Mediapro has reached the figure of 3,000 matches produced remotely. With this milestone, we are taking the opportunity to explore the keys to this type of broadcasting with Emili Planas, CTO of Mediapro. “For remote production to be a good option, the stars have to align. It’s not something that always pays off”, he warned.

*This article is included in the March issue of TM BROADCAST. You can access the full magazine here.

Remote productions have marked a turning point in sports broadcasting. And a pioneering top-level competition in its implementation was Grup Mediapro for LaLiga, which implemented for the first time ever this type of production 14 years ago, when there was still no integrated solution available on the market, carrying it out with in-house developed technology.

This football competition has recently reached the figure of 3,000 matches produced remotely. We take advantage of the pretext provided by this figure to explore all the particularities of these productions with Emili Planas, CTO of Mediapro, a voice of undisputed authority to address this issue.

The first thing to clarify is that there is no single remote production model and, depending on the type of sport or competition, the latter will require certain specifications. Emili Planas has told us how Mediapro implemented this technological innovation for LaLiga more than a decade ago: “What we did was set up our own structure with requirements that remain in place to this day. We did not start it as first system from which to settle and then gradually evolve, and instead we for it all and we achieved the most already at that time, succeeding in getting the entire team in charge of production, camera control, etc., experience the same operating feeling by doing it remotely as being inside an OB.”

“We were the first company to use remote production to cover a high-level remote event”

The challenge was, therefore, to ensure that latency was minimal while maintaining or even increasing the quality and security levels of the production. “In order to achieve that minimal latency that would preserve the good feeling I have mentioned, we had to develop a series of rather curious solutions”, went on Planas. “Nowadays this is easier, but 14 years ago it was something very innovative. In fact, there was only one specific encoder model on the market that could achieve such minimal latency with the bandwidth available at the time and the required quality.”

“We devoted three months to the technical development process and the different tests concerning all the engineering, where the five areas that at that time made up the group were involved: engineering, Unitecnic; sets and OBs, which at that time worked as separate entities and were not integrated within the Eumovil company as now; Overon for the communications side, networks and encoding part; and Mediapro, in charge of the executive operation and coordination.”

After that time, and with the coordinated work of the five aforementioned units, Mediapro developed a model that is still in place today. “We were the first company to use remote production to cover a high-level regular competition,” stressed Emili Planas.

Over time, the use of this model was transferred to Division Two. “Interestingly, we started producing first-rate matches remotely, but with the evolution of broadcasts, the increase in cameras and the production requirements of LaLiga, both for the Top Division and for Division Two, the model we initially became suited to perfectly match the accentuated models found in Division Two,” explained the Mediapro CTO.

“In addition, it coincided in time with an increase in the available bandwidth in Division Two stadiums, which made it possible to implement this model, something that had not been possible before,” he remarks. Out of the 11 weekly matches, at least seven are currently done with remote production.

“What we have done, only more recently, is to enable a second remote production model,” pointed out Emili Planas. “The first is a model that is traditionally called REMI. It is a model in which, to put it simply, you transport all the signals to the production center, so you do the program there, that is, in the OB you only send, so to speak, cameras and microphones. ”

“With this second model we implemented,” he went on, “what we do is operate the OB’s devices remotely, and we do apply this in all Division One matches and in Division Two matches where we do not do REMI. With this model, there are a certain number of operators who do not travel, as they operate the systems that are in the OB truck from the production centers.”

Regarding the scope of these productions, Planas indicated that it allows the deployment of up to 12 cameras, two of them super slow motion, although it all depends on the specific conditions of each match. “The range is between eight and 12. And two of these cameras are are super slow-motion with three phases, so you have to send four signals from the stadium to the production center to have the composition plus the three phases.”

“Are we talking about hybrid SDI-IP productions?” “We are transitioning,” says Planas. “Although we’ve been using IP networks for contribution for over 20 years, until recently, all local workflows were SDI. Right now, we are combining IP-2110 workflows with SDI workflows, and in many cases, we use hybrid models where IP and SDI are combined as effectively as possible.”

“No one would be able to tell whether one of our productions is made remotely or locally. This was our main challenge and goal when we designed the system”

Software-based solutions

On the other hand, when asked about the use of cloud or on-premises systems for this type of production, Emili Planas indicated that it is more accurate to talk about software-based solutions. “Whether the software has its deployment in the cloud or on-premises is something else. In general, what we do is classify the solutions based on their use, whether they are software-based or a bit more traditional. At present, we are using software-based solutions that we open in the cloud or on-premises based on variables that do not have an operational sense, but are rather related to the availability of systems, the flexibility of one with respect to the other, economies of scale, or the number of times the system is used.”

“In this regard,” he pointed out, “if we massively use a system on a permanent basis, it is not usually possible to mount it on the cloud, although sometimes it may be the case that it is. It usually pays off just to buy it. If you have, of course, the rack rooms, the maintenance equipment and all the staff and the operating capacity that we have”.

“Sometimes we resort to the cloud,” he added, “because it’s the best option. If you want to create subtitles using AI with a speech-to-text system of certain quality, cloud-based systems are working really well, compared to others, isolated in your rack room that don’t learn or grow.”

“This is a bit the spirit of it all,” he concluded. We are making parallel products for LaLiga through software-based systems, as the multiscreen for example, but not under this remote production model, because, although it would be possible to do it, it would not be likely to meet the quality requirements we mentioned.”

