Making DASH Workable: The DASH-264 Recommendation

going to give a brief introduction about what we’re going to be talking about and why we’re here to bit before I do that so the topic here is dash 264 towards over-the-top interoperability for I do that I’ll say it’s a effort by the dash industry forum for I do that I want to quickly introduce the panelists I’m Mike movie from Qualcomm technologies go ahead will Michael stop among managing director of cos lips and Andrew pop off I’m from by drm and see to my name is Yan ortman the fraunhofer we make audio codecs okay great thanks we’ll get back to these guys in a minute but let me give you a brief introduction I don’t think you really need to see this picture but this is a sort of a motivation for why in particular a lot of these companies are involved in that video is just taking off over the fixed internet but even more over the mobile internet and there’s a lot of interest in having a standard for delivering that I not going to really talk about this about HTT adopted streaming I assume if you’re here you already understand it this kind of gives a cartoon of the technology and the different components the only thing I’ll say about this slide is that there are a lot of different parts to it there’s the the client side there’s the encoding side there’s the ingestion side there’s the delivery side and this is really a good motivation why to actually deliver something from many different content providers to many different clients you need standards and interoperability which is really why we’re here the MPEG standard as people know there’s a lot of input from various groups it’s been ratified about almost a year ago at this point and the dash industry forum is kind of what grew out of that a little bit about the dash industry for me it grew out of the dash promoters group and one of the most the key objectives there is the of the dash interoperability group which is the part that is actually focusing on the technical work that’s being done within the dash industry forum focusing on interoperability it throughout the industry so the objectives are to promote and catalyze market adoption of mpeg-dash in general and in particular to publish interoperability deployment guidelines and so on the dash in our operability working group is really it’s actually kind of hard to read this off my own slides developing in public because I can’t I wish it was here it was developing and publish interoperability testing procedures support conformance and reference software development coordinate on technical matters with other standards organizations generation of identical technical requirements based on deployment experience to profile in our improve mpeg-dash and to generate technical white papers and deployment guidelines and a common repository for dash related code points so all these are kind of important ingredients of the dash interoperability working group which is a sub working group of the josh i f overall so i think this was touched on a little bit this morning there’s there’s the interoperability points there’s the MPEG DASH part as iraj and others we’re talking about this morning the dash spec itself is a huge body of work and what we’re really trying to do here is to say let’s focus down on the parts that we really want to start to point and interoperate with and so there’s specific features that are chosen out of the MPEG DASH itself but in some sense the interoperability goes beyond just what’s in dash it goes to which codex including the levels and profiles that we’re going to interoperate on whether what subtitling and closed captioning we’re going to use which drm baseline technologies another transport protocol specific aspects including metadata so within the dash in interoperability group this is kind of the things that have been really been concentrated on in terms of interoperability guidelines and the basic one is the dash two six two four base and you see certain dates up there for completion of certain stages and I think at this point we’re pretty much on target with those dates you also see listed a bunch of extensions and we’re really not focusing on those completely at this point we’re really trying to get the base done before we really completely focus on these extensions but the extensions are also going to be important and they focus on such things as hevc video which we know there’s a

lot of interest in within the industry in other extensions for example higher quality video and audio poop so the dash 264 base iop your overview is that it’s based on the iso base media file format those are what the segmentation is based on for video it’s h.264 up to a main profile level 3 for audio it’s HEA cd24 subtitle it’s empty time text so all these things that the three the video audio and subtitle both points are beyond what’s actually specified in the dash back and they kind of go beyond it and these are really being nailed down because they want to be able to have interoperability between these different vendors drm baseline iso/iec 2300 one common encryption doesn’t say what kind of DRM to use it just says what kind of encryption and then various drm solutions should be able to work with this and then in terms of dash itself it’s a restricted version can but combination the iso base media file format live and on-demand profiles i’m not going to go into details that’s not really the main point here the main point is to let people know about this work and also just talk about it in the panel discussion in the dash interoperability working group