Garbage In, Garbage Out: Knowing What Affects a Video s Quality

Garbage In, Garbage Out: Knowing What Affects a Video's Quality

You may have heard this phrase before, which means that you can't get something from nothing. Ever tried making a soup with bad ingredients, thinking it would still taste good? The same principle holds true for video production. Four primary factors influence the quality of your video footage: source format, image quality, sound quality, and subject matter.

Source Format

The source format is the container in which the video is stored, whether it's a digital recording encoded on miniDV or DVCAM tape, an analog recording on VHS or Hi8 tape, or a direct MPEG file recorded on your digital camera. Each recording medium has inherent resolution limitations — some formats can store more information than others. The more information the medium stores, the higher the quality of the recording.

The following list outlines the resolution capacities of common recording mediums. For the purposes of our discussion, this list is restricted to video formats and excludes film formats, such as 35mm or 16mm motion picture film. The video formats are compared by line resolution, which is the number of horizontal scan lines the format can capture. Line resolution is not the definitive attribute for video quality, but it does measure the capacity of potential visual information captured by the camera or recording device. The most important factor to remember when comparing line resolutions is that video is primarily targeted at television screens that, in North America, display 525 lines of resolution.

Note

Practically speaking, television sets display a maximum of 483 lines of visible resolution — the remaining lines convey other information such as the sync pulse. Although High Definition TV (HDTV) is capable of displaying 1,080 lines of resolution, most multimedia producers will not be using HD cameras to record video.

VHS, VHS-C, or 8mm videotape: These tape formats, especially VHS, are the most common video formats in use by consumers today. The average household VCR records video in VHS format. These tape formats can record about 240 lines of resolution. This resolution is less than half of the potential scan lines that can be shown on a TV screen, which is why the VHS recordings of your favorite TV shows don't exactly match the live broadcasts. These tape formats are analog in nature — the signal they record is not digital and can deteriorate from copy to copy. To translate this tape format into digital video for use in a multimedia production, you need to capture the footage with an analog video capture card.
S-VHS or Hi8 videotape: S-VHS and Hi8 video use the same tape sizes as their VHS and 8mm equivalents, but they capture up to 400 lines of resolution. This resolution capacity comes very close to the 525 lines of resolution that a television can display. You probably won't encounter many S-VHS camcorders anymore, but Hi8 video camcorders are still very popular and in wide use today. While the video quality is a noticeable improvement, the video signal is still analog, so as with VHS, you must capture it with an analog video capture card.
miniDV, Digital8, or DVCAM tape: The new breed of video recording devices for consumer, prosumer, and professional use is the DV (Digital Video) format. DV formats use the DV codec to digitally record video, while storing audio at near-CD or CD-quality sampling rates. The native resolution for the DV format is 525 lines, but the actual resolution a DV camcorder will record varies from model to model. Most high-end DV camcorders are considered broadcast quality, meaning the video image is good enough to use on a television show such as the evening news. Many computer systems today ship with an IEEE 1394 (also known as FireWire or iLink) port to capture digital video from DV recording devices. When video is transferred digitally from a camera to a computer over an IEEE 1394 connection, there is no loss of quality.
Betacam SP, Digital Betacam: These tape formats are for serious video professionals who work in the television industry. Betacam SP has been the industry standard for network television for more than 30 years. While Betacam SP is an analog format, Digital Betacam (also known as DigiBeta) records video with a proprietary Sony codec. Both formats capture 550 or more lines of resolution.
MPEG file recording: Many digital cameras can also record MPEG video at various sizes and resolutions. Most digital cameras use the MPEG-1 codec, which was used in the original Video CD (or VCD) format. Sony recently introduced a new Micro MV format, which records video in the MPEG-2 format used by DVD Video discs, at similar line resolutions to DVD Video (525 lines). Most MPEG recording devices capture at line resolutions equal to or less than VHS quality.

As a general guideline, we recommend using a miniDV camcorder to record video that you intend to use for multimedia presentations within a Flash movie. Several factors beyond the recording format affect the quality of any recording. In the next two sections, we discuss variables that influence the visual and audio portions of a recording.

Web Resource

If you're a beginner videographer and want to know more about video resolution, check out the following links:

www.bealecorner.com/trv900/respat/
hometown.aol.com/ajaynejr/vidres.htm

Image Quality

Regardless of the source format that you use for your video recording, the recording device may have other limitations. The recorded resolution of your camera may be significantly lower than the upper limit of your recording format's line resolution. For example, the DV format is capable of storing 525 lines of resolution. However, your specific DV camcorder may have an effective resolution of 490 lines. The following variables can affect the quality of the picture recorded by the camera. Note that this list is by no means exhaustive for the professional videographer — we chose these topics as a general summary to guide you in your video production.

