The run time is 30 minutes.
The full transcript is below:
Lee: Hey, happy Monday, how are you doing Joyce and Jan?
Joyce: Pretty good.
Jan: Pretty good.
Lee: Okay, that's fantastic. We're going to jump in here and we're going to speak about what Global IP Solutions has been doing, but before we begin; GIPS, as you're more commonly known, has been a sponsor of the eComm event previously. For the event this month GIPS is a Platinum sponsor. First of all, I would like to get a handle on some of the reasons why GIPS decided to sponsor again?
Joyce: Certainly, surely I think there are a lot of conferences out there, but eComm is one of the more unique conferences that has a lot more substantive content. I think we find that the audience is a great mix of strategic thinkers, innovators, and entrepreneurs in the relevant space. We find it very refreshing from most of the usual conferences to where it's pretty level sales pitch. I think we've found eComm is definitely more long-term strategies and new innovations which obviously, we want to be at the forefront of.
Lee: I appreciate the kind words. In talking of the forefront of communications, GIPS has been really generating quite a lot of noise around what's getting called HD Voice and wideband codecs. I've even heard talk of the standardization of a new codec. I would like to know where GIPS fits into all of this.
Jan: That's a good question. One of the things I would like to start with is more to clarify what is a codec and what isn't a codec. I think there is a lot of confusion for people who are not really digging into the dirt there.
A codec is obviously an important part, but it's not everything that makes the HD Voice experience. If you look at the analogy of a car, you have a lot of different things like tires, body, steering wheel, etc. and the codec - let's say that's the tires. It's one very important part that makes the whole thing run smoothly, but without all the other pieces, you don't get the full experience.
That said, we're very much focused on HD Voice, and the things we do, in addition to codecs, is things like jitter buffer and packet loss concealment. Those are things to address what happens on the network. We also address things that handles on the device, like sound card. For video, it's cameras, microphone adjustments, speaker adjustments; all those things are necessary for the great experience.
Maybe, the most obvious thing for the end user is taking care of issues like echo, background noise, and the level of the signal, the automatic gain control. Those parts address what happens at the end points of the users. All these parts are important to deliver this full HD Voice experience. That's what we focus on. Codecs, we do that, and we'll get back to that I hope, if I get a chance, but also for us it's really important to look at the whole picture where you get the total HD voice experience by addressing all the other issues.
Lee: Okay, I just want to technically clarify; you spoke about packet loss concealment, jitter buffering, and so on. I thought I knew something about codecs, certainly, and had assumed that smart codecs like the one that GIPS used to use to power Skype, the iSAC codec. I thought it had packet loss concealment and jitter buffer handling and so on, built in.
Jan: It's not normally part of the codecs. Any codec, be it proprietary or a standard codec, doesn't include a jitter buffer. Jitter buffer, as you say, is extremely important for the performance over a packet network. Packet loss concealment is to some extent included, in terms that many codecs have something to handle packet loss, but it doesn't have the logics to apply it in the right time, and when to do it, basically, so the codec is very pure, just compress the signal or decompress the signal, but doesn't do any of these things that effectively handle the issues in the network.
Lee: Okay, so if I used to take say, an HD codec that maybe is in the public domain, the SILK codec, would I get jitter buffer and packet loss concealment etc with that codec?
Jan: No, all the standard codecs, the same with SILK, it does not include the jitter buffer. It will have parts of the packet loss concealment in the codec, but not all the pieces that you need to put it together. Again, of course, on top of that, the other things I mentioned you need in terms of echo cancelation, noise suppression, gain control, and put the whole thing together into a high quality solution.
Lee: When it comes to putting these things together, you spoke about handling devices like sound cards, microphones, cameras, and so on. Then you're saying that some of the packet loss concealment and jitter buffer technology is outside the codec anyway. Is it not possible for companies to assemble these themselves, without using GIPS?
Jan: Oh absolutely, and people do. It's just a matter that to do this really well, you need a certain expertise and you need a long experience of doing this. That's what we bring to the table, so we can do it at the highest possible quality, and you get the benefits of time to market. Our customers don't have to deal with this. They can get it done immediately. If you do it internally, it takes a long time, and most people will actually not succeed well enough to get to a quality level that's close to what we can offer.
Lee: Can I hassle you there and say we don't get to quality level you can offer. I think what I've gained from speaking to you in the past is GIPS is really a research and development company and the R&D to integrate the things that you speak of costs a lot of money. What GIPS is doing is basically allowing companies to minimize their overhead in R&D and I guess there is some kind of business model where companies pay a royalty fee to you but you're one centralized R&D in assembling these for devices, networks, or whoever. Is this correct?
