Select Page

How Haivision SafeStreams Transport (SST) Transmits Live Video over Cellular and Internet

Last updated Jun 29, 2023 | Published on Nov 22, 2019 | Podcast, Featured, Bonded Cellular, Webinars


Jim Jachetta:                      Great. Okay, we’ll get started. Good morning, everyone. This is Jim Jachetta. I’m the CTO and co-founder of VidOvation Corporation. VidOvation, we’re the master distributor in the US for a company out of Rennes, France called Haivision.

Jim Jachetta:                      I’m sure we’ve talked to many of you about the Haivision technology. We’ve deployed Haivision with Turner Sports, CNN. Haivision technology has helped make shows like Live PD possible, and today we have a very special guest. Some of you guys and gals might be tired of hearing from me on webinars, so we’re very happy today to have Samuel Fleischhacker.

Jim Jachetta:                      Samuel has been with Haivision as a senior product manager for almost two years now. Samuel is a 20 year veteran in the broadcast television space, in coding video transport over IP. He’s worked for brands such as Phillips, Telos, Thompson, and Harmonic, and we’re really happy today to have Samuel tell us more about the special … we get asked every day, “What is the one thing that makes Haivision different? What is the big differentiator?”

Jim Jachetta:                      It’s a lot of different things. There’s a lot of uniqueness to their technology, to their software, to their antennas, to their modems, and their protocol, et cetera, but I think if we were to start with one thing, it’s the Safe Streams Transport technology, and Samuel is going to tell us more about that today, so thank you, Samuel.

Jim Jachetta:                      How are you?

Samuel F.:                           I’m fine, thank you, Jim, and welcome, everybody. I’m very pleased to present this first webinar session, which is part of a larger set of sessions organized by VidOvation and Haivision. In the next session, we will focus probably on the product more especially, or on some use cases, but today we will, as you mentioned, talk a lot about our technology which helps us to perform [inaudible 00:03:06] role and assists the transport of video over, and manage IP networks.

Samuel F.:                           This Emmy-awarded technology is named SST, which stands for Safe Stream Transport Protocol, and on Haivision, which is a French company created 11 years ago, we are today on the third generation of SST, so the SST V3, which is implemented in our latest generation of products.

Samuel F.:                           What is SST? SST is a technology, a patented technology, which will help you to transport videos, audios, and data wherever you are in the world, in the field, even if you are moving with your camera or with your stem, and to deliver these to your control rooms.

Samuel F.:                           The important point is that we are totally network agnostic, and that is a trend of our solutions, meaning that you can transport the videos over wifi, over traditional IP networks, LAN or WAN, as well as any type of satellite, so KA satellite, KU satellite, BGAN satellites, which have the benefit of having a different contractor, different usage.

Samuel F.:                           And finally, also, and this is the most use case we have: on cellular networks, 3G, 4G, as well as 5G networks, so we are totally agnostic. You can use any of these networks, as a standalone network, or as combined networks, as we will see later. We do not compromise on quality.

Samuel F.:                           With the standard technology, if you are transmitting video, you can see some issues like blurriness, blockiness, jerkiness, and with the safe stream technology, the promise is that the quality will be there, whatever occurs in the field. Even if you have a lack of connections, we will find a solution thanks to this technology to provide the picture to your control room, as demonstrated there.

Samuel F.:                           The markets we are targeting with this is of course, and first, news and broadcast, as well as live entertainment, like Live PD, I will present later. Sport coverage. I will focus on Turner Sports, and especially when that happened lasts year in France. Online media, education, and public safety, so all of that can be used and an be supported by SST.

Samuel F.:                           How does it work? What are the use cases? One of these use cases is SST’s contribution in the field, when you want to do it live, so this is typically you are wherever you want, in a city, on the countryside, you are in your car, you can be on your motorbike, or you can be fixed.

Samuel F.:                           Even though you can be with a drone, or with a helicopter, and you want to transmit videos to your control room, which is in Atlanta. With SST, you can do that. You can have multiple connections from multiple sources, which will all be transported to your control room.

