Editing Software is Easy as AB

Magazine: Government Video
September 1999
By Sheldon Liebman

As computer-based tools continue to evolve, it’s often helpful to evaluate them in the context of the process they replace. For example, comparing a paint package with the process of taking brush to canvas instead of comparing it with other paint programs. This month, we’re going to look at software packages designed for non-linear editing compared with the process of using a traditional A/B Roll editing system.

Getting to the Source
When you set up an A/B Roll system, it isn’t a requirement that you shot the footage yourself. You just need to have a system capable of playing it back. Obviously, you can’t play a Betacam tape on a U-Matic machine, but if you get the source in a format you can use, then you can work with it. For the purpose of this story, we’re going to assume that you can get source material as digital files in a popular format such as an Apple QuickTime file or a Microsoft AVI file. We’re going to further assume that the ultimate deliverable you want to create is a similar digital file.

By making these assumptions, we don’t have to worry about capturing video onto a computer or creating a tape once our project is completed. If you need to start with tape or end up on tape, there are two options. The first is to purchase a hardware/software NLE system from a company like Accom, Avid or Media 100. These turnkey systems include the components necessary to import and export videotaped material. You can also build your own system by purchasing products from companies like Fast, Matrox or Pinnacle, just to name a few.

Ignoring the input/output issue, we can concentrate on the software only solutions. Packages in this area include Adobe Premiere, Apple’s Final Cut, EditDV from Digital Origin, in-sync’s Speed Razor and Ulead MediaStudioPro. These products just keep adding features and the latest versions were on display at the recent SIGGRAPH trade show in Los Angeles. One of the reasons they keep getting better so quickly is that most now support the use of “plug-in” technology. With plug-ins, you can add new capabilities without needing to purchase another version of the software. Dozens of companies exist today just to supply these plug-ins. Traditional editing systems just don’t have anything to compete with that.

The Editing Console
In a standard A/B Roll setup, the input and output machines are connected to an editing console, which is also connected to a video monitor. The console is designed to control all your machines and to let you specify the locations and order of the material you wish to put onto the finished tape. In a software-based editing environment, the computer, keyboard, and mouse replace the editing console and the computer display replaces the video monitor.

Work It
Now that we’ve looked at the physical setup, we can delve into the actual process of working with the source material and creating the program. This is the point where software-based editing and traditional editing go in completely different directions. The differences are based on two very unique aspects of computer-based editing compared to traditional methods.

The first is that you can have all the material available to you all the time. Instead of having a box (or shelf) full of videotapes that need to be shuffled around constantly, non-linear editing systems can utilize as many files as you need to access the source material. If you need material from Tape #1 in the beginning and the end of the project, you don’t have to load it twice.

The second big difference is that using editing software doesn’t limit you to two tracks of video. When this type of software was first introduced, it closely followed the concept of using only two “layers” of video. Today, most editing software allows dozens, or even hundreds of tracks of video that can be used for editing as well as compositing.

Since it’s all done in software, you can create a project today that uses two layers and a project tomorrow with ten. No additional hardware is required and you don’t have to change any wires. You also have access to more than just “cuts-only” editing. All non-linear editing software packages provide dozens of transitions and effects that can be used for simple effects like wipes and dissolves to complex effects like page turns, mosaics and more.

Bin There, Spun That
The key to using NLE software effectively is the concept of bins. In most systems, virtually everything is organized into bins. Usually, bins are displayed on the upper left side of the computer screen and can provide a lot of information. If you click on a bin that has video, it can tell you the name, date and size of the file, but it might also offer a description of the contents, the video format it’s in and the starting time code.

Bins are also used to show libraries of transition effects. Many packages actually provide very small previews of the effects when this bin is open using the letters “A” and “B” to easily convey how the transition is applied to the foreground and background video channels. Since all of the previews typically have additional settings, this information can usually be accessed and adjusted by clicking on the effect. For example, you might specify how quickly a page turn occurs or how many times the video spins as it travels in or out of the frame.

When you want to add a video clip or an effect to your project, you can usually just select it with the mouse and drag it down to the spot on your timeline in which you want it to occur.

Timing is Everything
The timeline mentioned above is another key area of editing software and virtually all packages place this in the lower half of the display screen. Early packages had five “channels” in the timeline representing Video A, Video B, Transition, Audio A and Audio B.