Optics: The characteristics of the lens (or lens system) that your camera uses are known as the optics. The optical quality of video camcorders varies wildly. Many camcorders boast Carl Zeiss lenses, which are known for their precision and accuracy. Some lenses are constructed of plastic, whereas others are glass. Neither lens type is necessarily better than the other, but the only way to accurately judge the optical system of your camera is to conduct extensive testing with a resolution target.
CCD: The CCD, or charged-coupled device, is the electronic plate that captures the image being rendered by the camera's lens. You can think of the CCD as the digital film that's being exposed to the light coming through the lens. CCDs come in a variety of sizes, ranging from ¼ inch to 1 inch. Although these sizes may sound incredibly small, even ¼-inch CCDs can capture amazing detail. The larger a CCD is, however, the more information it can capture. Most cameras only have one CCD, but some cameras have three CCDs — one for each color of the RGB (red, green, and blue) spectrum. These are known as 3-chip or 3-CCD cameras, and they have better color fidelity than single-chip CCD cameras. Color fidelity, simply put, is the measure of how closely the recorded color of an image matches the original colors of the subject.

Tape quality: The quality of the tape on which you record the video can also affect the quality of the overall image. Each tape brand has unique characteristics, such as the type of metal particles that are magnetized onto the tape. Some brands (or stocks) are more durable and can withstand the rigors of repeated recordings. In general, you should always record on a fresh tape — and preferably one listed as premium or high quality.

Tip

Make sure you use the same brand and model of tape consistently throughout a shoot. There are slight color variations and differences in quality in each brand and tape type. Preferably, all of the tapes used for a shoot should be manufactured from the same batch, and as such, you should buy bulk boxes of your preferred tape brands and formats.

Shutter mechanism: The shutter mechanism is the device that controls how quickly the CCD samples the image rendered by the lens. Most camcorders have shutters that record interlaced video, which records two interwoven fields (or separate halves of a picture) to create one frame of video. Have you ever noticed how the image on your TV flickers when you pause a VHS recording? You're seeing the two fields of the frame alternating. Computer monitors do not display interlaced video. Rather, they use progressive scanning to minimize a flicker effect. Some higher-end camcorders have progressive scan shutters that record the image with higher apparent resolution. We discuss de-interlacing video later in this appendix.
Exposure: You should make every effort to shoot your video footage with the correct exposure. Exposure refers to the amount of light captured by the CCD and the shutter mechanism. Some camcorders do not allow you to control the exposure — it's all automatic. The biggest pitfall for most videographers is underexposing video footage or shooting in low light. When you try to record a poorly lit subject, you tend to get noise in the darkest areas of the image. Video noise is a random pattern of dots that shows up on the image. Finally, make every effort to properly white balance your shot to suit the color temperature of the dominant light source — some camcorders enable you to control the color temperature. You'll notice the effects of color temperature in the video samples on the CD-ROM. You learn more about white balance later in this appendix.

So what do all these variables boil down to? In a nutshell, we recommend that whenever possible, you should shoot video with a camcorder that has a superior optical system with three CCDs, use high-quality tapes, and properly control your exposure. Avoid shooting in low light, unless you are shooting with a particular effect in mind, such as infrared lighting.

Sound Quality

Every videographer should consider how audio is recorded during production. Most video camcorders have a decent built-in stereo microphone, but professional videographers equip themselves with accessories beyond those that ship with the camera. Review the following guidelines for capturing sound with your video recording:

External microphones: To achieve the best audio recording, put an external microphone as close as possible to the source of the sound. If you want to record a person talking, an external microphone, such as a wireless Lavaliere mic that's pinned to the person's shirt, collar, or tie, will produce a much cleaner recording than the microphone on the camera.
Balanced vs. unbalanced audio: Most microphones you will find at electronics stores use a stereo or mono 3.5mm mini-adapter that plugs into the microphone jack on your camcorder. These microphones are considered "unbalanced" audio sources, due to the nature of their internal wiring. For many video shoots, this may not pose a problem. However, professional audio engineers use balanced audio for all sound sources. The cabling for balanced audio tends to be a heavier gauge than that of the consumer variety. Many balanced microphones have a 3-pin XLR connector, and most camcorders require an XLR adapter to connect these sources to the mini microphone jack. Many professional video cameras have built-in XLR jacks.
Sampling rate and bit-depth: Unless you're using a DV format camcorder, it's likely that you will have little or no control over the specific sampling rate and bit-depth used to record audio on the camera. DV camcorders enable you to record with either 32 kHz or 48 kHz. 32 kHz audio is recorded at 12-bit, which is suitable for recording dialog and live action. 48 kHz is recorded at 16-bit, which is suitable for recording live musical per formances or any other scene requiring high fidelity.
Audio levels: One of the most overlooked aspects of video production is monitoring the audio levels of your source while recording. Most camcorders only record audio with AGC, or Automatic Gain Control. This "feature" allows the camcorder to determine how much an audio signal should be boosted or minimized during recording. Professional audio engineers prefer to manually monitor and adjust sound levels while the recording is taking place. Undesirable side effects of using AGC include amplified background noise during silent periods, audio distortion, and sudden jumps or drops in audio levels. Whenever possible, listen to your audio source through headphones connected to the camera.
Unwanted noise: Do your best to minimize any background noise when you are recording. The microphone(s) connected to your camera will pick up more noise than you may realize. Each microphone has a different "pick-up" pattern, determining the range and direction of sound that it will receive.

