CitrusTV Video Wall
Walk into nearly any television studio (local news, network news, sports, etc.) and you'll see video walls are staple of television production in 2017. These large displays (made up of many smaller televisions) are used to play video, display graphics, or become interactive elements of a telecast. As the general manager of CitrusTV, I'm always looking to the industry to see what trends are taking hold and how we can best prepare our members for what they'll experience in the workplace while simultaneously upping CitrusTV's production quality.
Originally, I intended to replace one of our aging standard-definition flex plasma screens with a new 70" LED TV. This would certainly have improved things by itself, but I quickly realized this new monitor would still be placed in front of our "B" set which is used for interviews and other segments that don't wind up on the desk. This "B" set used a backlit vinyl backdrop which was originally stapled into place but later made into a velcro removable vinyl which could be swapped between a morning or night skyline scene. This left something to be desired and looked, frankly, horrendous on air. You could clearly tell it was a wrinkled vinyl and despite much fiddling and cursing, we could never seem to get a satisfactory look with no wrinkles using the velcro method.
Immediately my thought went to a video wall. This way, it could serve as a "monitor" into which we could display video or graphics as well as solve the issues with the backdrop. As I began planning and sketching what this could end up looking like, I ran into a few problems.
- The "B" set sits about 6 feet away from the nearest wall. This was originally done to give space for the fluorescent backlights for the vinyl backdrop as well as to allow for storage space for cases, shelving, etc. in the studio. This space gap meant that all of the monitors in the video wall would have to be ground supported or flown from the grid above rather than being easily wall mounted.
- The opening where the backdrop sat was 94" x 47". If you're good at math, you'll know that that's not even close to a 16:9 aspect ratio meaning I had to get creative with with the configuration of monitors. (If you're keeping score at home, it's actually a 2:1 aspect ratio). This meant that using four, or even six TVs arranged in a horizontal grid would be problematic and not adequately fill the opening without losing a ton of screen real estate on the top and bottom.
- I needed to be able to feed the display with a simple AUX bus from our video switcher without the need for any sort of controller software or any need to create video content specifically for the wall. In other words, if I wanted to roll a video on the air or roll it in the wall, I should be able to utilize the same piece of video.
With those goals in mind, I began looking at cost. I quickly realized that so-called "video wall displays" which have bezels only a few millimeters wide would be prohibitively expensive. With panels ranging from $5,000 to $10,000, such units would be difficult to fit within my limited budget. Plus, I had to keep in mind my odd dimensions where video wall panels are typically designed for normal landscape orientation.
I began looking at who made a television with the smallest bezels. I found some consumer-grade panels from Samsung, LG, and the like, but I also wanted to consider longevity. Since these would be used in a broadcast environment, I wanted to ensure that whatever panels we used would have good guarantees and appropriate duty cycles. After careful consideration of dimensions, I figured out that the best thing to do would be to use three large televisions mounted in portrait orientation. The panels I used are NEC 65" TVs which are commercial grade panels (albeit at the lower end of NEC's line) with a decently narrow bezel for the price. With the panels picked, I had to figure out how to give them signal and mount them. The dimensions of the televisions actually meant that when it was all said and done I overextended the opening by about 2 inches on both the left and right and about 5 inches on both to the top and the bottom.
As mentioned previously, these panels could not be mounted flush against a wall; they had to be freestanding. This meant I'd have to anchor them to the floor or fly them from the grid overhead. I quickly ruled out flying the panels since I'd have to use impossibly long poles to get them close to where they needed to be coupled with the fact that the spacing of the grid put the nearest possible hanging point about a foot away from where the wall needed to sit. I instead went with the floor mounting option. I'm a Chief mount guy through and through (always well built, easy to install, easy to adjust) so I looked there first to see what options they had. Chief makes a number of multi-display mounts which support various configurations of video walls. They did have a 3x1 configuration (which would have worked with my design), but it was more than $2,500. So, I elected to use some Peerless A/V mounts wherein I'd be mounting each TV individually and then the TV would mount on poles. This worried me a little bit because if I didn't space the floor pieces exactly right, these mounts allowed for no lateral movement. I could easily move them up and down on the poles, but if I was even a little off when I bolted the mounts into the concrete floor I'd be in for a lot of headaches.
Thankfully, finding a way to drive the wall was much easier than physically putting it together. My requirement for controlling the video wall was I didn't want to have to have a complex process to drive the wall. In other words, I didn't want to have to make graphics and then have to rotate them to fit the portrait orientation or worry about using a specific piece of software to drive the wall. I wanted to be able to send it a simple AUX bus from our switcher and roll whatever content we wanted in the wall. I found a neat little device from Black Box Networks called the VideoPlex4. It takes any standard video signal in and allows you to display it on up to four monitors. Each output can be independently cropped, rotated, and positioned. You could duplicate the display four times, crop it into four even pieces and more. I was able to use it to chop the images into three pieces, rotate to portrait and account for the bezel widths (more on how I did that later). The VideoPlex is a cool device because it can actually take a 4K signal, so if you were splitting in the input into four pieces it wouldn't have to upscale the image at all. In our case, we're sending it 1080i, so it upscales the image to fill the monitors.
Installing the wall went better than expected. I used a Hilti hammer drill to drill 18 holes into the concrete studio floor. I was concerned about cracking the tile, but with a sharp bit, the holes were clean and precise. I used a shop-vac next to the hole to keep the dust to a minimum and kept a bucket of water to cool the bit. After some careful measuring, everything ended up lining up pretty well and getting things leveled turned out to be pretty easy. The biggest problem was actually that the studio floor turned out to be just a little bit out of level on one end, so I shimmed it with a few washers under the front side of the floor plate.
Photos above during the install process.
Once the wall was installed and cabled, I had to set up the controller. This required a little bit of math, but essentially I used two patterns to aid in setting up the wall. One that has lines every 10px horizontally and the other that has lines every 10px vertically. This way, I could see exactly how much of the image I had to lose to make everything line up in one seamless image. Then, I could simply enter coordinates into the video wall controller software to make the images match.
I also used the 10px patterns to see how much of the wall I was losing on the top and bottom as well as where the individual displays lined up. Using this information, I created a Photoshop and After Effects template with guides to show where to build graphics. Our sports department, for example, created a cool graphic to compare three players where each one appeared in each physical monitor and as they talked about each the other two were blown out of the space and video took over the other two panels. It looked very cool and it added a neat visual element to their show.
Overall, I'm thrilled with how well the wall came out. It came in well under budget and added a tremendous amount of value to our shows. It looks better and performs better than walls I've seen that cost well in excess of $100k.