Streaming Media West '15: EuclidIQ Debuts IQ264x, Tackles Complex Video Problems
[This sponsored interview was conducted at Streaming Media West 2015.]
Jose: Hi, this is Jose Castillo, and welcome to Streaming Media West, Almost Live 2015. Today we have a very special guest with us. It's CEO Richard Wingard. CEO, co-founder of Euclid. I've got a couple questions for you, would love to chat with you a little bit. What does EuclidIQ do exactly?
Richard: Thanks, Jose. EuclidIQ solves complex video problems. The kind of problems that have been really a pain for the industry for over a decade, and we do it in an elegant and scientific way. We're at the show today to announce our product that we call IQ 264X, which is part of the family of products that we call IQ 264 that addresses the need for the x264 encoder market. It's a plugin that plugs right into the x264 encoder.
Jose: Got you. H.264, why are you trying to solve problems around that when everybody else is talking about the next level?
Richard: This is a great show to talk about that. Everywhere I went, and I've gone the past two days, people are talking about the argument about whether and when HEVC will be a product. That leaves us with an opportunity. The fusion-style base of H.264 decoders out there are much larger than the install base of HEVC decoders, and we think that gives us a real sweet spot to target.
For example, we give them about a twenty percent bandwidth reduction for H.264, and even for x.264, which is considered one of the better H.264 encoders. So, if you look at the sweet spot of 1080p, sixty frames a second, and then I heard just an hour ago that they're looking at HDR as a better encoding technology than HEVC, we'd be foolish not to take advantage. We think that our clients are going to want to look at the largest addressable market for the decoder perspective.
Now, having said that, HEVC, the other alternatives in the open sources you probably heard about, we're looking at those options, because the VP9, VP10 ... We're looking at those options, because fundamentally our technology in concept would be additive, cumulative to their compression claims, so we are looking at doing that. Right now lots of opportunity with H.264, and right now with our deployable X.264 technology.
Jose: Got it. Now, you note that your technology does encoding directly, and that you're not post-processing the video. Does that mean you can increase the quality of x264 video encodes?
Richard: That's exactly right. We can do one of two things. We can either give you the bandwidth savings, or we can increase the quality. A quality you can actually see in the video. The choice is yours.
Jose: Got it. I love it. Other solutions on the market, and again, you've mentioned some of the competition, but they claim to lower bandwidth, too. What makes IQ264 different?
Richard: In today's world you have to do a couple things. You have to lower the bandwidth, you have to give your clients the option of improving the quality, because that's a competitive benefit in the industry, and you have to be fast. If you don't have those three things, it's going to be tough, I think, to be relevant, so what we did was, we integrate directly into the encoder so we don't have that pre-processing, post-processing bandwidth, that problem with taking hours of the encoding to work.
On the other side is, we can always give you better quality. Sometimes better quality and even lower the bandwidth. That gives our clients a real option in this marketplace.
Jose: Definitely. Tell me about the future of EuclidIQ.
Richard: The future is the part I love the best, but we're going to pay the bills with what we got out here on the floor today. What I like about our approach is that this API plugin is ... Nobody's said a real 20 percent bandwidth savings isn't important, so we know we've got something of value. Second thing is, we're not done.
We could come back to our clients with more bandwidth reduction, or quality improvement, and that's because we have a huge group. For a startup we have a ... We just hired another PhD from Princeton with a lot of video experience. We have this brain power, and we can continue to work on the algorithms to do more in the encoder, so our next step is to take advantage of our ability to do better rate control based on our quality optimization.
Jose: Got you. Tell me a little bit about maybe some of the other projects that EuclidIQ's working on.
Richard: That's a great question. One thing that I would say is that we've been working on very interesting, complex video problems for over a decade, and as we sit here today I've spent my life dealing with proprietary decoders, and now I have one that the technology works very well with compliant decoders. I spent eight years on the INCITS L3.1 committee, and everybody was waiting for HEVC to show up, even though it would require a different decoder.
What's interesting about that is we actually have, I think, over 45 patents on technology that was way in front of the marketplace. We were doing modeling ten years ago, where we model objects and features in the video. One thing that we came upon that we have patents on that's extremely exciting, and would mean significant benefit to the industry is what we call model re-use.
Model re-use is the ability to take the encoding of data for one video, and help with the encoding of data in many, many different disparate videos in a video repository. We have done a lot of research, and we have been able to validate that there are patterns in these videos that are redundant and can be tracked.
That's sort of brings this concept to some level of maturity in a prototype.
What that means is that when you look at all those redundant models, you can actually get more compression than we ever thought of before beyond 264, 265, even EuclidIQ, because this redundancy is mathematical, and these features, like a feature in a LeBron James basketball game is redundant to a feature in a sitcom or a Brave New World, or a Band Of Brothers.
That's a real breakthrough. Now, in addition to being able to do more video compression with what we call smart models, we're able to do some other things that are unique. We're able to do things like search videos in a novel way, index videos in a novel way, and archive videos in a novel way that's never been done before. We think that's a future leg of our technology, and we're excited about that.
We're also very excited to have IQ 264x out there in the market, because a lot of companies are going to x264, and we want to be that second step for them to get more value.
Jose: Excellent. It sounds like you guys are doing wonderful things now, and looking forward to the future. This has been a great conversation with CEO and co-founder of EuclidIQ, Richard Wingard. This has been Almost Live 2015 Streaming Media West.
This article is Sponsored Content