Samuel F.:                           This is a live field contribution, but SST is going to do much more. Let’s imagine that you have affiliates in USA, but anywhere around the world, and you want to transmit them your programs because you have some contract with them. With SST, you will be able to transfer and to transmit all of this video to your affiliates.

Samuel F.:                           Then that commences the content distribution, so what’s a typical use case? Let me talk about French people. This would be shown between a lot of French speaking TV channels in Europe, located in Belgium, in Switzerland, in France, as well as in Africa, and in Canada, who are exchanging live all the time with this technology.

Samuel F.:                           So, this is a content distribution. You can do, also, remote or at-home production, as demonstrated here. Let’s imagine that you have an area with cameras. With our technology, you can, of course, deliver the video from this area to the control room, but you can also remotely control the cameras, so we could say that SST is a bidirectional assistant.

Samuel F.:                           SST, in a nutshell, is able to transport multi-types of content, so as you understood, it can be live. We are totally agnostic to the encoding formats, so in our solution we are performing traditional H264 AVC transplant encoding, as well as HEVC encoding, in any type of resolution.

Samuel F.:                           SD, HD, full HD, ultra HD, with the European frame rate, American frame rate; we are totally agnostic on that, okay? One type of application is file transmission. You have many applications where the reporters are making a shoot in the field, and we’re recording it on our unit, and after the day, when they’re back to the hotel, they want to transfer it to the end CR.

Samuel F.:                           They can do that with our solution. It can also transmit bi-directional data. Not only five. That is technically what is used. We kind of established a VPN between two networks to transfer the videos from, let’s say IP cameras for instance, or to chat, or to navigate on the web wherever you in the field.

Samuel F.:                           We have also the capability to manage intercom, also named ISB. Obviously, the person in the field will be able to chat and discuss with the person in the control room to get instructions and to give it feedback during a live. We have the adaptability, also, to provide video return, to give in the person which is in the field constant information about the program of a camera which is on the air.

Samuel F.:                           But it’s really to give teleprompting information during an interview, so this is a very interesting feature, and finally, as mentioned earlier, the remote control from the control room to the field devices. This is an example. It’s a PTZ, but it can be anything else. Multi-contents support. Video, files, data, and comments. That is all including in SST, and of course on our transmitter products and receive products.

Samuel F.:                           In a nutshell, SST which was awarded this year, with an Emmy award at NAB, it’s a transport protocol which will guarantee quality of services and security thanks to transmission robustness, packet reordering, intelligence and retransmission, error correction. I will describe it in the next slide.

Samuel F.:                           It also aggregates different networks, even if they are established networks. This is a very interesting feature, and it is adaptive, meaning that you can now guarantee on unmanaged IP networks the bandwidth, so you are adapting in real-time the video resolution, the video bit rate, according to the capabilities of the networks, and we are tracking, we are monitoring very shortly any change to make sure that your live will still be there.

Samuel F.:                           Let’s focus now on the feature of the day, network aggregation. Networks aggregation means that you can aggregate links and load balance the traffic between links which are relative to cellular networks, to wifi networks, to IP networks, and satellite networks. That is what we do typically and which is demonstrated there.


Samuel F.:                           We have solutions with many similar connections, up to eight, for instance, for one of our products. For the purpose of this slide, I would use this through three cellular networks and one wifi, so on the encoder, on the transmitter side, we will receive the stream from the camera.

Samuel F.:                           We will process it. We will then encode it, and we will spread the data of this video encoded stream among all of the streams, first, second, and third stream, and the wifi networks, and this has to maximize which is available. This is very useful because, for instance, you can move with your product in the field, and while two sim cards with one operator, let’s say AT&T, two other sim cards with T-Mobile, and two other ones with Sprint, and two other ones with Verizon.


Samuel F.:                           And wherever you are, you have a guarantee that all networks are there, so you optimize the bandwidth, but at some point, if one network is not there if you are on Verizon, it’s not a problem. You will simply not use the network which is missing, but you will do the other one to guarantee that you still deliver the stream, so this is a smart and intelligent system which is able to maximize the bandwidth thanks to this.

Samuel F.:                           We have, also, the capability with Safe Stream to say that some links are high priority networks, and other ones are low priority networks. Two examples of that. Let’s imagine that you have some contracts, special contracts, with two operators, so you have special prices, and with the other one, you have no special contract, so you can declare the link or the operator from which you are contracting as high priority, and the other one as low priority.

