Examining a 'Switch' to HD

Avoid vaporware in choosing new HD gear

Avoid vaporware in choosing new HD gear

AMES, IOWA: Over the last few years, IPTV's digital conversion has focused on design, purchase and installation of RF systems. We have not completely taken our eye off creating content, but with limited resources, the federal mandate, and the lack of available grant funds for production equipment, moving the studio transition along has had to wait.

The delay has actually been more of a benefit than a detriment. Early on, we purchased hardware to acquire and edit in HD, so we have begun building a library of HD content and gaining valuable experience in working in the format.

During that same period, real HD production equipment has finally started to show up on the market, although there still appears to be a lot of vaporware and limited selection. We've really noticed this as we began to investigate converting our online production control facility to HD.

And I do have to specify HD and not just digital. There are any number of SD production switchers with various quantities of inputs and M/E's but when looking for HD devices, the field rapidly narrows. We first started looking for HD production switchers almost four years ago and found virtually nothing available. The manufacturers claimed to be waiting for the market to develop, but I wondered how it was supposed to happen with no hardware available.


Our existing online room consists of a Grass Valley 250 production switcher, DPM-700 digital video effects system, Pinnacle still store and a Dubner 30K graphics system. Ideally we plan on replicating this room in HD, which should be doable.

But there are a number of challenges we face. When we initially began discussions with our technical directors, one of their chief complaints about the GVG-250 was that it only had two and a half M/E's, which at times, limits production capabilities. So we started looking at three and four M/E switchers but quickly discovered that a single M/E on several of the models we looked at have all the capabilities of our existing 250. Just counting M/E's doesn't really yield meaningful information without understanding the capabilities of the M/E bank and the needs of the facility. To really make sense out of the selection process, it is essential to let the users evaluate the operational characteristics, layout and ease of use. Stacking multiple capabilities into a single M/E may make great sense in an automated or edit-controlled environment but may be meaningless in a live or live to disk program if those capabilities are buried in layers of soft keys and menus.

One critical thing that the engineers need to be aware of when evaluating the choices in switchers is what capabilities are actually deliverable and what are in development. As we move further and further into the software-driven digital architecture, products are inevitably released to market before the software is finished. This can be absolutely devastating. Being a sports fan, I have loved watching the network sports in HD but I have been surprised by some of the glitches that I have seen in what are fairly simple transitions and effects. When I ask about the specific problems I am amazed at how many of them are known software issues that are being worked on. I won't single out any manufacturer because the problem seems to be ubiquitous.


This becomes even more critical when looking in depth at the switchers because they are now so much more. When you look at switchers like Grass Valley's Kalypso, Ross Synergy or Sony MVS series, they blend switcher, DVE, still store and more into a single package. For me this is something that really needs to be closely examined. I have never been a big fan of multifunction boxes for a number of reasons.

In many cases, to access the various functions requires either soft switches or a complex control panel that is difficult to use in a live environment. Perhaps even more critical is the impact to the CPU in the switcher that has to manage all of these various tasks. To the best of my experience, no CPU's really multitask; they task switch, albeit very fast, but they are still doing one thing at a time. Now start stacking complex tasks for the CPU to control and the load can increase to the point where the efficiency and the capabilities of the system are compromised. I am not suggesting that this is the case in the switchers that I mentioned but performance under load is a factor that needs to be measured and evaluated. Sales demonstrations tend to be formatted to show off the capabilities of the switcher in its best light and may mask any flaws; and remember in our area we are looking at HD data rates and not SD.

Probably the one area that I am most dubious of is the idea of buying a system for SD with the idea of doing an upgrade to HD. Let's face it, we're dealing with digital technology, which at its base unit is a commodity with fluid pricing and power. What are the chances that three to five years from now, there will be an upgrade path for an existing SD switcher that is cost effective and can match the capabilities of a current system? I think it is false economy to look at SD as a gap filler, thinking that the hardware will be reusable. It also seems self-defeating to say that the local stations don't produce content that merits HD. If we don't think our content merits the highest quality, why would the audience?