Curated by: Luigi Canali De Rossi

Wednesday, October 5, 2005

Search Audio Inside Podcasts, Radio and TV Shows: TVEyes and Podscope

Sponsored Links

The realization of the Long Tail and the opportunities it delivers to the small publishers, niche musicians, independent video-makers is tightly weaved with the ability of Internet entrepreneurs and open-source developers to design effective search tools that enable true exploration of specific media universes like books, radio, podcasts, TV, images, tags in detail and with precision.

Photo credit: Karl-Erik Bennion

The more such new advanced search tools will make it easy to search and find beyond what the traditional Google text-based search approach provides, the more the Long Tail of content will gain acceptance and recognition.

Think of podcasting today. Which podcasts get to be subscribed to? It is a handful of names that get 80 or 90% of the subscriptions simply because there is no easy way to find out more about the others. So, in the doubt of subscribing to some majestic crap, individuals prefer to subscribe to the popular few podcasters that were there first or who made enough noise to get top visibility.

"But so are you saying Robin that even podcasting is a stardom universe in which only a few get visibility and attention?"

Robin Good: Yes. For now that IS what we have.

"But with so much variety of authors and content themes why only so few get that attention?"

Because unless you make it easy for people to explore, search and discover what they want by themselves, in a comprehensive, detailed and reliable way, people will only see what you show them: what is most popular, what appeals to most; often sacrificing on overall quality, depth and focus.

Can you search and find podcasts based on the actual content recorded in them?

Think now how powerful it would be to monitor and track radio and television programming from any station, country or language and be able to monitor in real-time when anyone program talks about something you are interested into?

Well, settle yourself for some great news as today I have here in front of me David Ives, the CEO of a company that provides search for keywords of interest across live audio and video feeds while having been introducing recently corporate and consumer products that allow organizations and individuals to do keyword searches against radio, TV, and now podcast feeds.

Let me introduce the gentleman to you:

(just click the play button here above to start listening - here below also the full text transcript of our conversation)

RG: Well good morning to you David and it's a pleasure to have you here. You are the CEO of a very promising and interesting company, at least from my point of view. Can you tell me the name of your company and what it is all about?

David Ives: Sure, I am the CEO and founder of TVEyes, that's T-V-E-Y-E-S and we started the company seven years ago to index live television and radio. And we have been building the business through a combination of B2B services, where we provide a search engine to corporations that are able to search for keywords of interest across live audio and video feeds. And more recently, we have been introducing some consumer products that allow consumers to do keyword searches against radio, TV, and now podcast feeds.

RG: That's absolutely fascinating, and it sounds to me from what I've seen of your technology, that you are one of the very first being able to not only do this, but also to bring it to market in an accessible and cost-effective way. But, let's start at the very beginning. David, do you want to tell us a little bit about yourself and your background?

David Ives: Yes, my background is software development. I've been a custom database designer for over twenty years. I had a consulting business prior to founding TVEyes, and in fact the formation of TVEyes was prompted when one of my engineers was day trading, back in 1999.

And instead of writing code for me, he was making money by trading stocks in short intervals. And we thought wouldn't it be a great idea to write some software to watch television specifically the financial channels so whenever one of the CEOs or one of the tech companies was mentioned on a financial or news channel, we would be able to fire off instant alerts to users around the world who could perhaps have a trading interval so they could buy or sell a stock before it made it's move.

So we built the first version of the TVEyes software to provide strictly text alerts, and we now have advanced the service to full video where we are capturing 24/7 audio and video from 135 radio and TV channels in 7 countries.

And we allow a combination of commercial access to be able to watch these live or near real time clips from any internet connection, and we also have certain stations we have made available to consumers so that they can find relevant video clips according to their needs.

Now, what's actually unique about our capabilities is we are actually creating a spoken word index against all audio and video signals.

So we are not searching on the metadata, and we are not dependant about descriptive data that someone may attach to an audio/video file. But we are using voice recognition techniques to actually create a spoken word index across all of the spoken words within an audio/video file, and frankly it is the only way to sufficiently search an audio/video file, just as we've come to depend on full-text search for Google, Yahoo and other search engines.

RG: Fascinating stuff; do you feel you are the only one indexing audio out there? ...I have been exposed I think in the past to... well I don't want to shrug some wrong names, but I recall a major search engine company that sells search technology to large enterprises, I think showing me that capability though at very high cost and capable of creating spoken indexes, also of recordings in different languages. Are you aware of any other company doing that yourself?

David Ives: Yes, there are a few providers of voice recognition technologies.

In fact, we're probably aware of everyone in the space, and what makes us unique is we are approaching the problem from multiple angles. So, we are not dependent on any one search technology to create our spoken word indexes. We have worked closely with law enforcement agencies in multiple countries to search varying qualities of audio files, and the proper solution really involves a clean interface, which we believe we've created with our products along with a combination of voice-recognition engines on the backend. So, we do feel we are very much in the lead with the voice-recognition capabilities that we have.

With that said, we know this is an exciting space audio/video search is, and it's attracting large companies like Google and Yahoo and AOL, so we are seeking relationships and have made good relationships with some of the big companies, specifically Yahoo and AOL, to distribute our search engine on their platforms.

RG: So, to say it in simpler words, TVEyes and its connected services make it possible for people today to access the hyped and much buzzed about Long Tail of Audio and Video by providing a unique way to find content by providing the actual content which is in video clips and podcast shows; stuff which is normally accessible only by searching the title or the description, or the short notes that containing descriptions and references to this content. Is that correct?

David Ives: That's right, Robin. In fact, I was doing some searches for your name in our podcast search engine, which we call Podscope, P-O-D-S-C-O-P-E, and what I found is that a number of podcasters will reference your name and your website as a part of their podcast, but your name is not included anywhere in their descriptive data or show title. So, without our search engine it would be impossible, unless you listen to every single one of these podcasts, it would be impossible for you to find just those instances where your name was mentioned.

