CJ Interview: Qube Cinema’s Rajesh Ramachandran On The Past, Present and Future of Qube Wire
Published in "celluloidjunkie.com"
One of the products and services that made a big splash at CineEurope in 2017 was Qube Wire, which was just being launched commercially by its developer, Qube Cinema. Some of the buzz around Qube Wire surely had to do with curiosity; both over what the new offering was all about, and why a company like Qube, which had built its reputation as a digital cinema hardware manufacturer, would build a platform for global theatrical content distribution.
Like many, we wanted to learn more about Qube Wire. Since last year’s CineEurope we have been given demonstrations of the platform and its extensive features which are geared toward both distributors looking to manage their content assets, including digital cinema packages (DCPs), key delivery messages (KDMs), and rights management, as well as exhibitors who wish to receive and manage digital cinema content, KDMs and device information. (Full disclosure, I’m a partner at Covergent which currently has a commercial relationship with Qube Cinema).
During this same time, Qube Cinema has grown Qube Wire to a point where it is now servicing 127 countries. What this means, as we soon discovered, was that Qube began assisting distributors, independent producers and studios in delivering their theatrical content to cinemas in territories all over the globe. Qube has also been working with exhibitors around the world to update the digital cinema device information in the Qube Wire database, which in turn provides the cinema operator with a publishable facility list message (FLM), more specifically; it provides an FLMx feed, which is the industry-standard protocol for exchanging theatre information. FLMx is based on SMPTE ST 430-7.
If all this talk about KDMs and FLMx sounds confusing, that’s precisely why Qube Cinema created Qube Wire. But part of the problem in trying to convey precisely what Qube Wire is and who exactly should be using it, is that when it comes to digital cinema, it covers such a broad spectrum of use cases. According to Rajesh Ramachandran, the President and CTO of Qube Cinema, Inc., part of the reason for developing Qube Wire, was so that its users wouldn’t have to worry about three-letter acronyms, or how content is getting delivered or whether keys will be issued for the right devices. Ramachandran and the team at Qube looked at the major issues surrounding the adoption of digital cinema distribution and playback and came up with Qube Wire; a system designed to overcome them.
We thought it would be a good idea to speak with Ramachandran in the weeks leading up to CineEurope to get an update on Qube Wire, a year after its launch. We intended to learn how the service has changed in the interim, what we can expect to see at this year’s show and in the near future. As you will see from the following interview, what we got was not only a full explanation of Qube Wire, but a short dive into the history of digital cinema development which lead to the need for such a platform.
For years Qube was known outside of India, its home territory, as a hardware manufacturer. Are you finding it difficult to change the perception the market has of Qube from a manufacturer into a service provider?
A couple of years ago, we changed the name of our parent company in India which provided both hardware and services, from Real Image to Qube Cinema for that very reason. We’re now one brand that provides both products and services globally. And as we add more services that we have in the pipeline, this will become clearer to everyone.
What made Qube decide take some of the content and exhibitor services that you provide in India to a global market?
We need to look back at how and why we developed a digital cinema platform in the beginning. Yes, it was clear that we were going to be making servers and products for exhibitors, but back then, there wasn’t a distribution system or mastering system available at all. Everyone had their own closed systems at the start of digital cinema. So right from the day that we had our server, we had to have a mastering system and a distribution service because there was nobody else that was providing that.
So, if you look at the three areas of focus that we had – mastering, distribution and products – they are all kind of in lock step with each other. With distribution, in addition to creating the products, we also had to have a services group. Even with exhibition, we grew services – the VPF integration work that we do. If you look at our evolution and our growth and these three areas, we’ve essentially been a distribution company from the day we’ve been a digital cinema company. We work on about 2,000 titles every year. Although we may not be as popular as Deluxe Technicolor, we clearly handle far more titles than they have just because of how prolific the Indian film industry is. A portion of those titles get an international release and we have been facilitating international distribution as well. In fact, when we started with Landmark Theatres back in 2005, we had deployed Qube systems to the theatres and the DCPs were made using QubeMaster Pro. The point is, we’ve been in the distribution space from very early on. As the content and the Landmark network grew, we actually had a mastering and distribution group working out of Los Angeles. It was a smaller operation in LA but the India operation grew bigger and bigger and now has a few hundred people working in the mastering services group now. That’s not the world’s perception however.
