Overlay.TV is proud to preview the first episode of The OverView (click for all episodes) with Amber MacArthur, our new weekly series for online video fans.  Enjoy and be sure to leave any suggestions or feedback in the comments.

In this embed I have disabled the create, on/off function, and carousel to give an example of a cleaner implementation of our player.

Thanks for watching!

creator · eepybird · lighting · online · overlay · Overlay.TV · overview · stickynotes · The OverView · Video · viral

One of the overlooked features of Overlay.TV is how easy it is to integrate us with your existing analytics package, be it Omniture, FireClick or any package that utilizes transparent gifs to deliver the tracking pixel.  This can be done in a couple of minutes to any video, and of course you have the option of adding other widgets to the same video (product, clipart, text, etc).

To do this you first need to generate the report or goal in your analytics package and determine the increments of engagement.  For example, if you have a 3 minute video, you probably want to measure how many people made it to 1 minute, 2 minutes and 3 minutes respectively.  Whatever page you embed the video in will record the number of views the player gets, and the player itself will measure how many people hit play.  This last stat is available in any Overlay.TV video by clicking on the info pane at the bottom right.

Once you have generated the 3 pixels, you simply advance your video to the appropriate point in the timeline and open the Link widget.  In the Image field, paste in your transparent pixel.  Complete this for each of the pixels you have created.  Now you go ahead and embed your video in your page, start building traffic, and let the pixel and your analytics package do the rest.  To get the most of your experiment I would create a custom report that graphs usage statistic across the 3 pixels.  This will measure drop-off engagement etc.

If you are using other widgets like the products in our affiliate catalogues I would suggest placing them near that pixel, either right before or right after based on what you want to measure or understand better.  This will then give you a map between engagement and revenue or referrals, which will make your report even more powerful from an ROI perspective.

I am planning on building an example of this with a report - so if you would like me to do it to your video, leave a comment here with a link to your video and a way to contact you and I will work with you on this.   If you have other ideas or analytics issues you are trying to sort out, let me know and I would be more than willing to work with you on them.

analytics · engagement · tracking pixel · Video

Gather round kids it's story time ...

Once upon a time in a land not so far away there lived a girl, a girl who loved to create. One day she happened upon a website by the name of Overlay.TV, and her creative skills flourished. She created Overlay after Overlay, adding clipart, widgets and products to every one. Her favorite thing by far though was to find just the right video to use, and create just the right products to add to her Overlays (and maybe make a little money on the side, but I'll save that story for another day). She would search high and low for the perfect video and the perfect products to add to her Overlays, everything from clothing to iTunes, and manually paste in the URLs. But as time went on, the girl who loved to create found that she was so good and so fast at creating Overlays, that to manually add the URLs every time just wasn't fast enough. "There must be an easier way," thought the girl, who loved to create. "If only there was a tool that lived in my browser that I could call upon to harvest products and grab videos quickly and easily, without having to leave the site I'm on every time." What was a damsel in creative distress to do?

One day, while admiring her proverbial garden of Overlays, she found the answer she was looking for right there on her Overlay.TV MY PAGE. It was called a Bookmarklet, and did everything she wanted and more. It lived in her browser, it harvested products quickly and easily, and gave her the ability to start creating Overlays on videos right from their site, with just one click of a button.

Now, the girl who loves to create uses her Bookmarklet whenever she wants and so can you. Install your own Bookmarklet  and use it to create products and grab videos to Overlay. Read ahead to find out how to use the Bookmarklet and live happily ever after ... (more...)

Bookmarklet · Overlay.TV · overlays · Platform · products · tips · tool · tricks · videos

.

In Part 1, we setup the Player SDK, tried out the test player and created a skeleton widget that can be animated on top of a World of Warcraft in-game video. Now we get into the fun part. We're going to query the WoW Armory, initially with a hardcoded player name so that we have some results that can be styled in the ViewPane. At the end of this post, the view should be functional and show some stats for a hardcoded player. In Part 3 we'll make a configuration pane which will allow an author to configure a player of their choosing.

(more...)

armory · in-game videos · MMORPG · Player SDK · widget · World of Warcraft

.

Online gaming is huge and watching videos of other players to see how they play and to learn their techniques is a pretty standard activity in the gaming world. In fact, there are YouTube gaming stars like Leeroy Jenkins, Athene and Ming and many more that have in-game videos with millions of views.

What their videos are missing however is the data the gamers crave. Graham, an Overlay.TV developer and a recovering gaming addict tells me that the first thing that people want to know when they are watching an in-game video is what the players’ stats are. So if it’s a World of Warcraft (WoW) in-game video, the viewer would typically find the players’ server and player name in the title or description of the video or sometimes rendered in the video somewhere. Then, they would visit the WoW Armory and search for the players name to see all their stats. Things like the players’ guild name (their “team”), ranking and level, their weapons and a wealth of other information is provided in that portal.

