At the second MotoGP round in Austin, I spoke to Sergi Sendra, Director of Dorna Sports TV Production, about what goes on behind the scenes when bringing MotoGP to TV audiences around the world. Mr. Sendra graciously found more time for MotoMatters at Laguna Seca, so that we could ask him about the popular slow motion shots, among other things. You may want to read the first part of this interview, here, before continuing on to the conclusion of the interview below.
MotoMatters: Now that we have some idea about the complexity of the TV production, I’d like to know how you manage the logistics of getting everything from race to race. For example, last weekend we were in Germany, and now we’re at Laguna Seca, so in a couple of days you had to get everything packed up, flown across an ocean, and then set up again.
Sergi Sendra: The system is designed to work so that we can install everything in two days, that is normally between Tuesday and Wednesday. Some people at the circuit may begin on Monday, because the main cable for the cameras at the track is sometimes run by a company separate from us. This company usually starts Monday and continues on Tuesday and Wednesday.
But the main equipment arrives, let’s say late Monday or early Tuesday. And then we spend two days installing everything. Thursday is the day that we have to do all of the tests. Maybe there are some issues when they finish the set up, and the technicians we have are now a standard group of key people.
Now the way we work is more specialized, and we gain time by spending the money on somebody who knows the job, so it takes them less time and they are more effective. And this also helps to avoid many mistakes because the cabling process is as important as [anything else].
If you don’t put the cables together properly… For example, every circuit has different layouts, this affects us profoundly. So the plan is that the way the things are installed is as effective as the way you dismantle them. If you don’t think about the dismantling process it will take you longer.
We spend Sunday doing that, after the races, we usually finish around 10pm in Europe and between 11:30pm and 12 midnight when we are abroad, or when we have to jump from Europe to America like we did in Germany. For example in Germany, everybody left the circuit at 11:30pm.
This is a big challenge and now it’s something that is really positive for us as our people can sleep or have time to catch a flight. It’s not a brilliant time to finish, but this is something we are working on for the future. You learn year by year and you try to simplify things, use special packing cases, use fewer cases, compact the things we can. If every year we bend something then this changes the logistics. Once a part of the process is stable, we say “Ok, now we will compact this and make it better for the following year.”
MM: So the cable contractor arrives on Monday, and your gear is in transit on Monday…
SS: The gear goes to the airport at night, and on Monday will be travelling by plane to wherever it is going. Usually this takes the whole night, because you get to the airport, unload the equipment from the trucks and then load it onto the plane and it goes. This is a process that Dorna has managed for the last 6 or 7 years and we have learned a lot by now because we know which equipment must arrive first, second etc. as there is more than one plane.
MM: So while the gear is in the air, they start running the cables at the circuit…
SS: It depends. There are different situations. Sometimes the cable comes on the plane and sometimes the cable is at the circuit. For example, in Australia the cable is at the circuit weeks before.
MM: And that’s not Dorna cable?
SS: There are different cases. In Europe, it belongs to the same company. In Malaysia and Japan it’s the same company, in Australia and America it’s local companies.
MM: But that’s not Dorna’s—you don’t own that cable?
SS: No, but if you want to know more about cable, which so important and this is interesting because it is just increasing and increasing. Just the radio frequency equipment that we manage, and that is for the on-board cameras, the antennas for sites around the track, to cover the pit cameras and the boxes and cameras in the paddock, the helicopter and so on—all of this requires 16km of fiber. This fiber arrives with all of the equipment and will be distributed on Tuesday and Wednesday.
MM: By ‘fiber’ you mean fiber optic cable?
SS: Yes, fiber optic. This is something that we own and we need to implement it at every circuit because many of our cameras depend on this cable. Also, the fiber now allows us to be more efficient with the audio, with the commentary positions. Really, the fiber optic cable has become one of the top issues and the biggest advantage is that you can manage a lot of signals with a very thin cable.
MM: Are you gradually moving toward all fiber optic cable?