Would you say that the international market, outside of India, isn’t fully aware of the full scope of what Qube does as a company?
We didn’t actively communicate or create the right perception of what we do in India. In India, we are seen as an all-encompassing provider. In India, people understand that Qube is a major force, although the actual services that Qube provides might not be very clear, people understand that digital cinema happens and continues to happen in India because of Qube. You are right that the international perception has always been that we’re a server manufacturer, even though post facilities have always known about our QubeMaster platform, which we started commercially deploying in 2004. There is definitely a mismatch in perception. Our VPF business has been limited to India. We don’t do VPF integration outside of India. So, outside of India, people don’t see that side of Qube at all.
That’s kind of important though, because as a VPF provider you are forced to have a full suite of systems and services in place. Obviously Qube has a server and a TMS, but does your system handle accounting, back office, equipment monitoring, key management and content storage?
Yes definitely. Even across all of our VPF deployments we don’t outsource the technical support. All of that is done by Qube employees. There are multiple touch points when it comes to India. In terms of how entrenched and how recognized we are in India, the global perception does not match that at all. We’ve always been seen as an innovative technology company, but then there was not the footprint to be seen as a major force.
Well that’s a great segue into Qube Wire. Would you say the perception of what Qube is and what it does now is changing with Qube Wire?
Absolutely. When we started off with Qube Wire, even other server manufacturers, everyone was suspicious about what we were up to because they saw us as a server manufacturer. We had to explain that we are actually the largest mastering and distribution service, in terms of number of titles, the number of theatres might not be as much as a Hollywood title because the Indian titles go mostly to theatres in India, the Middle East, Europe & North America but the numbers are not close to an “Avengers” or “Harry Potter” release. So when I say we’re the largest mastering facility, I’m talking really the number of titles we work on and not necessarily the number of theatres we’re distributing to. When it came to rolling out Qube Wire, this is pretty much what we’ve been telling people. We’ve been sending Indian content to many countries in the world and we also have the experience to quickly turn around a title because of the very last-minute nature of the Indian film industry. So our systems are built at industrial scale so that we can accommodate these last minute workflows. And our systems need to be incredibly reliable because there’s simply no chance to correct any missteps when you’re delivering content so, so close to release date.
How were you distributing content without Qube Wire, or were you using a similar solution internally and have begun commercializing it over the past year?
It wasn’t called Qube Wire, but for instance in the United States, we would sub-contract it out to Deluxe Technicolor and we had partners like that in various territories and they would handle the distribution. In some cases, the distributors would have people fly out of India with dozens of drives and then they would just use local couriers. So about four years ago we were just looking at how distribution was happening with Indian content, but also how distribution was happening with Hollywood content. We saw that it was always a hybrid, or multi-modal approach. But the one problem with everybody’s approach, though they were all using partners in-territory, was all the details that the customer, in most cases a distributor, had to worry about. When we look at how logistics is done in other areas, there is a level of transparency and the granularity of information that is available to the customer is available… we were not having that. So, we would hand off a 200 or 300 theatre job to a content service provider somewhere in the world and for the next 48-hours, all we knew was that they were working on the job. We had no visibility into where they were in the process. If all went well, that was not a problem, but if they had an issue or there were issues outside of their control, these were things that would hit us at the last minute. Our customers didn’t know either. Suddenly we would get a bunch of tracking numbers, sometimes after the content had been delivered to the theatre. Clearly, that was not ideal.
Today when people order something from Amazon, you go to amazon.com and you can get the status of your order every step of the way. If in a B-to-C world you can provide that kind of transparency, then why not in a B-to-B world like digital cinema distribution?
Add to all of this, the fact that it was multi-modal with sneakernet, there was broadband delivery and satellite delivery… it was too much for a distributor to have to worry about all of that. We wanted to create a system that allowed the distributors to not have to worry about how content was actually getting delivered or whether it would actually get to the cinema.