Samuel F.:                           This means that it would first try to use the high priority link, and if you do not succeed to reach the target bit rate that you have configured in the system, you will use the complimentary networks, the low priority networks. Another application that we have or we deployed is some people are using KA satellite as high priority network, and as backup systems, the cellular networks.


Samuel F.:                           You can do that with SST. SST, of course, is spreading the packet over all the networks, so everything is arriving not in the right order, so our technology guarantees that we will reorder packets correctly, so that is literally done by the SST. The second point is the automatic repeat request or ARQ.

Samuel F.:                           Let’s imagine that you are transferring, you’re transmitting one packet. Packet number one. This one is correctly received, so no problems, but the second one is not received, so you will still continue to deliver packets, packet three and so on, but at some point you will be able to say, “Oh, one packet is missing.”

Samuel F.:                           And you will do some retransmission, so within an intelligent system that does not retransmit all packets, but only the packets which are missing, and our system is reordering the packets to make sure that nothing is missing.

Samuel F.:                           On the complimentary corrections, we have FEC. Two-dimensional FEC. FEC stands for forward error correction, so let’s imagine, as illustrated there, that you have 16 video packets to transmit, but at some point one is missing, so you have two unconnected. The first one is to use the retransmission of this packet, but for some technology, you cannot use that, because the latency is so long that you cannot do that.

Samuel F.:                           For instance, with satellite, so in that case our system is able to add additional packets which are named protection data, as illustrated there, and this protection information which are relevant information will be able to rebuild the missing packet, so of course you are having additional packets, you are increasing the bandwidth of your systems, but you can recover your packets without retransmissions.

Samuel F.:                           That is very interesting, and what is interesting with our technology is that it is fully adaptive, so the size of the table which is there, 16 packets, is automatically adapting according to the quality of the network. If you have a very poor network, the size will be small, and if you have a good network, the size will be high because we lose not many packets, so that is what we do with SST.

Samuel F.:                           What’s very important, also, with SST, is that we are able to configure the encoders to say, “Hey, I want to target six megabits,” for instance, “with two seconds of latency.” We can do down to 500 milliseconds, okay? But let’s say two seconds, which is typically used by broadcasters. If you are configuring this multiple camera, in the field, you have a guarantee, with our system, that the actual data delivery of the video will be perfectly synchronized.

Samuel F.:                           You can do a frame accurate switching in your MTR, with three cameras in the field, with Haivision technology, and configured with the same parameters, so that is a very nice feature. We are supporting adaptive video resolution and bit rate, because when you are using unmanaged networks, for instance cellular networks, you are on your motorbike, you are moving from cells to cells, and you are losing some networks.

Samuel F.:                           Some networks aren’t appearing. You may have some congestion, so at some point, the available bandwidth, it’s not stable. It’s moving all the time. We are measuring that, and we are adapting accurately, many times per second, the bit rate of the video we are encoding.

Samuel F.:                           Of course, we try to reach, at any time, the target you have configured, but if we cannot we are viably declining or decreasing the bit rate, and at some point if the bit rate is too low we are changing internally the resolutions, dynamically, to make sure that the live is still there.

Samuel F.:                           Of course, on the reception part, we will upscale the resolution to make sure that from the starting point and the ending point the resolution stays stable, but in between, it’s varying according to the resolution, according to the bandwidth of the network, and again, because it’s an unstable network.

Samuel F.:                           If you have a very stable network, you don’t care about that. You can configure the system to say that you want to have a constant bit rate. We can do that, and you will have a stable bit rate and delivery, but other systems works with managing unmanaged networks. What are the benefits in a few words about SST?

Samuel F.:                           First, it is the reliability. You make sure that your system is working 24/7 every day, anywhere. The second point is the best quality of services thanks to how well they do corrections, thanks also to the retransmission of packets, and other topics I described earlier.

Samuel F.:                           You also have the benefit of the cost reduction, because you will see that, during the next sessions, that the solutions that we have are really small. We have backpack solution, we have pocket sized solution, we have applications running on smart phones, so it’s quite less expensive than the big system which tracks satellite and so on.