Time starts at 0:00:00 at the left edge of the screen and progresses toward the right. In most packages, you can zoom in or out to adjust exactly how much time is displayed across your screen. This is crucial when it’s necessary to adjust the timing of events to a single frame.

Audio and video items are added to the timeline by selecting them from the bins and dragging them onto one of the available channels. When an item is positioned in this way, it’s automatically set up to start at the beginning and continue for its entire length. Placing another item on the same channel creates a cut from the end of one to the start of the next. When items are placed on different channels, transitions can be specified to occur between them, usually corresponding to the time period in which they overlap. As with video and audio, transition effects are usually selected from a bin and dragged onto the transition channel.

The Ins and Outs of Software Editing
If you only wanted to use complete clips, this might be enough to finish your project. However, it’s rarely that easy. Instead, you need the ability to set your in and out points to the exact frame you need. With editing software, this can usually be done either visually or from the keyboard. Visually, you can move your mouse to the start or end of a clip and just drag it in or out. Numerically, you can bring up the details for a clip and type in the time code for the in and out points to whatever you need.

In some packages, you may even find an option to compress or expand the time of a clip so that you can pick the footage you want and fit it exactly into the time frame you need. You can also slide clips around so that they occur earlier or later in your project.

Looking Good
As we’ve been describing the layout of editing software screens, we’ve covered the bins and the timeline, but we haven’t mentioned anything about the video and audio themselves. That’s usually located in upper right hand portion of the screen. This area can be used to play a low-resolution version of the video project as it stands, either in full or in part. Playing part of the project simply involves setting in and out points for the preview (instead of in and out points for specific clips).

When it’s time to actually generate the final video, all of these packages have a “Record” or “Publish” function that automatically creates every frame of the project, video and audio, with appropriate transition effects, and saves it to one of the digital file formats supported by the product. At that point, you can just stick it on a disk and deliver it to your client.

But Wait, There’s More
One of the nice things about using software is that it can offer more capabilities than traditional hardware editing systems. For instance, many of these packages have expanded beyond just editing and provide compositing functions as well. In this mode, you can create as many “channels” of video as you want and each new channel is placed on top of the ones before. If all the channels are displayed on the full screen, you only see the top layer. But, by keying out backgrounds or using 3D-style effects, you can actually view many layers of video at once and produce amazingly complex projects.

With all these layers and special effects, your simple software-based non-linear system can appear to be a multi-channel compositing and digital effects system. It can also be used to combine still images or graphics with video for other applications like titling. In fact, at least one of the software companies indicated that a titling module is under development for a future release. There really is no limit to what these systems can do as long as programmers keep improving them and third parties continue to develop plug-ins. If you’re ready to move into software editing, you’ll never want to do A/B Roll again.

SIGGRAPH Showcases Small World of Computer Graphics

Magazine: Content Creation Europe
September 1999
by Sheldon Liebman

For anyone who has ever been to a SIGGRAPH show, this year’s affaire, held from August 10-12 in Los Angeles, CA, recaptured some of the energy that seemed to be missing in recent years. Although final attendance figures were not available at press time, the show seemed to have a higher attendance and the people who attended appeared more eager to learn and to buy. On the other hand, the number of exhibitors was clearly lower, while the actual space occupied by those companies may have actually been higher than last year’s exhibition.

Mergers played some part in this trend, as Discreet Logic and Kinetix shared a booth, Transoft was part of the Hewlett-Packard exhibit, Play was combined with Electric Image, and many other companies appeared under new names. Another factor in this trend is clearly the size increase of individual booths. Just a few years ago, the largest exhibit at SIGGRAPH was probably only 15 meters square, but the largest booth at this year’s show was occupied by Intel and measured over 20 meters (70 feet) in each direction. In fact, it is quite possible that the largest 15 exhibits occupied more total space than the entire SIGGRAPH show of 10 years ago.

With fewer exhibits to see, the show seemed more manageable than in past years, but it was also more difficult for smaller companies to get noticed. In many cases, they appeared on the periphery of the exhibition area or on aisles that didn’t connect from the front to the back of the hall due to the monster booths around them.

Rather than trying to cover all the companies and products that were presented, we’ll focus here on some of the more pronounced trends from the show. Over the next few months, we’ll fill you in on more of the news from the show.