there’s been a lot of test cases and when we develop a test case it’s only gets onto the list if some company volunteers to provide test vectors so these things can actually be tested and I’m not going to go through this litany list of all these different test cases but I think there’s an over 20 at this point and there are company contributions that are supporting all these so that the interoperability testing can actually be done test vectors and conformist there’s there’s crucial work that for each test back case there’s one or more these test vectors that are being provided and then the conformance software targets to verify that the it passes that it meets the criteria for what is supposed to being tested their conformist includes testing testing that the NPD is conforming the segment’s formats are conformant the restrictions are being enforced that are coming from the dash 264 the media style streams themselves are conformant and the common encryption the drm is conforming ok finally I think this is pretty part of the discussion at this point but they’re the I think everybody understands why Dash is important y dash 264 is important is because otherwise there won’t be you know there would be a lot of talk about dash and people dash implementing dash this way or this way this way but there won’t be conformance across the industry so what we really would like is that if you have a client anybody that’s encoding with the dash 264 spec this client will be able to play it out and vice versa if someone is producing encoding according to the dash 264 than any client out there that supports dash 264 is able to play that out ok and then with that I’ll just say one word this picture is really a picture from the Mobile World Congress demo that we did back in 2012 and this kind of really made it clear why interoperability is so important there even in this little demo there are multiple partners involved harmonic alchemy and Qualcomm in this case and even the BBC and and there was a lot of sort of ad hoc interoperability testing that had to be done to make this whole thing work and so we really got the point from this that really interoperability is key here and with that I’ll pass it on to will great thanks Mike so I’m going to just briefly say a few words about Akamai how we do video and why we’re putting our weight and our influence and actually putting our money on the line when it comes to dash 264 we think it’s the right way to be doing Dash and we’re very active in helping define what desh 264 means so you probably know akima is a very large CDN we deliver a lot of content part of what we deliver is media we deliver about six hundred gigabytes of video data every second and right now we’re pushing about 2.2 million streams it’s live and VOD streams so we’re very large and it’s important to us what format these streams go out this rather busy slide I don’t have animation unfortunately otherwise it’ll make a little bit more sense but we have a cloud-based solution and I know cloud is a trendy buzzy term at the moment but it’s what we built 10 years ago and it’s been involving ever since then our media

cloud is called solar it’s a new brand names you can see solar out there to basically think of it as Akamai’s media cloud source files come in on the left-hand side and what’s interesting what with akamai is we don’t require customers to pre segment content so you can push up non segmented contiguous mp4 files one for each bit rate right now leave them on your origin or push them into our origin and then in the network and on the fly we will segment those and package them into the current suite of HTTP segmented formats which is Apple’s HLS Adobe’s HDS you can also do progressive download and smooth HD out of that and as you can see from the diagram in the yellow we’re going to be adding on to that mix impact dash and we’re going to be doing it under the restriction set of dash 264 and within that cloud we can add on advertising we can add drm while we’re doing it will further optimize this by removing the requirement to have the encoder and have the bit rates which is the direct line from the source files over to the origin you’ll be able to just give sahayak wat 1 high-quality mezzanine file and we’ll take care of making the renditions and packaging it on the fly so part of the attraction of us for dash 264 is well the traction put dashes to get rid of all all the noise on the right-hand side there and all the complexity there’s a lot we have to deal with when it comes to different manifest formats and it sure would be nice to be able to focus down we’re not in any way going to drop our existing customers are formats we’re committed to HD S&H TLS HLS and we’re going to add dash into that mix we’re going to add dash 264 because it’s a common sense set of requirements and it’s actually taking features that we’re already doing we’re already enabling on our network if you look at the segmentation it’s very similar to what smooth HD is using its iso base media file format that is also the intermediate format that we’re using within our network so there’s a very natural match for us there the common encryption is highly attractive it stops duplication of content it makes our cache is more efficient because we’re not storing separately protected content in different silos we will be deploying this we our dash deployment will be out in q3 of next year in terms of dynamic instantiation we will be able to do some simple PMD delivery next week if you so wanted that means it would just come into our network and go out we won’t