Graham thought that combining this data with the video would make a great widget especially since there’s an xml feed for the data in The Armory. I completely agree. That’s what Overlay.TV is all about; interactive video where relevant information about the video is accessible right from the video. So, what I’m going to do is write a few posts about how you can use the Overlay.TV Player SDK to create a WoW widget for in-game videos. Later on, Graham is going to show you how to give the player a WoW skin. In the end we should have a really cool WoW in-game video player that you can configure to show player stats at various times and locations during playback. And since it will be built upon the player core, you’ll be able to put other widgets in the video as well like live chat so players can talk real time about the action.

(more...)

armory · in-game videos · MMORPG · Player SDK · widget · World of Warcraft

Why in-video context inference is so difficult

Google AdSense changed the world of online advertising, using standard Information Retrieval techniques to match ad inventory with web pages. This has opened new possibilities for advertisers to get a better bang for their buck and send the right message to the right audience. Can the same be done for video? Easier said than done.

Web pages lend themselves nicely for textual analysis. Not only can content be stemmed, counted for term frequency and enhanced semantically, cues in an HTML document such as headings, page title and metadata provide a much richer base for context inference.

Video, on the other hand, is a much tougher nut to crack. It is unstructured to begin with and has much fewer metadata elements that could be of any use in drawing semantic conclusions. Nevertheless, video is what web users are now consuming more than ever. The number of clips served daily by video sharing and premium content sites is staggering, and so is the ever-increasing time users spend on online entertainment.

It is, therefore, not surprising that more advertising dollars are now being diverted to online media on the expense of traditional TV campaigns. However, the fundamental question still begs – can this money be spent more effectively delivering not only the right message to the right audience, but also at the right time?

Time introduces the most challenging factor in video advertising. While concepts discussed on a web page are linear for reading but parallel to access (the whole page shows as one piece in the browser), scenes in a video are temporal and can only be consumed sequentially. Not only do we need to wait for content to be streamed to our client before we can scan it, skimming over the timeline is not quite the same as scrolling up and down a web page. To this, lets add lack of anchors and in-content links to realize that inter- and intra- navigation in a video space is very different from navigation in hypertext.

So how can advertisers tap into this elusive medium and deliver contextual messaging? Currently, with great difficulty and without much success. One can obviously use the surrounding textual information of the embedding page as well as look at the video title, description and tags to make some assumptions as to what the video is all about; but that would be like drawing similar conclusions based on a whole web site rather than at the page level – to use a web analogy. Nothing in and around the video can actually tell us much about the particular scenes or even define when those start and finish.

Better tools are, therefore, needed - if visual data is of not much help then how about audio? That, indeed, is the flavor of the month in contextual video advertising with companies using speech-to-text techniques to transform audio-visual data to temporal text signals to drive good old semantic analysis. Alas, verbal information only captures a small fraction of what actually goes on in a video. It is enough to consider narration over background visuals in a documentary or the stylishly rich details of a music video clip to see how so much information is lost in this translation.

There is no silver bullet to automate this just yet. No clever algorithm – trained, unsupervised, adaptive or other – can come close to an average person’s ability to easily describe what they see in a video. Any person, you say? Why not use plenty of those then? By the laws of large numbers, they’re bound to reasonably describe it collectively.

Stay tuned for the next post on crowd-sourcing context inference in video and why advertisers should care.

No tags

Ok, so it's already pretty cool that you can appear next to your favorite artist using the video record widget but I'm experimenting with other possible features.

Tyler helping Radiohead out

What if my surroundings were removed and I could appear right in the studio helping my favorite band out with my wicked guitar skills? Kind of Holodeck-style. That would be really cool. Professionals do this all the time, normally with the use of green screens or blue screens and a video editing program that can remove the color of the screen so that another image or video can show through. It's called Chroma Key.

Although it's not very hard to make a functional green screen, I can't expect music fans to erect one and light it properly. More likely, they won't want to even change out of their pajamas and will expect the performance to work in their messy bedroom. That's ok, I'll see what I can do.

Instead of subtracting similar colors from the video image, I'm experimenting with subtracting a known background from the video. This requires first getting a picture of the background and then in each frame removing any pixels that are similar to the background shot.

I tried various combinations of BitmapData.compare() and a DisplacementMapFilter but in the end the best results involved processing each pixel in each frame. Yikes... What I really need to do this efficiently is the ability to create a custom bitmap filter. Custom filters are possible but they're new in Flash 10 with Adobe Pixel Bender and well, were not going to require a Flash 10 player just for holodeck support.

Here's a work in progress. It's not smart enough or efficient enough yet and it cuts out part of your body if it happens to be the same color as the background but it's still fun. Oh, and you need a video camera.