SS: Yes, even the track feed. The track feed is moving to all optical fiber. In fact, here in America we work with a local company, in Laguna and Indy, everything will be fiber. There is no triax. The old cable is very thick with a lot of signal loss, so fiber is definitely the best way and you can increase the speed of communication, no? We are in the 1 gig area if I’m not wrong, but you can go to 3 gig with fiber, it’s the only cable technology that allows it.
MM: When the telemetry for the bikes is shown on screen, where does that information come from and how accurate is it? On screen we see the lean angle, it shows the revs, braking and acceleration…
SS: All of the information comes from the bikes, from the telemetry of the bikes. We share the system but not the channel. I mean, the bikes have their own channels [to communicate with the teams]. From the beginning, when we agreed with the manufacturers to do this, we decided to have a complete, spare and independent parallel channel, it is called “CAN-Bus.”
This channel is dedicated to Dorna, which as I said is an independent channel: If there is something happening with the telemetry, it will not affect our output. Maybe we’ll have no data but when [the teams] go to the bike to pick up their information, they can do it on a separate CAN-Bus.
So, there are two different channels. One for their telemetry reading and one for Dorna which is live data acquisition. And for sure we are increasing the amount of information that we’re requesting from them. We started with RPM, speed and gear, then we requested the throttle and brakes, and finally we requested lean angle. So all of this information is something that we manage with our own electronic boards. We also have GPS, then we have boards dedicated with accelerometers and gyroscopes to pick up the balance of the lean angle.
At the end, the accuracy in some parts is total, in others it’s not. There are people who say “I don’t believe it’s 18,000 revs!” The information we receive we want to be as accurate as possible so that we show something with credibility. This is like saying if you give somebody five samples per second, you have less information than if you give 100 samples per second. Ok, we receive a lower sampling frequency and the final drawing [on screen] is not perfect, but you can feel the changes. It’s important to know that technically, we are in a situation where we are not giving wrong information, as this would be negative for everybody.
MM: So just to make sure I understood that, the channel that goes to the team’s data engineer has more samples per second?
SS: Exactly, they have more accuracy than us, but our accuracy is a true accuracy. It is not wrong information. And as you can see now, for example, the lean angle is really, really accurate because there is a magic number, that nobody wants to say, for the lean angle, but we know it’s close to 60 degrees. And one of the goals for us it that the [viewers] understand and learn something that we never did before.
When you arrive in MotoGP nobody tells you “Oh, the bike has two wheels and the lean angle is 62 degrees.” No. You have to start learning that the bike leans and this is maybe the nicest thing that our sport has compared to all of the other sports. This year, the goal has been to achieve a real number, but more than a real number is how we can best explain to people in order for them to understand that this is measuring degrees. And the degrees are in respect to a horizontal level when the bike is standing at zero degrees.
To show how the bikes lean, we will make some kind of a video which will be launched this year where we will compare a scooter, a custom bike, a street bike, a superbike and then a MotoGP bike. This is for people to understand, oh wow, we are at the highest level. And the lean angle number is important. Because I know, when people study, it’s easy to learn about linear measurements but people don’t like trigonometry! For us, it’s very important to explain it and to draw it in the best and easiest way so that everybody can finally talk about this concept as one MotoGP concept that we can have forever.
MM: So that’s interesting because I think of the things like the telemetry that Dorna presents as intended to entertain the audience, but you say that you put that up also to educate the audience and help them learn more about MotoGP. Can you do other things, like the telemetry, to educate the people who are watching instead of just entertain them?
SS: I think we can now bring the show to a level where you will learn and understand better and therefore you enjoy it more. The tires, the tires are another of the big surprises. Now with the slow motion and the high speed cameras, we are watching movements in slow motion that we’ve never seen before.
There is another family of concepts, like the spinning, the steering, the over-steering, that to people who don’t like or aren’t interested in bikes, it’s our slang. We have our own slang. For an inexperienced viewer, someone not a motorcycling fan, he will need to understand easily what he sees on TV and say: “Oh wow, the rear wheel is faster!” For us, we ask, “Can we see this? Can we show it? Can we explain it?” And then when you draw it and people understand it, it attracts people more because they’ve learned something that they didn’t know.