modify for what I’m talking about is a more sophisticated generation of the dash content within the cloud and that is slated for q3 right next yeah I’m coslett is a provider of software technology and tools and services aiming to reduce the cost for video delivery services and in order to do that we are looking for technologies that are helping to reduce the complexity of such video delivery services dash is definitely helpful but that is a very flexible standard it contains a multiple profile mpeg-4 and big too it’s containing a lot of options so what the dash two six forum promises is to reduce these options and simplify them the deployment forum for services so dash two six four contains a set of constraints we saw that before we have SD video HD video 3d video later each TVC including so it’s possible to have a modular approach to encode some one set of content for multiple devices includes multi channel audio features for example include premium features for premium content feature-rich subtitles and the other put I should know that we have a definite set of content metadata avail ever making it easier for the devices to recognize the content if you look at for example HLS it’s not very constrained you don’t know exactly what metadata you have available in the manifest so you need to deal with a lot of variability in this manifest in dash two six four we tried to make it to define set of metadata always available to all clients making intuited sorry aight revivalism is an example form utilizing dash to could you please wedding could you could you for a story okay come back okay okay here by drm one of the you know oldest

and leading drm providers on the market and one of our whole focuses of course to mean to minimize costs for our customers and we serve the major customers major brands in the market and for us it’s important to follow were you know the whole industry goes and mpeg-dash gives us and to the market actually a great opportunity to minimize the cost what we are focusing mostly is on encryption portions we provide variety of ways how the content can be encrypted on the fly using our encoding partners we have our own tools which perform the encryption we also do focusing on actually playback platforms were for us interpro bility is it’s a crucial thing and I think you know the way how the impact dash progressing is very exciting we already made some good progress in order demonstrating the whole technology we did some demos with the Olympics brt it was a good success story and today we are also progressing especially with a multi-channel audio support and 5.1 surround so we working together with front faux fur in order to build this thing and that’s actually it’s already done we do have a proof of concept it’s working today and I think the whole the whole story behind the impact dash I mean yes you guys said you know there is lots of different profiles there may be some complexity but in general as I can see it like for example industry already has adaptive bitrate streaming they know already how to do that how to work with that and I think for the whole industry it’s will be easier to switch to the MPEG DASH and start supporting in on the broad and wide platforms so and yeah now put your site up alright alright so just briefly from offers a germany-based rd organization that has many activities in the scientific area one of them is audio coding so we are usually known as the inventors of mp3 and co developers of the AAC audio format and so we are traditionally an organization that is very active in MPEG and other civilization bodies and so we’ve been involved in the MPEG DASH standard development and are now also actively involved in promoting the adoption of the standard as product adesh industry forum and since we are an audio company i’m going to focus on what’s being defined in dash two six four currently for audio and what we are working on as mike has on his initial slide we have a base set of features that is part of our interoperability document that defines 264 video and HD ACV two for stereo audio and the important part here is from any device makers that in order to support the dash based profile you can in most cases reuse existing decode implementations both for video as bells for audio we have adaptive bitrate support of course for video but also for audio in the profile and as mentioned before there are various test vectors that are available to dash I eff members to conduct in to availability test Mike also mentioned a few of the extensions that we are work right now for video as well as for audio and I’d like to just briefly talk a little bit about the deep work we are doing in for the multi-channel audio extension so as you can see on the slide we are we have formed a task force with current members Dolby DTS from her for Sony and by dr m that is working on that multi-channel extension the scope here is basically to enhance to you the audio support beyond the base profile so add in a multi-channel audio to deliver premium services and we defined a couple of requirements that a compatible clients should support at least one of those ideally all of them or more of them so obviously if you want to play multi-channel audio on the device you should be able to decode it and once an