Jan: That's a great summary of how it works. People come to us basically, and by integrating our software, they integrate our expertise into their engineering team. They get all of that, but they don't have to pay for all of it because we do it for many. They can get the same benefits more quickly, without having to invest as much money. So, in the end, it's a cheaper way to achieve the results than to try to do it yourself.
Joyce: It's also future proofing, in that sense, where we focus our specific R&D in those areas, where as new technologies and innovations come forth, we can provide that as sort of an extension of that specific R&D capability.
Lee: Can you give an example of new innovation? Again, I have a habit of hassling a little bit to get where the meat is. What kind of new innovations have come along, that you will then develop for and pass on to GIPS customers?
Joyce: Certainly, we're talking about audio or voice right now, but video would be a great area where innovations are coming about, that we can provide as they continue to develop. We'll talk about video capabilities and video codecs, such as the H.264 SVC standard. That is something that is up-and-coming, not fully taken off in the market, yet, but all our customers will have the advantage of having that expertise incorporated into their products, should they wish to do so.
Lee: You mentioned devices, so I know at the last event you spoke about building for the iPhone [Slides | Video]. Which devices do you cover, generally speaking?
Jan: Well, we pretty much cover anything from DSP-based solutions all the way up to PC solutions. We may be mostly known for the PC cell phone type of applications, video phone on the PC, but we also do a lot on embedded devices like the iPhone we mentioned; we have several customers offering VoIP over the iPhone, or Windows Mobile, or Symbian devices. We are looking at doing something for Android very soon here; that's one of the most current development projects that's ongoing. It pretty much spans the whole spectrum of devices. Our focus is to build something that's that adaptive, that you can put on any type of platform and you can get the same quality, same integration for our customers; they don't have to work with different types of integrations for different platforms.
Lee: When it comes to SILK, which was announced at the last eComm event in San Francisco, we've got an HD codec there. A lot of people, including myself I must admit, were confused as to what that means for GIPS. Since I thought GIPS is the "codec company," which had previously powered Skype, I wanted to know does this mean that it affects GIPS in some fashion? Can you pass some kind of comment on SILK? I know people will be asking.
Jan: I think it's a very relevant question. I think what we've already talked about here might explain part of that. We do so much more than a codec, and in fact, SILK is a very attractive codec for our customers to use, in conjunction with the rest of what we do. Again, we look at SILK as the wheels or the tires of the car and we provide the rest of it to put the whole car together. That means that SILK, standard codecs like G72.1 AMR [G.721] wideband, you name it; they all fit into our solution.
Lee: And GIPS also offered their own codecs, as well, I understand.
Jan: That's correct, and in particular, the iSAC codec, we're very well known for that and it's been deployed in hundreds of millions of end points.
Lee: That's the GIPS wideband codec, right?
Jan: That's our proprietary codec, which a lot of our customers use, but a lot of our customers also use standard codecs for interoperability, or can use SILK for wherever that's appropriate.
Lee: The iSAC codec is a wideband codec. Is there an evolution of iSAC towards, say, ultra wideband for example?
Jan: Going back to we don't stop our development and this is what you get when you work with GIPS. You get something that continuously evolves and gets better and better. When it comes to the ultra wideband or super wideband, whatever you may call it, that has until recently not been a big thing. Now when it is, we of course offer that as well. So iSAC now has a super wideband mode. We'll add other features over time, like series support, etc, anything that makes sense in terms of addressing the needs of the market.
Lee: Okay, so what you're doing is knocking down time to market for companies and you're also knocking down the cost of providing voice - audio, I should say, to be more general, and video solutions for devices, software, and so forth. It's about time to market, it's about saving money. This is the focus of GIPS?
Jan: In a couple of sentences, that's a very good summary of what we do, yes.
Lee: Okay, so that R&D, because I must admit for me, when I looked around the website I wasn't 100% clear. That's why it's really good to do this interview. Especially since GIPS is a sponsor, I really like to know what it is that GIPS is doing and get in a nutshell the sort of value. So it's research and development. If GIPS is an R&D company, then can you give me some idea of what's new? What's in the pipeline?
Joyce: Sure, I think we mentioned it a little bit, where we've been dealing with audio and voice over IP for quite a while. What's relatively newer for us is on the desktop video conferencing side. I think as you know, in the industry, real time video, tele-presence, video conferencing itself is really becoming more mainstream as more, and more collaboration applications are coming to mind.