“The technical and operational quality achieved remotely is exactly the same that we would achieve by doing it locally”

What determines whether a production of these characteristics is carried out remotely or locally? “Logistics, the distance that the OB truck must travel is what weighs on the decision,” said Planas. There is a clear advantage: if the OB truck can do without the mileage, you become more environmentally friendly and save costs. And we strive for both [Since 2014, Mediapro estimates that they have saved 14 million kilometers, which is equivalent to 7,000 tons of CO2]. So, if a Division One OB truck has to cover the same area as for a Division Two match, it is much more efficient to send that truck and do a few miles, instead of sending one that is further away, even if it is a lighter truck or a smaller van. Through logistics, we optimize Division One and Division Two resources to improve efficiency. ”

In any case -Planas made it clear- the reason for deploying or not a remote production with these top-level models is not one of a technical or operational nature. “This type of remote production can cover competitions of what we call Tier 1. When top-level competitions are not covered with these remote production models, it due to other reasons. The technical and operational quality achieved remotely is exactly the same that we would achieve by doing it locally.”

“In fact,” he remarked, “no one would be able, not even a spectator specialized in football productions, to differentiate whether a production we are making is remote or local. This was our main challenge and goal when we designed the system. “We aimed to ensure that there was no perceived decrease in quality or increase in latency compared to other matches and, then, when synchronizing with radio stations, one wonders why one match has more delay than the others. These considerations were relevant when defining the model and continue to be important to us when defining any new remote production model.”

Emili Planas insisted on this point: remote production cannot be tackled in generic terms and each sport, of course, presents its particular challenges. “There are other competitions or types of production in which latency may be less relevant or where, beyond objective quality, a quality evaluation plays a bigger role. Here we can deal with models in which costs surely can go down, but at the same time they will have their own limitations. If you have to do a replay in super slow motion, for example, lower-cost solutions will give you a lot more problems…”.

“These are the elements,” he continued, “that influence the decision of how this remote production model is made as compared to other models that we also use and that do not quite reach this level of maximum quality.”

“For remote production to be the best option, a lot of stars have to align”

In the case of LaLiga, Planas concluded: “Working remotely, with the principles that we apply to carry out a high-level broadcast of LaLiga, does not make any difference at operational level as compared to working in the OB.”

A different matter is dealing with other sports, where perhaps other specific models can be applied for remote production that in the case of football would not be possible: “There are certain remote production models that do pose limitations, which some sports can accept without any issues. Motorsports, for example, where the time between an action and its replay can be somewhat longer.”

“In basketball and football -Planas argues- the replay has to be prepared immediately after the action happens, because it is usually at that moment when the game stops. That’s when you have time to show it and that’s when the viewer expects it to be shown. In other sports you can wait longer to have the replay ready and it is not relevant in most cases if you take a little longer or a little less time.”

So, going back to the motor example, it is not that difficult to select the camera for the replay of an overtaking or an accident, for example. “You can choose between two or three cameras, but usually the good shot is very clear. In the case of football, there are many cameras where the action appears, and choosing the best one is not so easy,” illustrated Planas.

In football, the choice of the correct shoot requires the operator to have a certain degree of experience to act quickly, and a well-coordinated team is also needed to achieve the same operability as in an OB. This differential element of football, in short, means that not any remote production model is worth it, but very specific models are needed instead.

However, this does not mean that, with the right model, production can always be done remotely: “For remote production to be possible from a financial standpoint, a lot of stars must be aligned. It is not something that has come to save our lives and that from now on everything will be done remotely. It doesn’t always pay off.”

It is necessary here to make a distinction between the most affordable remote production models, which democratize production while making it possible to broadcast competitions that otherwise had not been  possible to broadcast until now, and top-level models. “When you have connectivity and pluggable devices, remote production is much easier than sending an OB. Working in SRT and high compression modes, you can do remote production based on cloud or software systems. But -no offense intended- we are talking here about a Tier 3 production model, designed for sports that until now could not afford high-level production because it is more expensive. ”

Does it mean that this model is suitable for all types of production? No, it doesn’t. “If production quality must meet strict standards, not only at a technical level, but also at an operational level, due to the speed of response it demands in the event of uncertainties that occur during the competition, latency and so on, then you have to go for a production model, whether remote or traditional, that is a bit higher.”

But this is not the only ‘star to align’ -referring to the metaphor used by Planas-: “Remote production allows you to mutualize the operation and for the same control, for example, to play a game at noon and another at 6 pm. You have that control doing two matches in one day instead of having an OB truck that will only to one match. What if in the competition calendar all matches are played at the same time? Well, then one of the stars becomes misaligned.”

Another important factor is the distance traveled by the OBs, and Planas provided here the example of the rugby competition in Argentina: “In Argentina, we were considering covering the rugby league with remote production. When we analyzed all the options -which took us a long time-, we realized that all the clubs in the league that we were going to cover except one had their venues around Buenos Aires. That is, operators would take the same time to go to the production center as to the stadium. Suddenly, the distance star was misaligned.”

In short, and serve this as a final reflection: “When we want to set up a remote production model, we always analyze absolutely everything and evaluate that the stars are as aligned as possible in order to offer the technical solution that best suits the needs, cost and operation that has to be done. Just one model is not good for everything.”

It is necessary to distinguish between the most affordable remote production models -which democratize competitions broadcasting- and the highest-level models

Postscript

Although we have focused on the technical area in this article, it is worth mentioning, as an appendix, the advantages that this type of production entails on a human level, as it frees a good part of the team from making large trips in OBs and allows them to better balance work and personal life and helps reduce accidents.
Despite the reluctance that could initially arise among producers, Emili Planas indicated that now they are all delighted, and illustrates with this anecdote the comforts that this new form of production poses: “The other day someone said to me: ‘What I like most about remote production is that I have the bathroom behind me, I don’t have to look for it in the stadium and carry the relevant badge or ask a favor from the owner of the bar next door.'”