area is you have to phone like like this one connected over HDMI tune home theater system and so you should be able to play the multi-channel audio on your home theater system you can do that by decoding on the device or you can do that by passing through the bitstream to an external a B receiver there’s also always the option on a mobile device to down mix multi-channel to stereo so for instance for headphone output there are further requirements that we are currently discussing so I’m inviting everybody to join the discussion we are also defining test cases just like we’ve done for the base profile we will have test cases and test vectors that will be available for interoperability tests and I think all the parties involved committed to have the initial test vectors available by the beginning of december when we have our next dash i have face-to-face meeting in san francisco obviously as you can see from the member companies in the group they are various codecs being proposed and will be tested did all the family of codecs dd+ and true HD the dts-hd family and the legacy DTS as well as the AC multi-channel family we includes a TAC multi-channel and impacts around so that’s as I said ongoing work and picking up many features that we’ve seen and being used in broadcast already so that’s another attempt to enable dash functionality on those devices without for instance adding IP costs to two hybrid clients for instance right thank you um so the net for the next part I’m going to leave it pretty open-ended I have one question in my own and I’d like you to be thinking of your own questions to ask the panelists but let me put to put the one question I have which i think is a relatively hard question interesting question for us and I’ll tell you what the motivation is at Qualcomm we are building the dash client into our product as we’ve announced a long time ago and we do have customers coming to us at this point asking for deployment of dash one of the interesting things is that they’re not necessarily members of the dash industry forum so when they come to us they’re not necessarily asking for exactly what’s in dash 264 and that is somewhat of an issue so we kind of like to solve that problem i’ll put the question to the panelists or in general how can we improve the likelihood that customers demand in deploy dash 264 anybody want to take that on sure my bag I can start well I think one of the major concerns from the from the people who is not really involved in developing this and not really involved in the in the Dashon 264 group is well they they should know how this works so I think from our standpoint it’s crucial to bring the life examples and use cases and you know in a very simplistic way with the demos the available content so that you know these people can actually see not the standard and you’re not the whole day’s big documents with the different midlands tall so variables and read it but they just need to look at and you know see how its work that’s one of my vision okay for example and it just to add to that I mean obviously I fully agree with Andrew here it’s one thing to have an MPEG standard it’s another step closer to reality to have an interoperability to document like dash two six four and but in order to add credibility you have to have an interoperability testing program like the one we are launching now which is available to you know dash I have members but also as Andrew pointed out showing real world use of dash is very important we started that with the above the Olympic Trials earlier this year where you know you could actually see this working in a broadcast environment as I mentioned you know we have you know just it’s not a commercial but just to show you it like if members of the dash industry forum can collaborate inside the organization or you know in parallel to that we can actually have real world data implementations like on this google phone we have the by dr m dash client that’s also used the basis for the HBO Go app and here we’re making native use of the codecs that are built into android in this case you know HT 64 and

also now AC multi-channel and we can have a real world dash compliant product using existing hardware and adding the best functionality to the clients being deployed in those devices and if I plucked is into you know my TV and the home theater system we don’t have speakers here but I would now be able to watch a premium HBO Go service and excellent video quality with the Patriots pitching and multi-channel AAC audio and all compliant to the dash two six four specification so I think that’s an you know that’s part of the message that we need to bring across this is real this helps deployment of dash based services I think one interesting point is that it helps to reduce fragmentation so dash two six four can actually deliver on the promise and that you include once and deliver the same content you need to do it once properly and so I lost track before on that you need to you need to include it once properly in preparing for multiple devices multiple frame rates multiple bit rates multiple resolutions up to 3d put all the features you you might want to use in one content set and be assured that every device that is interpolated tested using the h.264 win plated back at least the base profile but probably the additional features on top of that and you don’t need to have with each new device code generation and you including on the server you need to add every couple of months new content on the server include it once properly and rely on the infrastructure of the h.264 the test vectors and the drive of the group to make sure that is the devices we play that back and you don’t need to worry anymore new devices new theorems one format multiple drn’s multiple devices they have we have lots to say we we can increase the likelihood in a number of ways the