Interestingly enough, we're just actually in the midst of completing a survey of over 1,000 business professionals. We did this across the U.S. and Asia. As you may know, Asia certainly has a different mindset when it comes to using real time video due to the networks and so forth. But a couple of interesting points that came out, for example, is that you'd be surprised to find that about 60% of users actually have used video conferencing or video chat capabilities relatively frequently.
We're seeing that consumer level technology, maybe that's Skype or Yahoo Messenger, Google, they're pushing this capability to the business level. What we find interesting is that video was seen as a lot more useful in communicating internally at companies, so a lot of employees at different sites, for example, would be using these types of Internet-based capabilities. I think that's going to drive the technology into what's widely going to be known as unified communications, and really seeing that being adopted at the desktop level.
That's part of the reason why we've invested our R&D into, for example, the H.264 technology. That's a standard video technology that will provide the interoperability among multiple solutions out there. Part of that comes with a new innovation called SVC, scalable video codec.
If you think about video conferencing today, you're going to have multiple participants, so let's say there are four different participants. You can have two out of the four, very high bandwidth, office environment. One might be in a home, DSL, fairly robust bandwidth environment. Another person might come in from a mobile line. If that's the case today, a lot of the technologies will relegate pretty much everybody on that video conference to the mobile quality. For efficiency's sake, that's how it's done today.
With a technology like SVC, you really can deliver the optimal quality for each participant, so the two that are in the office with very high bandwidth capabilities are going to get the HD level quality that they can get, while the mobile user will get the mobile level quality they can only receive. These kinds of technologies are really pushing the quality in almost a lifelike experience, along with the HD Voice, to the forefront with a lot of users.
Lee: I wonder if you're able to tell me a little bit more technically about SVC? I know that good people tell me it will be "everywhere" within the next few years, but I wonder if you can tell me, to some kind of technical level, how it achieves the optimal quality?
Jan: Sure, basically instead of compressing the video signal into just one stream that you have to take as it is, it creates different layers. It starts with the base layer; the basic lowest quality is supports. Then it adds enhancement layers on top of that. The idea is you can then, depending on each of the user's bandwidth constraints, screen size, etc, you can peel off the layers to the appropriate amount of data you need. Then you only have to transmit that data to that person, instead of sending either the smallest that works for everybody or sending too much to someone. It makes it very flexible because it's so easy to just decide what layers you want, and you can do that in a very low processing way so it doesn't cost a lot of processing either.
It's not hard to build something with layers. The trick is to do it in a way that it doesn't mean that the bitrate for the highest person goes up tremendously because the layers introduce overhead. That's the beauty of the SVC standard, that they've achieved a way of doing that in terms of low overhead and very high efficiency.
It also comes with some other improvements. Because you have layers here, you can find ways to protect the stream against packet loss better. One of the biggest issues we have with video is you lose packets, and then you lose your history, and you get a freeze of the picture or you get a lot of pixilation. By protecting the base layer, you know you'll get a decent quality all the time, and then you can let the enhancement layers get through if they get there. If they don't, you still do pretty well. That's some of the basic and most important features of SVC.
Lee: So SVC, H.264 as we see, you'll agree is going to become exceptionally popular over the coming years.
Jan: Oh yes, I mean; we hear it everywhere. All of our customers are talking about it. They see the benefits of it.
Joyce: And it's a technology that makes sense for the market. At some point, yeah, maybe it will be irrelevant if there is enough bandwidth everywhere, but certainly, you're going to have scenarios where the differences will occur, as more and more people do video conferencing.
Lee: Has it been tested over Air interfaces, for example, the extreme bandwidth constraint of mobile Air interfaces?
Jan: Absolutely, that's the idea, that you can go down to pretty much any scenario, and we do a lot of testing over those kinds of networks. Today it's mostly audio, but it's starting to happen on the video side as well.
Lee: When it comes to the mobile devices, are you planning on taking this H.264 SVC technology to devices?
Jan: Yes, absolutely, I think it's one of the key areas. Again, back to that; those are very constrained scenarios and that's where it's very important to be able to support this without having to dumb down the communication for everybody else that's on the same video conference.
Lee: Is that available today, or in the pipeline?
Jan: It's in the pipeline; parts of it are already available but the full support is coming very soon.
Lee: Now, everybody knows GIPS in terms of IP-based audio delivery. Going by what you're saying so far, GIPS is pushing harder into the real time video, it seems, and is there some kind of synergy between the two because sometimes it seems it's better that the audio can speak to the video side somehow, instead of using separate solutions for each.
