Is it feasible to automate transmission at low cost?

The issue of automation of transmission operations is usually linked to monumental investments in very complex equipment, difficult to maintain and with control systems that only the initiated understand. Almost always the image that comes to mind is that of a system mounted in four or five racks equipped with semi-transparent doors through which you can see a few video recorders and a tangle of servomechanisms that move frantically transporting cassettes from one side to another.

VTR manufacturers attacked the problem of automated playback by simply combining the components available at the time. They resorted to installing a series of recorders next to a cassette warehouse, and the greatest design efforts were concentrated in the delicate robotic systems that powered the recorders. Control systems were generally based on ordinary microcomputers that ran software dedicated to controlling the machines and routing systems needed to deliver audio and video continuously.

These types of systems required intensive maintenance programs to ensure adequate performance. Normally it was necessary to suspend the operation for a couple of hours a week to recalibrate the servomechanisms and the control system, which constantly had to be adjusted to the variations in the performance of the mechanical parts. Removing a recorder for maintenance could represent a one-hour shutdown, because in order to work safely on the equipment it was essential to stop the tape insertion system.

Over time these systems fell short of the demands of large networks. The main desire of the channels was to be able to accommodate last-minute changes. Normally these changes could be marked a few minutes before the broadcast, but what was unpredictable was the appearance of conflicts of the type "the-commercial-that-follows-is-forty-minutes-later-in-the-same-cassette"... When the customer wanted more flexibility it was usually necessary to install more transports and duplicate the tapes with the most used clips. This meant more money, more space and more complex systems.

The landscape changed completely with the appearance of the first DDRs (Digital Disk Recorders) really functional. Small companies specializing in remote control systems and automation in manufacturing plants attacked the automated transmission market by presenting an alternative working model: More compact systems, equipped with one or two VCRs, one or two DDRs and offering performance comparable to that of monstrous systems equipped with six or more recorders.

In these systems the protagonist is the software. By having a recording space with random access, transmission operations could operate non-linearly, accommodating changes seconds before each cut. Normally the system keeps on the discs the most frequently used clips , and uses the machines to record, a few minutes in advance, the material that is kept off-line, in the tape bank. With a system of this type it is perfectly possible that several hours pass without using a recorder for live playback , because the material that is going to be used immediately is always kept online .

The most recent advances in this line do not really come from our industry, but from that of software. Around the same time that the first DDRs were launched, a significant offer of administrative systems specially designed for television operations began to appear on the market, which gradually began to get involved with the management of equipment and playback systems.

Normally these administrative systems offer connectivity solutions that allow the commercial operation to be directly linked to the playback systems, in such a way that the playlists are generated automatically from the sellers' desktops. This allows to ensure rapid changes to improve the profitability of the spaces and also offer a high compliance rate. The appearance of these systems has led to the fact that in the large markets Prime Time advertising is practically auctioned in real time responding to the partial results of audience measurements.

The manufacturers' next step was to offer video servers with the ability to record and play back multiple channels at the same time. In an extreme case, channel one may be delivering a clip to the air while channel two is capturing the clip that immediately follows. And there is always the option of directly shooting a VCR, although the trend that is followed in practice is to leave the machines for the programs and the discs for the spots... And following this trend new players have entered the market: Manufacturers of non-linear editors, for whom the design of such a product is a relatively trivial project.

This technology has been developed to the point where it is possible to mount playback systems with the capacity to serve multichannel systems such as those operating in large subscription television operations. A tangle of interconnected recorders, DDRs and Servers form a playback system with enormous capacity. Thanks to the possibility of exchanging information in real time.

In the real world, automated operations are taboo for local channels and small cable operations. It is possible to resort to the totally non-linear systems of the latest generation that offer a very attractive cost/benefit ratio, but that in any case are usually beyond the reach of this type of company.

In practice these operations work with linear equipment and manually, and it is almost always necessary to resort to the pre-edition of commercial blocks in order to avoid surprises "on the air". Obviously the flexibility is minimal and the operating costs increase by having to "lock" equipment for the task of editing the "rolls"...

Thanks to the large cost variations in the market for hard drives and DDRs, new options are beginning to appear on the market. Take the case of a single-channel DDR equipped with 72GB of storage and a software package for running playlists.

We are talking about a box of two or three rack units with the capacity to make instant playback of up to 8 hours of stored material with an MJPEG compression of 8:1, which can be enough for many cable operations or open channels of small markets. Sounds great...

An option of this type can be very attractive, but you have to be cautious. Hopefully a DDR controlled by a PC and running a low-cost software package will not offer the same performance as a complex system with a tradition of several years of development and, in addition, constantly enriched by the feedback of many users. A "homemade" system is going to suffer from exactly what makes conventional systems strong: Robust integration, which is the key to flexibility and efficiency.

There are really no major technological developments in this field. The components are the same as those that the conventional manufacturer would use, and apparently the deficiencies in the control systems can be compensated if you have well-trained personnel. In fact, we only have the technological device to make random playback of video clips , but you have to make it part of a work system to make it functional. And that is why special attention must be devoted to three fronts of the problem:

The first is system power. We have a PC that controls a DDR, but does not necessarily handle the VCR that is going to be used to capture the material on the hard drives. And this represents a twofold problem: First, that it may be necessary to reconfigure the DDR to work in a linear environment while capturing material. And on the other hand, it will be necessary to manually mark the clips already digitized and, this is a wasteful and especially critical work to ensure good performance.

The second potential problem is the integration to the operation. Will it be possible to "trigger" the playback from the switcher of the master? This may require a GPI-like interface, or in some cases the availability of a serial port on the DDR. Generally, a solution based on "shooting" from the PC is not going to work well due to the delays inherent in a general-purpose operating system, such as Windows. And, of course, you have to consider the compatibility of the signals. Do not forget that very often analog inputs and outputs are optional.

Finally, connectivity must be considered. If the complexity of the requirement increases, it is possible that the control via RS-422 is insufficient to ensure good performance. If several DDAs are going to work in parallel, it would be better to have network connections that ensure good speed and eventually allow the DDR to be fed directly with the playlist. On the other hand, we must solve the computer problem of transporting the information of the playlists between the commercial area and the playback system.

A good engineering job is always based on the prevention of failures and that is why it can also be convenient to get redundant storage. With some arithmetic you can examine the cost that can represent the sudden failure of a hard disk of US $ 1500, which is exactly the value of the second disk that can "save the life" of the channel in an emergency.

In conclusion, it is perfectly possible to install a small automated playback system using low-cost DDRs. The necessary software is on the market and it is to be expected that in a very short time it will improve drastically to solve many of the current shortcomings. But what needs to be judiciously reassessed is the operational cost of doing the system integration work "at home". And perhaps that's why it's wise to wait for this technology to mature.