Well, that was always the great promise of digital cinema – that distributors and exhibitors would have a real-time look at where content was in the delivery and playback process. The same is true for being able to monitor playback equipment. Why has digital cinema never lived up to that original promise?
Well, if you look at the motivations for Qube Wire, one is that it was too much of a black box of what was happening with the processes on one side, but then the distributor was exposed to too much of the nitty gritty about the modes of delivery and who was going to deliver their content in each territory. So, if you step back and look at the DCI spec, the clever thing about it is that it allowed for a disconnected architecture. There is nothing in the DCI spec that required online connectivity. It doesn’t even define what a TMS should be. How content and keys were dealt with at theatres wasn’t handled at all. It would have taken far too long to come up with a standard for that because there are just too many options that are available. Plus, the logistics and communications infrastructure are not at the same levels globally.
And if you look at the bane of digital cinema, even today, it’s keys. You’ll very often hear about shows lost because theatres were not getting the keys on time or they were getting the wrong keys. And the key file is so small that even an extremely low bandwidth connection could work, but because it was never addressed by the DCI spec and because there was no requirement on the in-theatre systems or the in-theatre networks to receive and provide status on keys we’re still operating blind on that. Of course, a bunch of us are trying to change that. Things like the automated key delivery and receipts, there are a few flavours of that being tried out.
There are still ways to go for the industry in that area. The standard incorporates digital certificates quite nicely but completely drops the ball on how the certificates are created and authorised and how they are protected. I mean the private key that a manufacturer uses to sign their certificates is equally if not more important, and there is actually no coverage in the spec about that aspect. What is the point protecting private keys using military level security on the server when somebody could be creating keys on their unprotected desktop in their office!
Maybe not big Hollywood studios, but do you think some distributors find it cumbersome to deal with encrypted content?
This is something we were trying to handle with Qube Wire. We wanted to make it easy for distributors to use encryption. In fact, other than maybe trailers, everything should ideally be encrypted by default. It should not even be a question. There is this silly notion in the industry that we should start off with an unencrypted DCP so that we can give it to people hassle-free, but then, when you’ve created a big buzz at Sundance with your movie, for example, it’s moot to then go create an encrypted DCP when there is an unencrypted DCP in the wild. Encryption is an all-or-nothing proposition. You either start with an encrypted DCP or you don’t.
But you’re right, you couldn’t fault the distributors because they were finding it extremely painful to have to deal with KDMs on one side and they also felt that they were locked out of their own content. You couldn’t just give them all the keys in an email and tell them, “keep that email safe”. You have to take a holistic view on this. You don’t have to burden the distributor or the content owner who doesn’t understand all of this. They want to create good movies and get wide distribution. That’s their problem. Security should just happen for them. It should not be a pain. That’s what we’re addressing with Qube Wire. It’s a system that always keeps its keys in a FIPS 140-2 bundle. We keep the keys more securely than the DCI server does. The keys are maintained in our private co-location facilities, but the studio, distributor or content owner can instead keep the keys on their own premises under their control using a solution called Wire Safe. That way, the keys can never leave a distributor’s facility.
So just to be clear, you’re saying Qube Wire allows for content owners to encrypt content from the start because it helps manage the rights, key delivery and key storage, correct?
We are saying that filmmakers, particularly the smaller filmmakers, didn’t want to bother with the encryption because it was such a pain, so we wanted to make it really easy for them to keep their DCPs encrypted right from the get-go and also provide them top-of-the line security for their keys. We’re not doing a smoke and mirrors key storage and key management system. We wanted to make it easy on one side and use the highest-level security on the other side.
Qube Wire also helps content owners not get locked out of their own DCPs. They can not only create an encrypted DCP, but manage their keys securely.
But to make keys you need a trusted device list (TDL). You need to know what equipment the content will be playing back on and where that equipment is; which theatre it is located in. TDL’s somehow became a proprietary to their service providers. Do you think that’s because DCI didn’t specify a TDL solution?
