quality audio equipment and remote microphones whenever possible to reduce camera noise. You should try to minimize any unnecessary noise in the audio signal such as wind or street sounds (cars, construction, etc.). Shotgun mics may be useful for minimizing background noise and lavaliere mics often work well for interviews.

If you are recording voice-overs in a studio, you should use professional equipment. The mics that come with computers (both Mac and PCs) don't usually produce the audio quality of a real, professional mic. If you are recording directly into a computer, beware of hard drive noise - this is often hard to hear when recording, but will decrease the quality of the final audio signal. Many computers' built-in sound cards introduce line noise, so it is usually better to record directly through your capture card or a digital recorder.

I have a new recording sound booth in my office just for voiceover work.

Green Screen

Properly executed greenscreen can significantly improve your video. For example, if you composite an actor in front of a digital still, the background image will be perfectly steady and noise-free. The lack of video noise and movement in the background improves both temporal and spatial compression of the video, which produces a higher-quality final video.

However, greenscreen work is very technically challenging and shouldn't be attempted unless you have the experience and equipment to do it correctly. Simply shooting an actor in front of a green backdrop often won't work - there are very specific ways you must design your set and lighting to ensure good results. Proper testing is critical, and poorly shot material cannot usually be saved in post-processing.

NOTE: Most of the DV formats (including the pro formats) use 4:1:1 color subsampling. This is fine for most projects, but is not optimal for high-quality greenscreen work. If you need to do greenscreen, you should use a format with higher color resolution, such as Panasonic's 4:2:2 DVCPRO 50 format, or an analog pro format such as BetaCamSP.

<<< BACK TO EXTRANET INDEX