Tim Siglin: Welcome again to Streaming Media West 2017. I’m Tim Siglin, Contributing Editor with Streaming Media, as nicely the Media Strategy Principal at ReelSolver, Incorporated. Today, I’ve with me Dror Gill, who’s the CTO of Beamr. So, Dror, initially, earlier than we go into among the subjects, inform me slightly bit about Beamr.
Dror Gill: Beamr is the main supplier of video encoding and optimization options, and primarily our clients are MSOs, telcos, and OTT suppliers. We present the fundamental HEVC and H.264 encoding options, that are very excessive density and top quality. We additionally present what we name video optimization options, that are items of software program that take current encoded video recordsdata and make the bit fee decrease with out hurting their high quality.
Tim Siglin: As a matter of reality, I labored at an organization referred to as EuclidIQ for a short time period, as their Strategy Officer, so I used to be very conversant in Beamr from that standpoint. You all acquired Vanguard Video final yr?
Dror Gill: Yes. Last yr we acquired Vanguard Video. We needed to have tighter integration between the codec and the optimization software program. So we went out, and we checked out 10 totally different corporations, and we discovered that Vanguard was one of the best video encoder on the market. We acquired the corporate in March 2016, and since then we have been working very intently with them on integrating the encoding know-how with the optimization know-how.
Tim Siglin: So, that capacity to go actually deep down into the encoder and add the optimization in, at that degree. I heard some actually excellent news out of IBC about you all, that individuals have been saying that integration has really yielded vital advantages, each on the AVC aspect and the HEVC aspect.
Dror Gill: Right. We’ve seen bit fee reductions of round 20 to 50%, usually, on content material. Back at NAB, we introduced this product, the Beamr 5x HEVC encoder with optimization in-built, and it has been doing very nicely.
Tim Siglin: And your foundation, once you say 20 to 50%, that is in opposition to one thing like x265, or … what’s your baseline on that?
Dror Gill: Yeah, that is once you evaluate it to another encoder, working in an everyday VBR mode. In VBR mode, you set the goal bit fee, that is the common bit fee you’re going to get. In our encoder, there’s a further mode referred to as CABR, content material tailored bit fee. When you run that, you set the goal bit fee, however the precise common can be decrease, and that is based mostly on content material. We do content material adaptive encoding on the body degree. For every body, we analyze what kind of content material and what number of bits we should always allocate, and that is how we are able to get a really low bit fee on every one of many titles.
Tim Siglin: Right. In reality, I simply had David Sayed from Brightcove, who was speaking about context-aware encoding. Yuriy Reznik from InterDigital has joined them, and it looks like there’s this … Between, you understand, after I was at Euclid, with Beamr, with what Yuriy’s doing, there’s actually a transfer to do video optimization, and I’ve needed this for a decade or extra, and I’m glad to see that we’re lastly there.
Dror Gill: I’m additionally very completely satisfied that it is lastly occurred. We’ve been the primary. We’ve finished video optimization since 2009. We began with JPEGs. And in 2012, or ’11, we began with H.264 optimization. Now, the entire business lastly understood the message. It would not make sense to encode your entire movies on the identical bit fee, as a result of every video has totally different necessities. And now you see that Netflix are doing it, and everyone’s into this per-title, per-scene optimization.
Tim Siglin: I feel we’re kind of half manner by way of the continuum, as a result of we have moved to per-title. I actually need to see us at per-scene, or per-shot, as a result of even inside the identical title, we have now so many various kinds of content material. So for those who’ll champion that for the longer term not simply of codecs, however of encoding, that’d be actually useful.
Dror Gill: Yes. And we’re one step forward of that, as a result of we’re already doing per-frame. Inside the scene, we’re adapting to every body of the video, and that is been our method from the start. This is actually how one can actually get the optimum bit fee for each bit of content material. In some circumstances, we noticed that clients despatched us a reel of various kinds of content material concatenated collectively. If you have got a naïve technique that simply appears at the start and tries to estimate what it’s, it would fail. But for those who take a look at every body, and attempt to discover the suitable encoding parameters and bit fee, then you possibly can actually adapt to the altering scene.
Tim Siglin: And clearly you need some degree of inter-frame profit as nicely …
Dror Gill: Exactly.
Tim Siglin: … but when you are able to do intra-frame, as you say, that is your beginning foundation.
Dror Gill: Right. You have to take a look at, not less than at pairs of frames as a result of preserving the temporal stream is essential.
Tim Siglin: Right. Yeah, spatial and temporal, you need to take care of each of them. So the 2 panels you’re on today–you mentioned you moderated one on HEVC implementations. And then the way forward for video codecs. I feel any person described it to me because the HEVC-AV1 Smackdown. What’s your tackle HEVC, AV1? Obviously, you personal an organization that has HEVC compression.
Dror Gill: Right. I used to be within the viewers of a 3rd panel that additionally talked about HEVC vs. AV1. In that panel it was extra of a battle between them. In the panel that I participated in, it was extra about encoding methods and streaming and issues like that. We’ve been large believers in HEVC from the start. And really, the corporate we acquired, Vanguard, has been creating HEVC for the previous 5 years. And one of many largest–I feel the largest–OTT supplier, it is well-known, is utilizing our codec for all of their 4K HDR sequence.