Samuel F.:                           We are totally agnostic. We can connect to public wifi, to private wifi, to any cellular network. The modem that we are using inside our platform, as well as the antennae which are embedded in our product, are worldwide compliant. For instance, we have French TV stations that come in Japan for the Rugby cups that were using the same products as the one they were using when the cathedral of Paris was burning.

Samuel F.:                           It is very easy to use, and you don’t care that, “Do I have to change something?” It works everywhere, so it is really very flexible, and it is mobile. You can move with it, we have some application helicopters, on motorbikes, on moving cars. You can do this very easily with this technology embedded in our solution.

Samuel F.:                           When we talk about our solutions, we have on one side this technology which is embedded on appliances, so transmitters, PRO3 series, AIR series, and MOJOPRO, as well as the backside, the pocket side, and the smartphone application I was mentioning, so this is real units.

Samuel F.:                           And we have the HE4000, which is an ultra HD, and multi-HD workable unit, multi-HD encoded, and the Stream Hub Transceiver, which is the unit MTR, which is receiving and re-aggregating the streams, so we have solutions of SST running on appliances, on hardware, but we have, also, SST running on the cloud.

Samuel F.:                           The Stream Hub can be installed in the cloud, typically on Amazon, we have some customers using their live like this, so we offer, really, an elasticity in your system. You can have your physical Stream Hub on your premises and if you want additional channels you can use an additional Stream Hub in the cloud.

Samuel F.:                           And we have, also, a version of SST which is Docker. Docker is a technology that can help you to include it directly inside your solutions, so you have some partner, some company performing cloud prediction, for instance, which are using our technology, meaning that our devices are directly connecting to their cloud [inaudible 00:24:01] thanks to this SST, in the Docker enabled version.

Samuel F.:                           Now let me explain you some use cases, because okay, technology is technology, but how is it used? I took some examples with American customers. Of course, we are working worldwide. We have customers in Asia, many in Europe, in Latin America, in the Middle East, in Russia, in Africa, and so on. Here, I just focus on the at-home production, distribution, and field live.

Samuel F.:                           With Turner Sports, CNN, and Live PD program. Last year, last September, an official event happened in Paris. It was the Ryder Cup, which is a golf tournament, opposing European golfer and American golfer, so it was happening in the suburb of Paris, and Turner Sports, based in Atlanta, was in charge to produce that. They had deployed cameras and crews in Paris, but all the production was performed in Atlanta remotely, and how did it work?

Samuel F.:                           Just like this. They had a truck. There’s a multiple HE4000 encoder, which is a multi-HD encoder, so one device is able to encode four HDs, so they were performing 16 HD 1080p encodings, and this was directly delivered over the Public Internet to Atlanta, to Turner Sports MCR, where they were decoding the streams and making the prediction with pictures coming from the field, but also people directly speaking in Atlanta, and they were returning, also, a multi of all the field signal as well as the program, from Atlanta to the truck, also using the SST technology with the HE4K encoder in Atlanta, and stream it in the car.

Samuel F.:                           It worked pretty well. It wasn’t at the event during three days, and Turner Sports used it again with them in Madrid for the final European soccer cup. Here you have some pictures of the event, with the case with cameras, with the truck where was the base at Turner, and the inside of the truck.

Samuel F.:                           But again, inside the truck, they were not performing any videos or predictions. They are just controlling the sound, they’re controlling the cameras. They were controlling the transmission. Everything was done in Atlanta. A good use case is CNN. We have developed for CNN a system that helps them to distribute some content to their affiliates around the world.

Samuel F.:                           Here, just to give you one of the examples, but we have some customers that subscribe to streams from CNN in Europe, in France, and Poland, and so on, so that is a principal, and for this CNN are using, in their headquarters, again, an HE4K encoder, which is performing HEVC encoding at 1080p/50. They are encoding four HD streams, and what is interesting is it’s a system running all the time, so it’s a 24/7 system.


