|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Editions | myCNN | Video | Audio | Headline News Brief | Feedback | ![]() |
![]() |
|||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
Review: Apple's 'Final Cut Pro' for video editing
(IDG) -- What's the big deal about Final Cut Pro? Plenty. Apple's Final Cut Pro (FCP) and its sibling, iMovie 2, are powering a new desktop video revolution that enables people to be content sources, rather than simply media consumers. With FCP, users gain entrance to the elite level of broadcast quality, otherwise reserved for half-million dollar studios. For most uses, any current shipping Macintosh model (except the low-end iMac) will do, although you might want more hard drive space, such as through an external FireWire drive. For high-end uses, Apple is working with other manufacturers on hardware to assist rendering and managing large files. These include the Matrox RTMac board and Pinnacle's Targa Cine for uncompressed SD and HD. What is Final Cut Pro?
Final Cut is a video-editing and -compositing program. Each frame works much like layers in Photoshop, but with a richer variety of file formats. Each layered frame is then organized into a long sequence, with layers containing audio, text (titles and credits) or an image with an alpha channel (transparency mask). FCP 1.2.5 introduced support for YUV color space. To make sense of this, know that Photoshop uses the Lab color space internally, which contains all the colors needed to down-convert to RGB or CMYK. Similarly, YUV is a color space that allows conversion of the video color into all other broadcast or film colors. The human eye is less sensitive to color variations than to intensity variations. YUV separates luminance (Y) and chrominance (UV), which allows encoding of Y information at full bandwidth and UV information at reduced bandwidth. Broadcast qualityFCP is largely independent of the source quality. It's a general rule of video and audio work that you begin with the highest possible quality, because you can save the final output in a lower resolution, as needed. If you use broadcast-quality source material, you'll edit and generate broadcast-quality material. And FCP can handle even higher-quality formats, such as high definition television (HDTV), an emerging standard with 16 different resolutions currently available. Sony has a format called HDcam, with quality slightly higher than the highest HDTV standard. This means you can convert down to any of the lower quality levels. If your target is the Web, you can save progressive or hinted movies of the appropriate sizes. Video is data organized in time. The U.S. television format--NTSC--runs at 29.9 frames per second. Film and the European TV formats--PAL or Secam--run at 24 frames per second. A perfectly good Web frame rate is 12. FCP weaknessHow is FCP's titling ability? In a word, weak. It's possible to create an editable text clip in FCP, but you're only allowed one line of text in a clip. To make multiple editable lines, make and format one, then clone it for the others. 2-Pop, an excellent source of FCP information, has tutorials for creating text in Photoshop and importing it as a clip into FCP. Another approach is the plug-in modules from Boris Graffiti, but both of these are modal, meaning you have to leave the FCP environment. Titling includes title effects. When you get FCP, do not neglect to get your free demo copy of Cinema4D GO. This is one of the finest 3-D text and object animation programs available. It is, incidentally, one of the few programs that support the new multiprocessor Macs. Cinema 4D and its Cinema 4D XL iteration are great 3-D text and object tools. Similarly, don't forget to pick up your entry-level version of Commotion. This rotoscoping and compositing program lets you place moving images in a scene or color a moving object (like a light saber). The pro version has elaborate motion-tracking features.You could win Emmys with these. How does it sound?Is FCP's audio ability respectable? Yes. Any experienced engineer will tell you that quality audio can be much more difficult to capture and work with than video. FCP sound editing has a resolution of 1/100th of a second, precise enough for editing an MTV video "to the beat." An audio engineer could use FCP for audio-only professional work and be perfectly satisfied. It's also possible for audio to be prepared outside of FCP and added back as synchronized clips. Portable editingIt's convenient to note that effects and video filters (dissolves, special effects, etc.) are also resolution independent. One magnificent benefit of this is portable editing. You can take any size source material, save a smaller version, transfer it to a PowerBook for editing, add special effects, etc., show your clients, get approval. Then export the edit decision list (EDL) back to the big format machine, and FCP will chomp away on the source material, and create a high format version, without compromise. A word about camerasThe term "megapixel" is driving the marketing of digital still cameras. Unfortunately, it clouds the discussion about good choices for digital video. For instance, a three-CCD DV camera, such as the Sony TRV900, has a much richer overall color depth than a single chip camera. Most consumer cameras have one video pickup chip. Three-chip cameras have three independent imagers. The primary advantage is greater color depth and sensitivity, and lower noise. Sony even has a pocket-sized three-chip camera (DCR-PC3000) in Japan, which we hope will make it to the states this month. With a three-chip camera, red, green and blue colors are imaged separately, giving 8 bits of depth (from 24-bit) to each of the three video colors. In nature, there are far more than 256 different values of dark and light or hues of color. Traditional film has very impressive latitude, but with three independent eight-bit sources you're more likely to get excellent results. A bonus for still photographers is that each frame of a good three-chip camera prints out a very impressive 8x10 on a competent ink-jet printer. This means that a good three-chip camera can be used as your complete image-gathering solution. Even though modern camcorders have image-stabilization circuits, consider a tripod. Cinema verite (the realistic ever-moving hand held camera) is interesting, but it's generally done with $25,000 steadycam mobile camera mounts. More important, most video compression takes advantage of frame difference. That is, the compression saves image material that's the same from frame to frame and only sends the changes. If the whole frame is moving, the