We proceed to push ahead in HEVC as a result of we consider that that is the best way to go when it comes to normal. The bit fee efficiencies are there, the standard is there. The market has not superior as shortly as we thought, and it has been hampered primarily by problems with royalty uncertainties.
Tim Siglin: It’s royalty uncertainties, however I additionally get a way that, particularly with context-aware encoding, AVC is sweet sufficient, except UHD or 4K. It’s ok, and if you may get a few of that additional profit, the workflow stays the identical.
Dror Gill: When we have been pushing content-adaptive encoding, we have been saying, you possibly can cut back 20 to 50% of your H.264 bit fee, so that you need not transfer to HEVC, and now that you’ve HEVC, we all know that with HEVC it can save you like 30, 40%, however with content-aware HEVC, which we have now in our Beamr 5x product, it can save you a further 30 to 40% on high of that. That’s because–and that is one thing that got here up within the panel that I moderated–with HEVC, there actually isn’t any perceptual instrument of defining the usual. So that is open to implementation. The effectivity you get in HEVC is due to higher mathematical fashions, compression block sizes and stuff like that. But the fundamental perceptual precept of optimizing video and adapting it to the content material, works for HEVC in addition to it does for H.264.
Tim Siglin: Because, as you say, there was nothing added into the usual that was totally different from what was in AVC.
Dror Gill: Exactly. That’s why our content-adaptive works on HEVC, as nicely. Now we’re pushing HEVC as a result of we need to say, okay, you may get double profit, HEVC and content material adaptive.
Tim Siglin: Well, that is fascinating to listen to, as a result of I’d puzzled, as we went by way of the time period with AVC discount, whether or not we’d get comparable situations in HEVC, and to listen to you say you are getting each the impact of HEVC, in addition to the optimization on high of that, that is a really compelling argument.
Dror Gill: I’d even say that with HEVC as we speak, we get higher financial savings. And the reason being that H.264 implementations, such because the open supply x264, have been optimized and developed over time, and reached a really excessive degree. With HEVC, the implementations are solely starting–
Tim Siglin: Right, x265 is an instance.
Dror Gill: Right. And it is behind, in order that’s why you possibly can nonetheless save extra … Today, there’s extra redundancy left, as a result of the codec itself, the implementation, has not been optimized. But, even on the optimized H.264, we are able to save lots, and I consider this can proceed with HEVC. So the controversy of which codec to make use of, I feel perhaps final yr, at this present, it was nonetheless open. But this yr, in April, Samsung introduced their assist. In June, Apple introduced their assist. Now iOS 11 is out. By the top of this yr, we’ll have a whole bunch of thousands and thousands of gadgets supporting HEVC within the discipline, the entire 4K TVs in fact, the Edge browser in Microsoft. If you have got HEVC assist within the Skylake chip in your PC, helps HEVC. I consider that subsequent yr, 2018, would be the yr that we’ll begin to see huge deployment of HEVC within the discipline, to televisions, to cell gadgets, in the entire OTT streaming providers. We’re already beginning to see demand from our clients, extra demand for the HEVC codec.
Tim Siglin: And only a closing thought on that: This is why, from an AV1 standpoint, it is so compelling for them to get what they’ve, out into the market, earlier than HEVC turns into de facto. If you are considering from their standpoint. They have a window of time during which they will exit and argue in opposition to it, however the optimization piece that you just’re placing in there makes the argument much more tough.
Dror Gill: Right. Also, if we talked about maturity of codecs, HEVC, the software program implementation, we have been creating for the previous 5 years. AV1 goes to freeze, as we heard on the present, within the subsequent one or two weeks, however it’s a reference implementation, not an optimized one. So, certain, you possibly can spin as much as 100 servers on the cloud and do one stream, that is good for a demo, however that is not lifelike to do tomorrow, or in 2018, for stay streaming.
Tim Siglin: All proper. Well, I’m going to remain on the fence between the 2, nevertheless it’s excellent to listen to from you, the optimization piece, having extra profit in HEVC past what we noticed in AVC. Dror, I actually respect your time.
Dror Gill: Thank you very a lot.
Tim Siglin: Absolutely fascinating.
Related Articles
Video: Has the Time Come for Mass HEVC Adoption?
Beamr’s Dror Gill makes the case for widespread HEVC implementation now.
16 Apr 2018
SMW 17: VideoRx’s Robert Reinhardt Talks SRT and the Future of Codecs
Streaming Media’s Tim Siglin interviews VideoRx’s Robert Reinhardt at Streaming Media West 2017.
05 Dec 2017
SMW 17: Brightcove’s David Sayed Talks Context-Aware Encoding
Streaming Media’s Tim Siglin interviews Brightcove VP of Product Marketing at Streaming Media West 2017.
30 Nov 2017
One Title at a Time: Comparing Per-Title Video Encoding Options
Save on bandwidth and maintain prices down: Per-title video encoding options let publishers break away from fastened encoding ladders. Explore the advantages of 4 strategies.
01 Nov 2017
Companies and Suppliers Mentioned