So I think there are a lot of concepts that we need to take care of because some of them are very technical and very complicated. For instance, the spinning is measured in percentages over 100%. That makes things complicated, managing percentages, you know? If you talk about 50% you understand it’s half, if you talk about 115-120% it’s like “hang on a second, only 100% is possible.” You must explain that it’s a comparison between two different speeds.
So our goal, step by step, is that at the moment you show something the viewer hasn’t seen before, you need to explain it from a basic point of view to make it accessible and accepted in order to make non-bike fans like it, because then they’ve learned something by watching MotoGP.
MM: So, speaking of the high speed camera, how do you decide where to place it and can you change that easily? For example if you put it on a corner on Friday and it’s not as good as you thought, is that camera easy to pick up and move on Friday night so that it’s in a different spot on Saturday?
SS: Well, the high speed camera is a big camera, it looks like a studio camera so it’s heavy and it takes more than 2 people to move it up and down or left and right. It needs a platform and it has fiber connected to it, and at the other end it has a complex buffering system absorbing a huge amount of information. Because the level of memory that you need to buffer for one shot that instead of 50 frames per second is 1000 fps, it is a very large file. With the high speed camera you create very large files compared with other cameras.
In any case, this camera that we started using last year, is for us a wonderful tool to learn, to discover, to see and to show new things that we couldn’t see before. Our philosophy [for camera placement] is first of all let’s analyze the circuit before we go, and by analyze I mean we talk to different riders, we talk to different engineers, we analyze from our side, from the television point of view, in order to decide the best spots. We have a lot of information with the circuits because we’ve been there before (some circuits we’ve been going to for more than 10 years).
Usually the spot for the high speed camera is going to be directly proportional to the speed of the bike. Fast corners, ends of breaking points, but in fact we are learning circuit by circuit.
The difficulty is that sometimes we arrive, we plan the spots, and then when you see them after all of the work, you think “Ok, spot one was better, spot two was not what we expected.” But the important thing is that every time we go to a new circuit (by which I mean a circuit where we didn’t have the high-speed camera before) you learn a lot.
And also you discover the behavior of the camera, how it works, because if we put this camera on a slow corner, it’s going to take ages to see the picture. The concept of high speed is that one second of reality becomes 40 seconds. One second recorded at high speed means you will have 1000 samples of something that lasts one second. In one second you normally have 25 samples (as the normal speed is 25 frames per second) so your eyes see 25 samples in one second and your sensation of speed is normal.
If I put 1000 samples, obviously you need more time so this second becomes 40 seconds but the reality is you are watching an extremely low speed and this allows you to see things that you didn’t see before.
For instance, here in Laguna today, we’re having a big debate because we want to see on the first corner the bikes flying, with both tires in the air. The goal was, ok let’s go to the track, look for the position, so we were walking all around this straight. Why do we do this? How do you know the bike is flying? If you haven’t seen it, you can’t know... This is because some riders have said in the past: “Ah, at Laguna when the bike goes out of the first corner after the hill, for some time the bike doesn’t touch the ground.” We can only see this fraction of a second if you convert it into a long sequence of frames to slow down the reality. So if this fraction of a second can be viewed over 10 or 20 seconds, it means that for those 10 seconds you will see the bike flying.
This is the goal here. I hope that if it’s sunny this afternoon, and the pace is higher than it is now [Ed.: It was cold and foggy during this interview], we will increase the speed of the camera and maybe shoot at 2000 fps because then the time that the bike’s flying will be longer visually. So this is a goal here now, but it depends on the height of the camera, the heat haze making the tire look like it is low to the ground.
MM: I have the same problem with photography. Heat haze ruins many photographs.
SS: Exactly, sometimes you love the heat, sometimes you hate the it, because it’s showing you things that don’t exist in reality. The interesting thing is you don’t put the camera in position, make the shot, and move on. You have to analyze what you have done, and when you have done it you can then tell the people what they are seeing.
I know many people watch the TV and don’t see the whole information but we try to make it very easy. Today the shot of the bikes flying out of the corner, it will show the top speed on screen and we also tell the commentators to make sure that they understand which way the shot is going to go and when they talk they can explain. Or maybe we make a circle to highlight something.
