Integrated Codecs and the Set-Top Mindset

Today’s compact codec packages offer amazing processing capabilities and remote control features, yet integrators still struggle with unbalanced audio and control complexity.

By Tim Cape, CTS-D

Back in the early 1980s, it took two or three full racks to run a small videoconference room. We were using telephone line echo cancellers and cameras that needed 100 footcandles to look good. However, we got it to work with a lot of audio, video and room design — and a quarter million dollar or so budget.

The codecs were monsters by today’s standards, taking up a good 10 rack units with large circuit boards that had to be handled carefully and were partially hand crafted. For all the electronics, the codecs had only one audio input/output and one video/output. Back then, all of the audio processing and video switching was external. We configured and controlled the external AV equipment the way we wanted. Those were the good old days.

Now we have compact codecs that fit in one rack unit with loads of connectors crammed onto the back panel, not to mention all of the set-top units that are available. The amount of processing in the package is amazing compared to 20 years ago. New units not only have built-in modern codec algorithms, but they also offer echo cancellation, video switching, audio and video bridging, and even remote control. You’d think I’d be really happy about this. But despite the fact that these are definite improvements, things could always be better.

The Set-Top Rules

For some systems, a set-top unit works just fine, particularly if a mobile system is shared among several small spaces. But the kind of projects today’s consultants face usually require an integrated system design, which typically means a multipurpose system that must accommodate local presentation function and videoconferencing functions without using any add-ons or portable equipment. But it’s not always that simple.

Sometimes customers ask for videoconferencing or two-way video distance education as an afterthought. They just stick in a casual “and we’d like to be able to videoconference” at the end of a sentence. What the owner doesn’t always realize is those innocent words just kicked the facility design, system design and, the budget, up a notch. Although it’s not as big a jump as in 1980, it’s still significant.

Once videoconferencing is added to a presentation system, it’s necessary to use a codec designed for integration rather than a set-top box or a boxed codec pulled out of a manufacturer’s pre-configured roll-about. As codec manufacturers came and went, and videoconferencing technology evolved, a large market for compact, pre-manufactured systems emerged, and the set-top videoconferencing system became a commodity.

Videoconferencing by nature isn’t easy. Beyond the equipment itself, It’s much more dependent on lighting and acoustics than normal presentation environments. And with all the additional cameras, video switching, echo cancellation, ISDN and H.323 interfacing, it was natural for manufacturers to try to make it as easy to use as your home stereo. At any rate, a lot of effort was put into the set-top mindset, and a lot of units were sold.

Why would you need anything but a remote?

Meanwhile, while watching the box sales grow, consultants and integrators still had to integrate videoconferencing capabilities into large (sometimes multi-screen or multi-room) environments that also needed to be used for presentation. Instead of getting easier, the game got harder. Why?

The answer lies in the set-top mindset. As the market bulked up on set-tops, the manufacturers tried to adapt these units to the integration market, largely by repackaging without a camera and monitor. This introduced us to some frustrating issues. Remember V-Tel and PictureTel? They tried but never really got it. Now Polycom and Tandberg are on the case, and things have improved, but could still be better.

Complaints from the integration market come down to two basic issues: the need for balanced audio and control. In addition, we need to be able to program when the picture-in-picture (PIP) is present on the preview monitor for the presenter in a conference but not on the larger screen for the audience. We can use external equipment for this, but today most codecs incorporate routing, PIP insertion, and video scaling right inside the box, we just need to be able to access if from the control code.

The audio side has improved in the past few years with connections other than RCA jacks for audio in some cases and even a few BNCs for video. The bigger problem — as always — is control. Trying to unlock the mysteries of modes and video routing in modern codecs — supposedly designed for integrated systems — is like trying to reach the deepest level of a difficult video game.

The Short List

So far, in my opinion, Tandberg has led the market in making a truly integration-friendly codec. Things such as easy access to audio setup, echo cancelers on/off, and VCR audio interlock on/off are available in control code. This helps because anything other than the smallest videoconferencing application requires the use of an external mixer and echo cancelers. But for the current or would-be codec manufacturers out there, a few additional features would make our jobs easier. Here’s what I think we need:

• Captive screw connectors with balanced audio inputs and outputs

• BNC connectors for video

• Discrete (not toggle) commands for every control function, including PIP on/off, location, or size, available simultaneously via RS232 and Ethernet

• Complete video switching control via RS232 and Ethernet.

The last item can be the most frustrating. Though it’s sometimes useful to have the hand-held remote logic built in to a codec that may be used with a remote and built into a roll-about setup, we really need to decide what video comes out of what spigot and whether it’s with PIP, without having the internal switcher control dictate what we can access.

I know this could pose potential support nightmares for manufacturers, but I think it’s feasible, perhaps with an “integration” mode that allows for full video routing access. Or perhaps we just need special override control codes that bypass any normally toggled functions and don’t allow the internal modes to interfere. Maybe we just need a codec like in the old days that simply processes audio and video and leaves the external routing, processing, and control to other equipment, which is pretty much how we handle the audio side already for the most part. We could have truly integration-friendly codecs with the currently available technology and manufacturer expertise, which would be a perfect combination. Remotes are so “last-century”.

email us