Uncompressed NLE

Non-linear editors without compression and with the ability to natively handle D1 material are becoming more accessible. The reference level for SD production is within the reach of (almost) everyone. And many producers invest significant sums in a magic box that allows them to say "I work in 601"... only to discover some time later that their editor does not take them along the route that best suited them.

The development of digital post-production systems follows the basic trend of the industry: The supply is multiplying and segmented every day. Each manufacturer tries to position itself in a niche. Avid is committed to connectivity and integration. Media 100 to multimedia and Internet. Accom to the power of hardware. Discreet and Quantel continue to offer the same old thing: high quality. Presumably, Lightworks will continue to appeal to the power of the elementary. And many other manufacturers try to build bridges between high-level production and multimedia studios. They all have different products... and all are uncompressed non-linear editors.

The really curious thing about this matter is that the market generates pressure for us to produce in 601 environments while the distribution migrates by leaps and bounds towards the use of MPEG2 streams of very low bandwidth. What can the producer feel when he sees his D1 product on a cable channel that "travels" in an MPEG stream of 4Mbits per second? Here is the first question. Do I really need to operate without compression?

As always the answer depends on what is expected of the product. If it is short-lived material the answer is a resounding NO. But a product that is expected to be valid for about five or six years needs the quality of the D1 environment to be able to compete successfully in DTV environments. Let's remember the old maxim of the computer world: Garbage in, garbage out. A 601 signal in NTSC consists of about 480 interlaced visible lines. By becoming a progressive DTV signal of 480 lines the average viewer perceives a resolution improvement of around 40%.

We are talking here about a significant improvement in the sharpness of the image that can make perfectly visible the defects caused by the MJPEG compression algorithms. If this problem arises when converting our material into SD signals, what can be expected from the cases where our masters have to migrate to HD? This factor of 40% comes from totally subjective measurements, but it is the one used by the manufacturers of upconverters to generate the algorithms that will lead our material towards the wide world of digital television. And what does this mean? That if we need the extra definition that is obtained when mastering in D1 environments.

Each brand has a tradition that is somehow reflected in the current product. One of the fundamental differences is the user interface of each system. Have you ever witnessed a "cordial" discussion between supporters of Avid and Media 100? It's really amazing how passionate this topic can be among today's publishers. Normally one of these discussions can be settled by clarifying that any team at this level does its job competently... or else it would already be out of the market. This is the advantage of facing mature products. But don't forget to consult your editors...

Another important factor is the productivity of the team. It is one thing to have access to D1 material handling and quite another to have real-time equipment performance. You have to check the specifications very well, because sometimes the cost contains traps ... You have to be careful with the "almost"... if your editor does playback "almost" in real time, or presents you with the result of an "almost" cut immediately you are in serious trouble.

Don't forget to find out how many video streams the computer can handle simultaneously. Two streams is the minimum to be able to have 2D effects in real time. And generally teams that offer two flows have better response times in daily work. Some manufacturers offer the option to add additional video "channels" to speed up the work of composing effects in several layers.

Returning to the topic of software , consider some features that publishers tend to forget... Is there direct compatibility between your preferred team and the dominant brands in YOUR local market? You may need to make some sort of business alliance with your valued competitors. And it would be very convenient to be able to exchange files transparently. The same applies to sound studios... Can I transfer the audio from my projects to the audio station of the studio chosen by my client? Can I import graphics with ease? Can I directly capture audio tracks from a compact disc? Connectivity is Productivity.

More desirable features... Does the equipment include an editing listing processing package? After all, it's quite possible that your new non-linear editor will function as your neighborhood's online editing room. Does the manufacturer offer viable backup systems? Can the system use an external driver? The list is endless...

Let's talk about effects production. Real-time editing is not synonymous with real-time effects. And real-time effects is equivalent to additional hardware . Do I really need real-time effects, or can it be better to bring the work to a composition system? It all depends on your needs. Sometimes it is possible to assemble a PC for graphic work with less money than the option of 3D effects costs for its editor... but in that case it must undergo duty cycles that include relatively high render times. On the other hand, the use of packages such as Adobe After Effects or Commotion will give a more refined look to your products... and it will only be limited by the creativity of its people.

The alternative for producers who need to maintain the style of conventional equipment may be to acquire one of the editors that incorporates a renowned effects unit. The decision is yours...

Finally let's do a brief review of what it means to operate in a D1 environment. The ITU 601 standard contemplates the use of digital video signals by components with a spatial resolution of 720x486 pixels in NTSC. The funny thing is that the video can be encoded with a color depth of 8 or 10 bits. The difference doesn't seem significant, but it may be interesting to ask an engineer if it's better to use a 10-bit switcher or an 8-bit switcher ...

8-bit 601 video can present serious color resolution problems if special error dissemination processes are not applied to it. And it is usually necessary to handle the effects in 10 bits to avoid the appearance of visible defects. And here's the final question... What coding does your favorite computer use? For now, if we talk about reasonably expensive equipment, the invariable response is 8 bits. Some manufacturers have developed very ingenious systems that allow to obtain 10-bit quality with 8-bit processing. but that comes at a cost.

We leave the issue of storage in the pipeline, but keep in mind the following: SCSI technology barely manages to hold more than two simultaneous streams of uncompressed video. Think FibreChannel and data transmission rats of the order of 80 to 100Mb/S. And don't forget to do clear accounts: An 8-bit D1 digital video frame equals just under 700Kb of storage. The rest is pure arithmetic.