RG: It is also true that there are some podcasts out there that DO have my name in their title and short notes but which do not appear in your index. But, I guess that is part of the growth and of the fact that maybe you are not indexing the entire universe of podcasts out there.

I'm referring for example to a quite popular podcast from the two ABC journalists Hobson and Holtz which is called For Immediate Release; they've talked about my articles and reviews a number of times.

But I haven't been able to scout or find my content included in Podscope, your excellent audio search service, and so in fact I was curious to understand what do you attribute that to; if it's early stages still, the fact that I may have to submit to you or what else can contribute to helping this content and podcasters make sure it is included.

David Ives: Yes, we are... it's a combination of factors. We currently index over 4000 unique podcasts and we have a total of 45000 podcast files that we've created indexes for, but we are every day discovering new podcasts that we need to include within our index. We also have a number of people submit their podcasts to us so we are made aware of their existence. And then there's also the reality that voice recognition technology is not 100% yet. And, while it gets better every year, it is very much a combination of art and science and there may be instances where a word is not properly picked up by our index, and it is an ongoing challenge where the accuracy rates are greater than 80%, but, closer to 80% than 100% in today's state of the art. So it's constantly improving, but there will be some omissions due to the limits of speech recognition.

RG: Now let me understand better, what is the difference between the different services that you offer? I mean, does the interface of Podscope, which anybody can access at, provide me access also to a different content index than the one I could access from TVEyes?

David Ives: Yes, we have a professional product at Google's AdSense as our advertising mechanism, so when a user comes and does a keyword search you will see contextually relevant advertisements down the right-hand side. And, that is the revenue generator for the moment; however we will soon be adding some premium content that will be embedded within the search results that will allow our users to go and subscribe to premium content. Perhaps audio books, or periodicals, any type of spoken word content that is available on a subscription basis, it will soon be available within the Podscope search results and we would of course receive a revenue share by driving customers to those premium sources.

RG: So will we expect to have the opportunity for independent audio producers to pay and have premium positioning inside your podcast search engine so that they can promote specific audio products and the like?

David Ives: That's correct, and we will flag those paying advertiser so that there is no confusion, and you will know within our search results which ones are paid and which ones are based on our normal ranking algorithms.

RG: Great, that makes a lot of sense, too. And now for the big news: it has been running the grapevine and along the blog-lines as well that you have recently made a great partnership, (you have already been mentioning Yahoo) with AOL as well.

And what I understand you are going to be providing for AOL is to utilize your audio search technology to provide greater and easier access for the AOL audiences to all of the podcasting, audio-based content essentially, and I don't know if that includes video too, that they are going to be serving out. Is that correct and can you give me a little more insight into this?

David Ives: That's right. We are really excited about our recent announcement with AOL. They are going to be distributing all of our podcast search results on the AOL search engine, and it will be branded as an AOL search product but powered by Podscope. So, we will be credited as a search engine for all of these searches.

It will include both audio podcasts as well as video blogs and we will be providing the same preview functionality, which is key, which allows you to listen to a ten second audio snippet or a ten second video clip containing your key word search term. And the AOL deal will obviously give us significant distribution world wide that we will be launching in the coming weeks.

RG: Fantastic and I must say you do have some great ideas implemented in that Podscope search engine. I myself found out something that not everyone may know yet, and this is that not only you have this nifty preview clip, but also that the Podscope search engine does something truly amazing and extremely rewarding for the searcher.

When you search for something and you get the results, like the podcast where those words are contained, you can actually play that podcast from the very section in which those things are being said. So, if I search for Robin Good and Podscope reports me some results, when I click play I may have to wait a few seconds, because if in that podcast they were talking about me at minute 47 out of 60, there is, I assume, something in the background that is playing forward through that virtual digital tape until the point where there is someone saying "Robin Good" and playing it back for me, while also automatically giving it a head start of a few seconds. (check this out yourself by following this link and clicking on the play button under the first result). Nothing better could be expected by someone searching within audio. So my sincere compliments for, you know, that feat, I think it is really outstanding.


David Ives: Thanks very much, Robin. Yeah, I have a great team of engineers that I've been working with here at TVEyes and they continue to innovate and provide very consumer-friendly applications.

RG: It has been a pleasure indeed to find out about these things. For those listening I think that there is nothing better than going out and trying out for yourself Podscope because it is so straightforward, easy and simple to use, that I think it speaks much better than any extra thing we could say about it, beyond this initial introduction to any one user that wants to be on top of communicating and finding information effectively. So, my thanks to you for your availability David, for sharing this time with us, and I leave it to you for any closing remarks, recommendations, URLs, promotion, anything you have.

From Robin Good, here in Rome this is all today with David Ives from Podscope, the CEO of TVEyes, the premier company that indexes audio out of video clips, podcasts and audio shows of any kind, both from the mainstream media and the grassroots. Hats to you David, and over to you for the closing.

David Ives: Well, thanks very much, Robin. We are excited to be leading the way with audio/video search. While there has been a number of companies that claim that they have video and audio search engines, we believe that the fact that we're not dependent just on the metadata, that we're creating a spoken word index against video and audio files, that is the key to discovering the relevant information that is buried within audio/video files across the internet.

A Good conversation with David Ives of and

(click the play button here above to start listening)

Readers' Comments    
blog comments powered by Disqus
posted by Robin Good on Wednesday, October 5 2005, updated on Tuesday, May 5 2015

Search this site for more with 








    Curated by

    New media explorer
    Communication designer


    POP Newsletter

    Robin Good's Newsletter for Professional Online Publishers  



    Real Time Web Analytics