Samuel F.:                           This system is distributing the videos to a Stream Hub which is inside the cloud, and this Stream Hub inside the cloud is in charge to distribute to the different affiliates the appropriate video. For instance, the affiliate number one subscribed to receive HD number one, and the Stream Hub in the cloud is in charge to distribute it to this one, and affiliate one have a porte where we can change and say, “Oh, now I want to receive HD number two.”

Samuel F.:                           And it will be automatically done, and with do the same for all the affiliates. Of course, both affiliates can receive all of the stream, but generally they don’t want to receive all of them down, so you stream by some of them, so this is the system that is running right now. The latest use case I wanted to focus on is Live PD, for which there’s a field system, for which the customer, while looking for reliable, low-latency and synchronous video and audio transmission on multiple camera and multi-location.

Samuel F.:                           You’ll probably know this program. It is a program where police persons are pursuing people in the streets who are inside their cars, and it is very popular in the US, so they are using our solutions for that. The car typically … they are broadcasting live from many cities simultaneously, but are maintaining the synchronization between all of the cameras. The lip-sync is perfect thanks to SST, and what is interesting is that they’re using our solution not only on the policemen which are working in the street but also inside the police vehicle, which is running at 130 miles per hour, so which is very fast.

Samuel F.:                           This is an example of photos of the system. There is multiple cameras inside the cars, which are performing live simultaneously, and in the car there is some remote antennas and transmission systems you have demonstrated there, and all is received in the controller room and produced in the control room on live.

Samuel F.:                           You have, here, some examples of use cases of our solution with the team in charge of this program, and the control room where all the streams from the Haivision solution are received, and used to build the final Live PD program. This was the first station, so if you want to have more information I give you, here, two contacts.


Samuel F.:                           Ronan Poullaouec, which is our chief technology officer. He’s one of the Haivision co-founders, but is the head of SST technology, did design it, is the right person which knows perfectly how it works, how to fine tune, and if you have any technical question, I ask you, but command you, to contact him, and I don’t introduce me again.

Samuel F.:                           I’m senior product manager by Haivision.


Jim Jachetta:                      So, Samuel we got a couple of good questions, and these are questions that come up quite a bit. Some of the folks are asking, you know, what is the difference between the Safe Streams Transport or the SST compared to transports like SRT or RIST or something similar to that? What are some of the differentiators? Maybe you can elaborate.

Samuel F.:                           Yes, and the differences are essentially for SST, for instance, the aggregation of multiple networks, which is today not supported. Another differentiator would be some technology that I did not mention earlier is that we are also able to do encryption of the streams, so AES encryption.

Samuel F.:                           SST can do that, but others cannot do that. It is also the, let’s say, the method of corrections which are implemented in the solutions. We have adaptive FEC correction, which is very nice, so just technically talking, differences. But as we are also a company which wants to be compliant with ecosystems, open ecosystems, our solution, SST, which is between our transmitter and our Stream Hub, can be converted at the outside of the Stream Hub to be then connected to another system with SST.

Samuel F.:                           Meaning that, for content distribution in the cloud, you can use SRT with Haivision solutions as a kind of gateway between SST and SRT.

Jim Jachetta:                      Well, Samuel, you know, some of the applications we’ve done, I think one of the critical linchpins is the multi-camera application. Take, for example, the Live PD show. That was one of the earlier projects that we did where there could be an incident where two police cars show up, and there’s two photographers in the back seat of the police car, there’s a camera on the officer, a POV shot of the officer behind the wheel. There’s a camera out the front dash, so when they’re arresting someone there could be as many as four to eight cameras in close proximity with more than eight sets of microphones in close proximity.

Jim Jachetta:                      And in a live production, there’s no time to fix lip sync problems or fix cameras that are out of sync, so I kind of tease the Haivision guys that sometimes you guys are a little bit too modest about how great your technology is, so that’s my job, is to brag about Haivision. A lot of people out there claim to be able to do at home production, Remy production, whatever it is you want to call it. You can do it on a managed network, you know? You can do precision timing, protocol, have your cameras all in sync.

Jim Jachetta:                      The real challenge is maintaining gen lock and lip sync down to the frame. All the cameras are on the same frame, they’re all accurate. They’re not plus or minus two, three, four, five frames. They really have this figured out, and it’s a testament to this SST technology. Then, fast forward to the Turner Sports project. That was even more critical. This was a higher level broadcast application, and they brought 16 cameras from France, I don’t know, 5,000 miles back to Atlanta through a single public internet connection.