Motion Capture Creates a Commotion
At the NAB show earlier this year, it seemed that everywhere you turned, you ran into a company offering virtual sets or video servers. At SIGGRAPH, that feeling could be related to motion capture products. There are a number of different technologies to capture motion including wired sensors, optical sensors and metal appliances. In every aisle, at least one of these technologies was on display and in some cases, more than one. The two big vendors of optical systems were Santa Rosa, CA-based Motion Analysis and Tustin, CA-based Vicon Motion Systems. The advantage of these systems is that the actor being captured only has to wear very light sensors that are picked up by infrared cameras. This means the motion of the subjects is not restricted and they are free to move about however they want.

On the other hand, specialized staging is required to ensure that the cameras have an unobstructed view of the actors being captured. This staging is not required with wired and wireless systems provided by companies like Ascension Technology (Burlington, VT), Puppet Works (Ontario, Canada), and X-IST Realtime Technologies GmbH (Huerth, Germany).

All of these products interface with the most popular animation programs so that captured data can easily be applied to 3D models.

Building Models is a Blast
If motion capture systems make is easier to move objects around, laser scanning offers a similar leap forward when it comes to building those models. Every 3D animation system incorporates a modeling function (even, finally, Play’s ElectricImage), but most people don’t need to worry about how to model objects any more. From small hand-held products to cameras that can be pointed at large buildings, the process of generating realistic 3D models has never been easier, all thanks to lasers. Minolta (Ramsey, NJ) offers a portable system that not only captures the 3D data, but also includes the colour information so that a texture map is automatically built with the model. Polhemus (Colchester, VT) has a hand-held model that allows you to freely move around an object to capture it completely. The camera used in the Cyra Technologies (Oakland, CA) system isn’t small or light, but it can capture the data for an entire building just by sitting it across the street.

If you don’t want to use a laser system, you can purchase models from a number of companies, including industry-leader Viewpoint Digital (Orem, Utah), who was recently purchased by Computer Associates. Other products, like Microscribe 3D from Immersion Corporation (San Jose, CA), let you effectively “trace” a real-world object to turn it into a 3D model.

After seeing all the ways you could create a 3D model automatically, traditional polygonal and nurbs-based modelers didn’t seem nearly as impressive.

The Best Things in Life Are Free
One thing that was impressive was the ability to get powerful 2D and 3D software completely free. On the 2D side, Linker Systems (Irvine, CA) shocked the industry by creating a lower resolution version of their Animation Stand product that is completely free as a download from their web site at http://www.animationstand.com. The new product, called Animation Stand Personal Edition, is only available for the Macintosh and is limited to an output resolution of 256×192, so it isn’t suitable for broadcast animation. However, schools, web developers and professionals who want to try before they buy will find it very useful.

The free 3D software is called Blender and is available from Not a Number, based in Eindhoven, The Netherlands. In the free version, some of the more advanced features, like radiosity, have been removed. However, the company played a very impressive demo reel, which they promised was done with the free version. Even if you want the full version, you’re likely going to be able to afford it. A site license, which includes versions for SGI, Sun, Linux, Windows and BeOS, is priced at only $100US.

It’s a Small World After All
Perhaps the most interesting aspect of the SIGGRAPH show was that a large percentage of the exhibitors were not based in the United States. Companies from the UK, France, Germany and Spain were there, as were companies from Sweden and the Netherlands. At least one company was based in Australia and two others were from Belgium. As expected, there were a lot of Japanese companies, but there was also a company from Singapore, REVIVAL digital, who demonstrated a very impressive system for cleaning up noise and other artifacts present in video archives.

In every booth, at least some of the people spoke English, but not every company had an office in the United States. This is definitely a departure from previous years, where companies worked very hard to open a U.S. office in order to appear “real.” Perhaps the growth of the Internet and the development of the “global economy” have eliminated this requirement.

Let’s Party!
Since SIGGRAPH was held in Los Angeles this year, it’s easy to attribute much of the success to the show’s proximity to Hollywood and the film and television industries. It will be interesting to see if next year’s show, scheduled for July 23-28, 2000 in New Orleans, matches the energy of the show that just ended. If it turns out that LA is the key to success, we’ll only have to wait another year for the energy to return – SIGGRAPH 2001 is scheduled for August 12-17 back in Los Angeles.