You’ve pinpointed the monster in what the industry has created with this TDL. This is actually a really important issue. If you look at the first 13 years of digital cinema and solving this KDM problem, we were looking at how we get information from the theatre. There was FLM which was created by SMPTE. The spirit of it was always good but the first implementation was severely lacking. Then some adhoc extensions were added and we wound up with FLM-x. So now we have the facility list message that has all of the device information for the theatre and the FLMx is the exchange protocol. So at Qube we asked, why don’t we turn the TDL problem on its head. Instead of the industry trying to race to this one universal TDL, why don’t we say there can be many TDLs but as soon as there is a standardized method to communicate them, then they can cross check each other. We were all fighting for the wrong goal or the wrong target. It was not really about having a single TDL; it was about having a single FLM. We were so focused on a single TDL that we got blindsided and had theatres information being sent out in many different ways. Every digital cinema distribution services provider, like Deluxe, Arts Alliance and Qube and the distributors would all send out Excel spreadsheets and dealing with multiple phone calls to handle device information. But if you simplify the problem down to one source for a theatre, so that a theatre can provide their information via FLMx and then provide the FLMx via one feed, you’ve solved a big problem.
Are you allowing Qube Wire users to maintain their device information inside the system?
What we are giving cinemas is a user interface for them to keep their device information current. So that if they change a server from an auditorium, they can change their device within the system. When they get setup in Qube Wire, they are automatically given an FLMx feed that they are free to publish. We actually ask exhibitors to send their FLMx feed URLs to a number of parties that are collecting information. The idea is that they do that once and then all they have to do is make sure the device information on Qube Wire is correct.
Can updating the device information in Qube Wire be automated from each cinema location provided the right equipment is in place?
Not yet. That really ties into how a theatre network is setup. There are tools like Cielo, as an example, that use heuristics to determine whether devices behind a certain IP address has changed and they can update their systems. We’re beginning to tie into these systems to automate the process.
With Qube Wire, exhibitors can use a system like that or they can update their information online manually as their devices change. Today the problem is an exhibitor will change equipment and only send an updated Excel sheet to a few service providers. Then it took a bad key for the exhibitor to send the updated information to smaller content service providers. Now with Qube Wire they can simply distribute the URL for their FLMx feed which will be updated automatically when they update the info in the system.
So Qube Wire can store the device information but how does it solve the problem of trust? If I’m a distributor, how do I know to trust the information in the system was updated by an authorised cinema operator?
All that theatre owners can provide is their device information and the trust is something that the content owner has to choose. When trust itself can be subjective, how can you have one trusted device list? Because then we’d be assuming that every distributor trusts information in the same way. Each distributor has different relationships with each exhibitor, so the notion of trust could be different. The kind of steady state that we all seem to be going towards is that there will be multiple TDLs and all that we can expect from theatres is a good FLM. That shift is what we are now trying to promote within the industry. Trying to get people to see what we should be after, not after one single TDL.
Instead of having a single source of truth as a TDL, we say you have a single source of truth which is the FLM that a theatre provides. Multiple parties can go about aggregating this information and they might all arrive at the same conclusion, but then they all have flexibility in how they choose to trust. But the key really is that it’s a single source of truth on the theatre side and not a single source of truth on some digital service provider that says their TDL is the best TDL.
So then if you look at how theatres become the single source of truth, you have to give them the tools to publish an FLMx. The big integrators like DCIP have created their own FLMx solutions which have been used for a long time, but the smaller companies don’t have that kind of IT support or it’s just expensive for them to maintain a system to do this. That’s what Qube Wire Cinemas is doing. We clearly have a vested interest in getting good data, because then we’re able to make good keys through Qube Wire. That’s why we keep Qube Wire Cinemas free so that exhibitors have an easy way to maintain their information and publish their information.
What is good here is that we don’t limit who can consume that information. We encourage the cinemas on Qube Wire to publish their information to any party that asks for device information using the FLM-x feed that we create for them.
We’ve talked a lot about how device information gets updated in Qube Wire and is provided to third-party providers, but this is all in service of getting a KDM for a piece of content. How does Qube Wire handle keys?