Jim Jachetta:                      They didn’t even have redundancy in the connection. They used a single connection. We didn’t drop a single packet during the multi-day event, and the cameras were all in perfect sync when they came to Atlanta, and then other little things like I remember two days before the event Tom Sahara at Turner Sports called me up and he was like, “Oh my gosh, Jim, I just realized we’re using a European crew to shoot the Ryder Cup and they’re shooting in 50 hertz. I’m going to have to get standards converters for the 16 cameras coming in because we’re in the US, obviously we’re at 59.94.”

Jim Jachetta:                      And the Haivision product is so great, you pick the output standard, the output resolution you want, so we captured the content at 50 hertz and we were able to output 59.94 very easily. Maybe you can speak to more of that, Samuel? Or … ?

Samuel F.:                           Yeah, yeah, so maybe it will be much more focused on product than that, but we can say that we have outside the SST technology and especially on your computer’s side, what many things that aren’t everything, but on the twin up side, this is really a box that is offering a lot of technology where you can receive SST, of course, but also IP streams, traditional IP streams. You can do transcoding inside it. You can create a multi-view on it, so you can aggregate all the different videos coming from the stream to create a mosaic and to provide it on the big screen.

Samuel F.:                           You can redefine and automatically perform SST transmission externally, so really it’s a multi-usage box, and that is really what is appreciated by our customers, and the second point which is really appreciated by our customers is that this box is your box. I mean, when you are a broadcaster, this box can be a server in your premises.

Samuel F.:                           It can be your server, with our software in your premises, or it can be in the cloud, but at some point, you have control of it. You don’t go through, let’s say, a management system under the control of Haivision which can decide or not decide to let you run correctly or to change the perimeters. You have full control of your systems, and that is what is really appreciated by our customers compared to all the competitors. It is really that you’re changing something in the field is really easy to do, and you can do it. You are not obliged to ask the authorization to Haivision to do that. You can do that.

Jim Jachetta:                      Yeah, so some of our customers, you know, like the Live PD show, sometimes there’s gunplay. If the photographer needs or wants to take control of the Haivision, they can, but usually someone kind of in a transmission engineering role in master control is monitoring the links. Obviously, the links will go down when they have to do a battery change, but the technology is such that when the power comes back on the battery is reconnected, the stream just restarts, so as Samuel said you have that flexibility where the photog in the field can take control.

Jim Jachetta:                      Or, if the photog is less experienced, someone in master control can orchestrate and take control of everything, and just watch the Live PD show. It’s not everyone’s cup of tea. It doesn’t catch people at their finest moments. I hope I never get pulled over in my car and Live PD cameras are there to take my picture, but it’s an exciting show. You know, a lot of this technology was invented for news gathering where a reporter would go out and do a 30 second standup bit on the courthouse steps, or they’re in a press room covering a press conference.

Jim Jachetta:                      Live PD, sports, really shows the robustness. It’s not insignificant what Haivision is doing here. It’s quit amazing, so I would like to do one of these webinars maybe once a month. I know we have the holidays coming up next month, so if we don’t do one in December I think we will continue the series for sure in January. We have a lot of new things coming from Haivision, and maybe we’ll hear from Samuel again, maybe we’ll hear from Ronan, or both of them.

Jim Jachetta:                      And thank you so much, Samuel, for joining us today. I really appreciate it. Oh, and by the way, these slides, these great images, Samuel made these all himself, so thank you so much Samuel for all your hard work. We really appreciate it, and those of you who have tuned in, we will transcribe today’s sessions and put the slides and the video you online in a couple of days, so if you missed it, or if you want to re-watch or re-listen to this event, I’ll also publish it on our podcast channel.

Jim Jachetta:                      Thank you so much, folks, and you have the contact information here if you have any other technical questions, you want to do a demo, you want to reach out to VidOvation or Haivision, please do, and thank you so much.

Samuel F.:                           Thank you.

Jim Jachetta:                      Thanks, Samuel. Bye.


Continue Reading