the one the one positive to driving to people towards a conversion solution is it’s it’s genuinely mutually beneficial it reduces the friction in your ecosystem when you’re using something that is interoperable with other people in the delivery chain so there’s a natural incentive for people it’s the easier way to deploy dash is to deploy something that a lot of other people are deploying and dash 264 is a vehicle for doing that so I think you’re going to see a lot of companies if they try to go out with their own version there’s not really a neck comic incentive in the long term certainly to try to do that there’s a there’s a strong incentive for people to coalesce around the format i think it’s the dash is job to evangelize that format provide working examples provide the test vectors and certainly all all three of those are being worked on currently you also increase the likelihood by getting more content owners to work on dash 264 a dash 360 s bar is open anyone in this room can join dash i F and start working on it if it’s important to your business you should think about doing that there’s a lot of people who are vendors of solutions right now defining the requirements we do need more people who are consumers people who pay the bills so if you work for a studio you work for a delivery agency who work for Viacom with CBS or ESPN or NBC Sports you should think about making sure your requirements are well represented in dash 264 so it’s things that you want to buy and not necessarily things that we all want to sell to you I like that answer so I want to reiterate what will just said I think that’s one of the issues is that I think some of the the consumers so to speak or the people who actually have the content and who will be paying the bill so to speak for whatever reason there’s there are some in the dash i F at this point but not enough and I don’t know exactly why that is but it may be because they think it’s a very technically oriented group which it is but on the other hand I think what we need from those people is not the technical work but the requirements and the user stories and things like that to bring those two into the dash I off so we can all say okay that’s what we should be developing exactly like this so I think that’s one part of it and i think the other part is educational i don’t think we’ve done a great job of we talked about dash 264 if though there’s something out there that’s published then everybody knows what it is but that’s not really the case at this point we have obviously a very good internal version of the dash i F document but we haven’t actually published that out to the rain is it well that’s a good question and we say here we’re going to put that up we’re going to publish it on December 15th but

that’s subject to approval within the dash i have but i think that should be a good target that we should project due to this also this slide here also gives the the website the is a good place to go find information that organization is new so it’s not the best website in the world but it’ll get better over time will last because he’s actually wanted to mean people organizing that i am the web boston yeah yeah so at this point i’d like to open it up to the audience for questions for the panelists if anybody has yes yes yes the iso base media file format why you chose NT 4 and not not for example transport streams and also the eventual ultraviolence I’m not really familiar with that so how does that come into play so who wants to go ahead feel free and are you sure well we basically the impact their supports impact for as also impact to s transport stream so you can you can have both yeah but not inside that she’s not and but not it’s a yeah yeah so really you want to talk it so yeah there are numbers perspectives on this answer that one of the main ones is there a number of features inside iso base media file format that we find attractive the largest one being common encryption so if you’re in the drm space that we feel especially from now my perspective nuh CDN perspective and i mentioned it it also matches the format we’ve chosen to be our intermediate format in our network now that’s a selfish region one thing that’s good for you but that’s certainly good for us it’s also good for a lot of other vendors a lot of people see TS as a solution that’s now 20 years old or david probably knows exact date when it was published designed to solve probably a different problem than we’re trying to solve here and they see him and they see iso base media file format as a newer vehicle that is more flexible and more adaptive going forward it does raise the big question and your it’s probably running through your head of of a schism dash 264 is not compliant with HLS now dash is dash has a profile specifically built by apple to match HLS and that’s good that’s y dash is a broad standard this refinement of dash on this subset and restriction is not compliant as you pointed out with HLS NTS segments but you’ll see a large movement in the industry outside of Apple most people prefer to deliver segmented format the MP faults what Microsoft have chosen it’s what adobe is chosen for hds and it’s what dashti 64 is choosing and you had another part in a question about the Dare couple of things you can only do with iso base media format one is definitely common encryptions or encrypt ones and EDD REMS whenever you need another one is you can embed metadata in there so example use kids is the same content repurpose the same content for downloads you use for a dash play streaming so you can have an