In summary, you should record audio as close as possible to the source, using balanced microphones and monitoring the audio feed with headphones connected to the camera. For most video recording, either the 32 kHz or 48 kHz sampling rates will yield superior audio reproduction.

Subject Matter

Last, but by no means least, the type of subject matter you are shooting can influence the results of video compression in Flash 8. When it comes to video compression, the most important factor to remember is variability from frame to frame. That is, how much and how often does your subject matter change? Are you recording the Indy 500, panning racecars accelerating at incredibly fast speeds? Or, are you recording a time lapse of a flower slowly blooming? In general, you will achieve the best results with video compression if the subject matter does not move randomly or quickly.

Here are some general guidelines when choosing and shooting your subject matter:

Use a tripod: One of the most common mistakes amateur videographers make is to handhold the video camcorder. Unless you need to freely move with your camera or intentionally want to introduce shakiness to your video image, always mount your camera on a tripod.
Avoid quick and jerky movements: If you need to follow your subject matter by panning the head on the tripod, practice the movement a few times before recording. Try to be as slow and fluid as possible. The more quickly you pan a video camera while recording, the more likely you'll see compression artifacts show up in your final video in the Flash movie.
Avoid zooming: While it may be tempting, do not zoom the lens while recording your subject matter. It's better to use the zoom to frame the shot before you record. Of course, you may intentionally want to using wild zooming effects to re-create the next Blair Witch mockumentary, but be aware that any rapid movement from zooming the lens will likely compress very poorly in the Flash movie.
Lock focus: All camcorders can auto-focus the lens. If you plan to record a stationary object, you may want to switch to manual focus. If your subject matter moves away from the focus "spot" used by the camera, it may refocus on areas behind the subject matter. This type of focus drifting may yield very unpleasant compression artifacts in the Flash video.
Watch white balance: White balance refers to how the camera interprets color. You might notice that your skin looks much less appealing under fluorescent light than it does under an incandescent or soft white indoor light bulb. The human eye can "correct" the perception of a light source's color (or color temperature) with greater ease than an automatic setting on your camera. Be sure to match the white balance setting on your camcorder to the predominant light source in your video composition. Most cameras have at least two white balance settings: indoor (for tungsten light) and out door (for daylight). Some cameras enable you to perform a custom white balance. To set a custom white balance, focus the entire viewing area on a solid field of white and engage the white balance lock on your camera.

Another factor to consider is the area of the video composition that changes from frame to frame. In a previous example, we mentioned panning a racecar. In that example, the entire picture changes in every frame. Compression works best on video footage with the least amount of movement per pixel in the frame. For example, if you mount your camera on a tripod and keep the camera motionless, you can record the motion of a single subject, such as the flight of a bee between flowers. In such an example, the video compression with the Sorenson Spark codec is much more effective than it would be in the example of the racecar.

Comparing Video Sources

On the Macromedia Flash 8 Bible CD-ROM, you will find three video samples in the appD folder. You can use these examples to test compression settings in Flash 8, the Flash 8 Video Encoder, or a third-party utility such as Sorenson Squeeze or Wildform Flix. Each sample was recorded with a different camera. Refer to the color insert for samples of these videos.

sample_low.mpg: This video file was recorded with a Sony DSC-S70 Cybershot camera. While this camera is designed to shoot stills, it can also record MPEG-1 video files. This file was recorded in high-quality mode at 320 x 240, 15 frames per second. The microphone on this camera is mono (single-channel).
sample_mid.avi: This video file was recorded with a Sony DCR-PC100 miniDV cam corder. This camcorder has a stereo built-in microphone and a single CCD. The quality captured by this camera is average for most DV camcorders. Compare the color temperature of this recording to that of sample_low.mpg and sample_high.avi files.
sample_high.avi: This video file was recorded with a Sony DSR-PD150 mini-DVCAM camcorder. This professional camera has superior exposure and color temperature control. Among other differences with the previous samples, the audio was recorded with a wireless Shure microphone and an UP4 receiver, which uses a 3-pin XLR output. The PD150 can accept this XLR output.

When you open these files in a video viewer such as Windows Media Player or Apple QuickTime Player, notice that the .avi files recorded by the DV camcorders appear stretched horizontally. The DV format uses nonsquare pixels, whereas most display devices, such as computer monitors, use square pixels. Unless the video display compensates for nonsquare pixels, the video image will appear stretched. Flash 8's video import process will not automatically compensate for non square pixel footage, so make sure you resize the video in the Advanced properties, which we discuss in more detail in Chapter 17, "Displaying Video."

We provided these samples so that you may experiment with them in Flash 8. You will find that cleaner video (such as that shot with the high-end video camcorder) will exhibit less video noise in the compressed version within the Flash 8 document.

Tip

While we mention these general rules for better looking video shoots, it is also crucial to develop a look and feel for whatever you are shooting. Establish a set of rules that apply to everything in the shoot. For example, perhaps you want to shoot everything handheld, rack the focus regularly, underexpose the image slightly, dutch the angle on all wide shots, and so on. With today's cutting-edge video and film effects, it is more important to develop a style and operate to achieve a consistent look than to worry excessively over conventional dos and don'ts.