whole frame is different, and thus requires more resources. If you don't like to bring along a tripod, consider putting a 1/4 20 bolt and a wing nut through a spring clamp and clamp it to a nearby chair or post. Few accessories will contribute more to a professional output than a stable camera. Non-compressed digital videoYou'll normally be working with the DV compressed format and rectangular pixels. This is the format used by consumer camcorders. The DV codec has the right combination of file size, throughput and quality. iMovie, for instance, works only with DV source material. As you can see from the discussion so far, formats and aspects ratios can be quite complicated. The miracle of iMovie is its simplicity. FCP, on the other hand, can use many kinds of source material. One of the most interesting and most pure is non-compressed high resolution video. Sony makes a FireWire-based industrial camera, the DFW-VL500. Unlike consumer FireWire camcorders, all aspects of this camera are controlled by software, including white balance and frames per second. If you've used the video source control panel, you'll notice the image quality is blanked out. Not so with these industrial cameras. All controls are available. This particular camera provides resolution of 640 x 480 square pixels. Everything is controlled via the FireWire cable, even zoom and focus. This is similar in capabilities to the large broadcast cameras you see at professional football or baseball games, perhaps even better! Apple doesn't have drivers for these cameras, but TeraGlobal has developed them with the Apple's cooperation. You'll hear much more about this shortly, but know that it takes time for the computer to decode (decompress) the DV stream. If the system can manage a pure digital video stream, it has the advantage of begin with a non-compressed image and quickly convert to other formats for such purposes as real-time collaborative video conferencing. Fun with formats and aspect ratiosNormally the pixels that make up our computer images are square, but in DV format they are rectangular. If you take the same 720 x 480 rectangular pixel DV image and make the pixels square, you get a familiar 640 x 480 image size. The resolution of DV camcorders is 72 dpi. It turns out this equals a quite good 8x10 photograph, out of each frame! The television you're familiar with has a 4:3 aspect ratio. HDTV and the wide-screen format of DVD movies are 16:9. (HDTV format uses square pixels.) The video codec used by consumer camcorders is called DV format, which offers 5:1 compression. There is an intelligent rationale for each of these formats and aspect ratios, and fortunately FCP 1.2.5 can handle all of them. A final note on aspect ratios: Final Cut Pro 1.2.5 introduces anamorphic aspect ratios. This is the ability to squeeze one format into another. Most movies you've seen on television have been "modified to fit on your screen." If you want to see a 16:9 movie on a 4:3 television, you'll see black bands at the top and bottom. The consumer and prosumer formats have another specification known as 4:1:1. Broadcast quality is technically 4:2:2. In practical terms it's a little more difficult to do a chroma key (pure blue or green background is replaced by another image) with 4:1:1. The color is a little less pure, but not by a large margin. Storage, and tricks with FireWireFireWire, IEEE 1394, was invented at Apple, and it has become an industry standard for data routing. Sony, Microsoft and may other developers are doing some remarkable work with it. FireWire's current throughput is 400 Mbps, but is expected to double soon. The cables can be up to four meters long, but can use repeaters or a hub to extend this distance. The six-pin connectors (standard on PowerBooks and current Macs, except non-DV iMacs) can carry up to 5 amps of power. Numerous vendors offer external FireWire hard drives, and a few offer FireWire RAID systems. Portable hard drives spin at 4000 rpm, and 3.5 inch hard drives spin at either 5000 or 7200 rpm. A DV stream requires about 4.5 Mbps of throughput, so either of these is sufficient. For cross-platform users, Unibrain's FireWire firenet networking software ($49) enables you to network Macs and PCs with a FireWire hub (under $100) and FireWire cables. Unlike USB, FireWire is an addressed bus. If you buy a FireWire hub, and you have two computers attached, both computers can access the same camera, at the same time. Since devices on a FireWire bus are independently addressable, it should be possible to have multiple cameras "online' and switch to any source at will. I'm particularly eager to see someone demonstrate this. Additionally, the time-organized QuickTime data format enables the ability to control other devices or other movies across a network. For instance, if you want play a movie simultaneously on many computers at once (each frame at the same exact time) it's possible to have one movie master the playing of others. This is like having multiple projectors, all synchronized. One program dedicated to this purpose is VirtualVTR, which uses a MIDI track and the OMS Opcode MIDI drivers popular with audio engineers. Don't underestimate the little guyAmid all the high-end possibilities, don't forget that iMovie 2 is truly a miracle. It's difficult to make something this excellent and this simple, and most people have success with this software. Its limitations are also its strengths. You will be limited to DV source material, but you can export to any format that QuickTime supports. iMovie 2 lets you separate the audio and video tracks, permitting individual manipulation, such as allowing a narrator to continue speaking while new pictures are shown. There are good new video and audio effects, such as reverse and slow-motion video, and audio reverb to simulate a big room. Titling in iMovie is almost better than in FCP. The last frameMovie editing is not easy. Our culture is impatient, and accustomed to high technical quality and a seamless flow of information. There's a reason why Sesame Street and MTV kept their segments to less than five minutes--this is the longest as they can expect people to pay attention. In short, use excellent source material and edit tightly. RELATED STORIES: Apple reassure staff after stock plummet RELATED IDG.net STORIES: Intel's snappy new video cam RELATED SITES: Final Cut Pro Product Page | |||||||||||||||||||||||||||||||||||||||||||||||||||||
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Back to the top |
© 2001 Cable News Network. All Rights Reserved. Terms under which this service is provided to you. Read our privacy guidelines. |