I'll post updates as the code improves.

as3 · video compositing

While we have had a considerable number of blog posts singing our praises as an alternative to conventional advertising, reader reactions have run the gamut, from "AMAZING, very cool!" to "Só que eu acho que é muita interrupção, era informação o tempo todo" (translation: I think that it's too much interruption, there was information all the time) and one post from a fellah by the name of Greg, who felt it was both cool and distracting:

"Cool technology and I love the concept. It did seem a bit distracting to the video though with the flying objects. I only watched Sabatoge (sic), so I may have totally missed a different point of view. I think the possibilities are there, perfecting how it intrudes into my video watching will be the key to success." via Hypebot. This is the video Greg is referring to:

Of course we love praise, but we also appreciate thoughtful, intelligent feedback such as Greg's. However, there is something that we should clarify: you can turn the Overlays OFF. Yes, that's right, OFF. The choice has always been there for those folks that do find it distracting. It's up to the user to decide how many targets they choose to put in an Overlay from one to ... well the sky's the limit, and the viewer to decide whether they want to watch the Overlay with the targets On or Off. Point being is that there is a choice. Our goal is not to ruin the integrity of people's work. The Sabotage video, for instance, was done as a proof of concept to showcase the different elements of Overlay, but certainly this is not what we think every video with an Overlay should look like. Here's another tip, if you are of the less is more persuasion and feel that adding stuff to your video is a cool concept, but feel that the hot spots and links detract from the overall aesthetic of the video, the targets can in fact be invisible. Here is an example of an Overlay with invisible targets using a clip from the HBO show Entourage: As you can see, almost everything within the clip is clickable, from Ari's robe, to the bed, to the wall sconces, without any interference to your viewing pleasure. If you see something you like you can click it and buy it, otherwise you can just sit back and enjoy the show, interruption free. The choice is yours.

No tags

CHOICE - OPT-IN

The original idea for Overlay.TV was conceived while watching my wife watch TV. She would sit down to a line-up of shows that had been PVR'd the night before and be delighted with the fact that the next few hours of her evening would be spent happily viewing the shows she had selected. Of course, there would be the occasional commercial interruption that she would fast forward through but at least it wasn't like the bad old days where she was forced to sit through several minutes of advertisements.

My wife wasn't doing anything out of the ordinary. Everyone I know has their own variation of on-demand content consumption. Some people pause a show when it starts, make dinner and then resume viewing when dinner is done allowing them to skip past all the commercial breaks. Others fill their content plates using clips or full length shows found on video sharing sites, torrents and other peer to peer networks. With the Internet-oriented method of viewing content, users also shun the “traditional” advertising. Banners, pop-ups and the like have very low click through rates and the minute that a video sharing site introduces the television model of interrupting content to show an ad, many of their users become upset and simply move on. For whatever reason, the new generation feels entitled to free content. Free, as in not costing money or time.

The common attribute among these viewers is that they are watching content without paying any attention to the advertising. Traditional advertising models have not kept up with advances in technology and are severely broken. Any new model would need to work in the world where the user expects to watch what they want, when they want it. This was the first ingredient for the idea.

THE CONTENT IS THE AD

The second ingredient was this: Although people will choose to disregard the interruptive type of advertising that has been chosen for them, product placement is still partially working. My wife will watch a video or show and make a comment about the shoes, handbags, hairstyles or the lifestyles of the rich and famous. She might simply be curious about what an item is or she may actually be interested in buying something just like it. Other people have similar thoughts about the things they like to watch from tools, to gadgets to vacations. Given the right timing, the right mood and the opportunity to dig deeper into what is being shown, these people would most certainly either read more about or purchase items on the spot. If the mood or timing isn’t perfect, they might just set the thought aside for another day. If only they knew exactly what thought to set aside. What were those shoes that the celebrity was wearing? What was the name of that resort that the cast of the reality show were staying at? And where can I get me one of those diamond studded circular saws that was used on that home improvement video?

USERS

Assuming that you could come up with an unobtrusive system that allowed people to delve into what they were watching, how would that system know what the objects in the video are? Artificial Intelligence isn't even close to being able to identify objects in video at the detailed level required. Viewers want to know what brand, where it can be purchased and what is comparable. To do that, you need either the people that were involved in creating the content or people who know about the subject matter. That leads me to the third ingredient: the users.

Users would be able to identify, classify, tag and comment on objects inside the content. Users could be professionals working for advertising companies with obvious access to product placement details or they might be outside the advertising world looking to monetize content on their own site. There could also be users looking to be paid to use their familiarity with snowboard gear or their knowledge of fine china in exchange for marking up content. Getting users to identify objects in video would be straightforward in the specialized world we live in. All they would need is an incentive and that was simple too. Since companies pay for referrals, especially when they result in a sale, sharing that money with the user would be the perfect incentive.

CHOICE - OPT-IN. THE CONTENT IS THE AD. USERS.

Building a system that combines those concepts is what we've been doing for the last year and I know you're going to love it. Overlay.TV's model works with today's culture of choice where you get to choose when to watch advertising and what that advertising consists of.

advertising · concept · user-generated · Video