impact for a file on the server you use for dash play streaming which is containing in itself the metadata instructing the client how to use the file and you can let the same file download to the kind for an offline use case mpeg-4 is them is the more recent former this is a more compelling for that has more features I think it’s the better choice so there is another question you had if I think about ultraviolet anybody want to say what that is I think you asked about that right so what are qualities imagine ultraviolet as a file format I’m not really familiar with that so how does that come into play is that what you use it what they’re using it for as opposed to mp4 which you mentioned okay so I to violet is a system established by the by the major studios to provide a file-based internet-based download service for Home Videos it’s supposed to be the successor of blu-ray so when you’re not going anymore to the store to buy a blu-ray disc and when you want to bite it online then the solution to them hopefully the solution to that will be ultraviolet and ultraviolet is using the same file format it’s fragmented mpeg-4 what are two valid calls the common file formats it it hits a defined set of metadata ultraviolet metadata but at the end of the day it’s and I cepes media

file from wat fragmented with common encryption and busy not a valid ism a specification growing which is the common streaming format which is based on dash as well and it’s compatible to dash two six four so body can actually do this dash two six voice you can get ultraviolet content and that content is produced by the studios it’s not anymore the service provider with encoding it it’s the studio who is providing the ready-to-use mpeg-4 so you can take that put it on a server and basically directly streaming this dash and CSF is a very strict subset of restriction stricter than the h.264 designed to suit the studio’s requirements for content distribution so we’re working hard to make it so that a CSF file will be playable in a dash to 64 player but not all dash to 64 content will be playable in CSF players because they’re they’re a title subject but ultraviolet is one example also for a an industry standards body that is also looking at what the dash industry forum is doing and defining for dash two six four you know it’s an example for the fact that dash I F is nothing we are not living on an island and we don’t care about what the rest of the world is doing so we have tight collaborations with other industry bodies that are looking in deploying the dash standard so for instance the the EBU is working on a you know proposing how to implement dash to their broadcast members there’s discussions in the US for the cable industry the dash standard is the basis for HbbTV hybrid TV in Europe and so the data entry form is collaborating with tours organizations to make make sure that what we are defining for dash two six four is at least you know part of their specifications as well there might be other restrictions or additions but the goal is to align those various industry efforts to deploy dash and many of the members and SIF are also members in those arms david like there was a question and an answer there so i guess Empey to TS was 1996 okay so yeah I can touch it so a lot of people ask what is the VIP status for dash in particular and by reference to h.264 so it is it has not been clearly established a number of companies have made written declarations that it is does not contain the IP some companies have have yet to do so Ericsson being one of them and there may never and people say to me or doesn’t that scare you how can you commit to something like dash when when that’s IP status is unknown I personally hope that there is a patent pool around it and I hope it is similar in its fee structure to h.264 because that that is is flexible enough that it has allowed h.264 to become the dominant standard now there’s license fees attached to it but it’s not stopping everybody in this room from using it and it’s clear nobody gets nervous about using it because they understand how to pay for and how to license it so i would like to see similar clarity around dash I’m not scared by the concept of a license pool being established I am scared by it offering unreasonable terms and trying to gouge people so I think h.264 is is is a good precedent they’re not the way it started but the way it’s it’s it is now maybe it was driven there by wear them and maybe Google had a lot to do with driving it down to a point where it’s become ubiquitous but if dashes as ubiquitous h.264 is then we’re all better off so 11 comments on I don’t think anybody said it was IP free but i think what they’ve said is they’ve made particular IPR declarations that from their perspective they’re not going to charge these or it’s going to fit into their existing licensing model where essentially no one’s going to get charged anything more so so no one has explicitly said we’re going to go out and demand royal teeth at this point for dash that I know but on the other hand not everybody said that we won’t either so I have the I have the mic over here you guys can’t see I have a question about HLS and specifically the HLS new selective scrambling do you have any comments or future predictions for HLS actually I don’t know what is this selective scrambling you can teach us something oh