Today, any cinema on Qube Wire can ask for their keys to be delivered to firstname.lastname@example.org. The keys come into the main Qube Wire dropbox and from there we can look at the device that key is targeting and from there we know exactly which theatre it needs to go to. The beauty of the DCI spec is that Qube Wire could even put the key in a place where every theatre could access it! The KDM delivery mechanism doesn’t need security because the KDM is secure by being tied to the device. But what we do is that when a theatre signs into Qube Wire, they have their own Universal Inbox and they can see all the keys that have been sent to them and all of the content that’s coming their way or will be arriving shortly.
You just mentioned content which is how Qube Wire was first introduced, as a global digital cinema content delivery platform. So now that we’ve discussed devices and touched on keys, how does Qube Wire handle content, both for the exhibitor and distributor?
The unique challenge with the Indian film industry is that it’s prolific and it’s last minute. We’re dealing with content distributors who are constantly calling us for status updates. So, with Qube Wire, exhibitors also have a user interface and they can see the same level of information the distributor has, and they can see where their content is. They know when the content will arrive. When the keys will be at the cinema. That’s the piece Qube Wire gives exhibitors visibility into. Otherwise, the exhibitors operate blind. When they got a hard drive delivered was when they knew they had the content.
How does Qube Wire deliver content? How was Qube delivering it before Qube Wire?
In India, it was a combination of hard delivery and satellite on the Hughes VSAT network using our multicasting system called QubeCast. Or we can use Kencast, or whoever. And of course, we can do broadband where it’s available.
And you’re doing this with partners in different territories?
Yes. By using partners. That’s an important aspect of Qube Wire. We’ve been enabling partners to work with Qube Wire, some of whom have not been in digital cinema previously, as long as they have a good relationship with logistics providers. What they have to do with the drive duplication itself is really simple and completely automated. In Singapore, we enabled a company with IT services to enter the business.
How did you do that?
We gave them the equipment, what we call the Qube Wire Partner Appliance. That sits in their facility; they buy hard drives and duplicate them. The key is that for them to duplicate hard drives, they don’t need to know anything about DCPs. The system tells them to push buttons after drives are inserted and then tells them when the drives are ready to be ejected. They don’t need to look at the contents of the drive at all.
Do exhibitors need to log into Qube Wire to get all the information about their content and keys or can they be delivered directly to a TMS or server?
TMS integrations are currently being worked on. Email integrations are already in place. You can think of the Universal Inbox as a dashboard. You can always log into Qube Wire and see all of your information, but you don’t have to log into the system once it’s set up to be automated.
Well, after last year’s CineEurope everyone was asking what Qube Wire was. Do you think the industry has figured it out by now? And what is the future of Qube Wire?
I think with the Indian film industry, people have figured it out. If you look at the future of Qube Wire, in the near term, we’re ramping up with our direct to theatre delivery using the Qube Wire Theatre appliance which you saw at CinemaCon, deploying them into theatres. The first deployments are going to start in a matter of days in the United States. The other focus is getting non-Indian content on Qube Wire. The other thing we’ve been very careful about is for people to make the wrong assumption that Qube Wire is only for Indian content. The one perception we’re trying to fight really hard is that Qube Wire is just for India or Indian content. The way we look at it is that Indian content is trial by fire, because if we can make Indian content happen, we can make any content happen, just because how organized and systematic the other film industries are comparatively.
What are some of your near or long-term goals for Qube Wire?
Our goal is to expand into all the other film industries, some of whom we’ve already begun to work with, but to do so more broadly. We can even work with major studios and distributors in parts of the world that might be harder to get to, where the distribution infrastructure may not already be in place. We feel the true promise of Qube Wire is when a Korean language movie is playing in Argentina, which otherwise would not have played because there was no visibility and the logistics were too complicated. Catering to Iranian cinema, Korean cinema, Hong Kong cinema… some of the Eastern European countries that have a cinema culture, along with Western Europe. So, the approach is not to go for a fewer number of movies going to a large number of theatres. Instead, we’d like a larger number of movies going to a smaller number of theatres in many territories so that we expand the range of content that is available. Because the theoretical footprint of theatres is limited, if you look at any expansion our industry can have, it’s by content.