Jan: Absolutely, as you know, a video conference without audio is pretty much useless. But, an audio conference without video is still a good way to communicate. The best is if you can have both, synchronized, and work really well. When we talk about synchronization, one of the key things is that the voice and video are synchronized so you get what's called lip synchronization.
With our experience and capabilities of handling the audio, we can make that happen much better than if you just focus on the video side of it. That, and the bandwidth management, you have this bandwidth available. You have to constantly adapt to what's available and to decide how much you're going to use. One important part of that is to figure out how much should I use on voice, how much should I use on video, to get the best possible user experience. Our experience of doing all this audio stuff means that we know what is the impact of reducing the bitrate on the audio so you find the right mix and tradeoff there.
Lee: Okay, it makes sense to me. Again, maybe because I was a developer for so long, I think a part of me would want to hack together some solutions without coming to GIPS. Have you come across any people who became customers, for example, who have tried to bring these components together for themselves, and if you have, why did they end up coming to GIPS?
Joyce: I think you're absolutely right, there are certainly lots of capable engineering folks out there who probably can take a bunch of these pieces and integrate them together. I think what we've seen, without naming specific customers, and we actually do have quite a few that do have that experience, believe they can do it, and usually within about a 6 to 9 month timeframe, will discover that optimizing that specifically for the level of quality they want is harder than they had envisioned.
What happens is there are all these other pieces. Remember, for the applications that we enable, we're a piece of that. Having to not focus in that area actually creates a much better product overall because they can look at the various user interfaces and other protocols that need to play a part in creating these types of applications. We definitely have a number of customers who actually have tried to put the pieces together, achieved some level, but the quality that GIPS can provide is certainly not there.
Lee: Okay, so it's the fact that you're centralizing the R&D, being in the game a long time, and it's spreading the risk for everybody, etc.
Joyce: Correct, and Lee remember, we have been doing this - I think our man hours is something in the order of 200 man years, or something. You get a lot of benefit for kind of taking the advantages of this, if you will, outsource capability.
Lee: I must admit, I've known about GIPS for years and we were all very excited with the iSAC codec. It's really good to hear where GIPS is doing, getting that clear to core message. I really appreciate that. Turning to the conference which takes place soon, can you just give me some kind of idea what
Jan: will be covering there?
Jan: I'm going to focus on the H.264 SVC and the benefits, and how to use that. It's going to be a pretty single-minded talk if you will. It's going to be really focused on SVC, the benefits, and how we look at how that can benefit the video conferencing market here.
Lee: Is that a technical talk on looking at what SVC is?
Joyce: It will be a mix. I think
Jan: will touch upon the importance of high quality video, which obviously as we've just talked about, SVC can deliver on the overall communication experience and then dive into the technology behind it, which obviously is the SVC.
Jan: Really, try to explain this is actually going to work, why this is what you want to do, and how you do it in terms of some quick remarks on what we've found out at this point in how you do this in the best way possible.
Lee: You're scaring me. It's not going to be a GIPS pitch, is it?
Joyce: No SVC is certainly not solely -
Jan: I mean, it's a standard.
Lee: It's going to be an independent talk about SVC, right?
Jan: Yeah, we have learned stuff; we might actually be the only ones that know some of these things, and we'll share that information.
Lee: Okay, so it's insight. I'm glad you're not scaring me. You're sharing your insight of the SVC technology, not some kind of sales pitch of why we need high quality video. I've been hearing that for the last ten years, but you're actually going to share your engineering insight, as you did last conference when you spoke about the challenges of Voice over IP on the iPhone [Slides | Video]. So, it's going to be a similar type of talk as in saying the challenges you've found or sharing insight into SVC itself, from a technical level.
Jan: Sure, that's our goal, to show our insight and our knowledge about it.
Lee: I really appreciate the fact that GIPS as a company supported the last event, came along and shared insight into challenges of audio on the iPhone and you're supporting the debut of the European event, and are really showing yourselves as a leader in communications innovation, and you're kindly now offering to share insight to the public on SVC.
I think I would like to wrap up there. I think we've got the message clear on GIPS as a company and value offering, and what you're bringing to the conference. Once again, I want to say we all appreciate the sponsorship, because sponsors pay for 2/3 of the event. The communications innovation community couldn't have two events running in a year, one in Europe and one in the U.S., if it wasn't for sponsors like GIPS. Again, I want to say for myself and the eComm community, I very much appreciate the sponsorship. I appreciate you spending your time to give me insights into the company. I wish you good luck with taking over the market when it comes to audio and video.
Joyce: Thanks Lee.
Jan: Thanks Lee.
Leave a comment