no in the latest draft that was published they changed their scrambling mechanisms okay well I you know I don’t know anything about that specifically but i think a ROG made a good point this morning that the that HLS is not a standard it’s an ITF individual draft and probably will remain that way it’s just out there for information and it’s changed something like 10 times at this point and no one else’s made any input to that other than apple and so the whole it’s a changing thing all the time and I’m not sure if this scrambling is backwards compatible with the previous version that they had out there but that’s one of the issues they they have sometimes they put out a new version is not even backward compatible with the previous version and so that that kind of fractures their own market which is sort of interesting that’s one of the things we’re definitely not trying to avoid with with dash 264 in particular is to kind of get the best features from all the different companies together and agree on something that we’re going to deploy and really stay with that and deploy that for the next few years also taking into account that there will be extensions but not every player will play with every extension for example anybody else have any comments about I’m going to stay away from that one yeah I guess we’re the wrong guys to ask you about a chill hey we’re working on something that well on the other hand I mean our company supports HLS and I’m sure and we we do order about traffic is a chelate yeah we yeah there is also like some for example when we talk about the live stream right there is like 10 seconds segments which gives you like latency so it’s pretty pretty old too you can actually without your legs very the segment you can yes the trouble is you’re you’re always depict all a client you can make sure that it’s retry and it’s switching logic is compatible with the shorter segments with Apple you don’t control the client nor do you have any input into the heuristics they’re using to swell so you need to go with a segment length that works well for them and that’s that gets awkward at times and the the various drafts of hls are frustrating for many people there’s no place you can go to to get HLS certified if you’re making an encoder it’s your best efforts and if it plays back on an Apple device which I do own then it works well and that’s what apple care about and that’s probably all they should care about they are not trying to create the convergent standard for the internet it happens they’ve created something is very popular works well on their devices what Dash is trying to do is create a converged standard for media delivery that works on all devices including apples because and actually there was a point made this morning which I’ll say again that Apple was very involved in the standardization of dash in fact they made specifically they took care to be very involved because I think their position is that they’re not going to for business reasons they’re not going to deploy it at this point but if they’re ever do decide to deploy a standardized version of adaptive streaming they want to make sure it’s a good one and therefore they wanted to make sure they had good input into the dash standard I think HLS has a bunch of limitations when you’re looking at premium features there’s no defined lee how to do subtitles and there’s no defiantly how to do things like multi-channel audio nor is it clearly defined what the what the decode about the device is able to expect from a from the manifest I mean the minimum set of features in there is is really little so if you’re looking at future services that that try to improve the user experience I think you’re pretty much going very fast at limits using HLS speaking of women’s were at the end of our session so we’re kind of limited in that sense maybe we can take one more quick question what’s that ah the question we need to pimp this evenings yes okay we’ll do that thanks thanks for the reminder kept it really that I have one more question 10 thing that’s currently interested in the industry is advertisement specifically advertisement instruction on the client device is any work been done in within the best and therefore the mark-up of the playlist the actually addresses opportunity and it’s done have been done regarding advertisement and construction sorry it was hard for us to understand the question is okay yet so josh what does what yeah yeah I had so dash is well suited to the dynamic insertion of advertising into the manifest it has a concept of periods so the presentation is the entire thing the entire the entire show or the earth stream and it’s made up of a sequence a single linear

sequence of periods these periods can hold dissimilar content and in fact it’s designed to carry ads that are then linear ads and surjit inside your stream so the players know that when they encounter this and maybe different bit rates different representations different sizes that’s okay players can handle it you’re not forced to try to mold your ad to be exactly the same and to fit in to your primary carrying stream so dash has the notion of periods with Qualcomm we dem it dynamic ad insertion that bulb our world congress back in january so it’s it certainly works we’re looking for the ad insertion companies to be bringing it out you’ll see Akamai doing something along those lines next year so dash is a is a great vehicle for server-side ad insertion it’s a little harder to describe but it can even handle the case where you don’t know what time the ad is going to play like you’re watching a sports game and there’s some injury on the field and suddenly how tense an ad so it can definitely handle that case as well anybody else want to say anything about this before we okay