MM: So when you decided to try to get that shot, you moved the high speed camera into a position where you think it will capture the bikes in the air.
SS: Yes, this weekend, the first day it was at the Corkscrew, today it is at the first corner and based on the results we’ll decide what to do with it Sunday. Maybe we’ll move it to another position.
MM: You must end up with more good footage than you can show.
SS: The other day in Germany, at the start of the race, we had a special shot of all of the bikes going through turn ten, all going down, that was a beautiful shot. It could’ve lasted for one minute! It’s like you’re at a museum looking at a painting and you could stay there and listen to classical music, the shot was that amazing. It was just a fraction of what we had. It was like, ‘I’ll show you something nice, but now I have to go!’ This is because we didn’t know what we might miss.
The way we direct the races, we know where somebody will overtake, we know that prior to the overtaking, because we have studied the overtaking positions, and they don’t overtake in crazy places. They overtake in the hot spots. But we are not Superman, all we can do is study the hotspots in order to show the viewers what’s happening.
Then, because of the weather, the tires or something else, something on that weekend changes and Marquez overtakes in a place that was not expected. In Austin, it was a new circuit, so… But it was not expected by the riders, either, because in Austin, we discussed it with the people who had been to Austin for the test. We asked them about the points where we needed to watch out for overtaking, they said ok, watch out here, here and here. Of course they didn’t expect the overtaking of Marquez.
That happened in our first year in Austin, it takes us three years to learn where overtaking is standard, and also for them, they ride and they fight each other, they discover the places where they can go for it.
MM: You said that you can put the high speed camera up to 2000 fps. When do you do this?
SS: You can see something at high speed from 200 fps and higher. The physical limit of the system today is 2500 fps. We used higher fps on a top speed position, for example in Mugello. In Mugello we have a top speed shot of Rossi going perpendicular to the camera to see how he looks, and that was 2500 fps because he was going 340 kph.
MM: You talked about the huge amount of data that the camera is generating because it’s shooting that many frames in HD. But your other cameras, your regular cameras, are always running, right? To make sure they catch as much as possible?
SS: We record all of them, yes.
MM: So the high speed camera, it’s not always going. You must record only when you want to capture something specific.
SS: Yes, you see the electronic picture on the monitor, but you only record to capture certain scenes because there is a limit on file size. One of the problems of the high speed camera is that it’s not that you record everything and then you look to see what is best. At the moment this is impossible because of the capacity.
This requires us to be very smart. The cameraman needs to be very smart, the replay people need to be very smart, and also the way we manage the camera must be smart. For example, we might say “Oh, this shot of Valentino is very good. Ah, it’s out of focus, let’s do it again, oh now he’s in the pits, wait for another lap, oh now it’s raining, we don’t have another chance…”
MM: That sounds like my life as a photographer!
SS: This is similar to you, I mean, you know you have chances and you need to take them. It’s like when you travel and you go to a shop, if you like something, buy it! Because tomorrow you don’t know if you’ll be back there. We’ve learned that when something is working you need to squeeze it because you won’t be given too many opportunities.
MM: So when you capture the high speed corner, the signal goes to a server somewhere, then someone sees the footage and decides to use it. How long does it take for you to say, “We got it, let’s put it in the feed?”
SS: It depends, there are 2 ways to do it. One way is you could take the risk, and while we’re loading it [from the buffer], you just put it on TV. But you don’t know what you’ll see. When [the camera operator] records a two or three second clip it goes by like a short video clip piece of noise, pffft! If you expand this noise, only then do you see if it’s out of focus, if the framing was good, and so on. But while he’s recording it, your intuition can tell you that it’s a good shot, but when you expand it is when you realize if it’s something incredible or not. But at that point, then we cannot change it.
As you one of the busiest people in the paddock, Mr. Sendra's time is in high demand and we thank him again for sparing some to talk to us. And again, thanks to Pol Bardolet for assisting with the scheduling of the interview.
Thanks also to PHOTO.GP intern, Kerry Port, for transcribing this interview.