My audio-tweet was from Srini Kumar, developer of TinyVox. He wanted to know what I thought of the app's "voicemail on Twitter" approach and its retro cassette tape aesthetic. I said I'd be happy to check it out on the condition that we conduct our interview asynchronously, back and forth over TinyVox. So we did, and I learned more about communication than any social app has taught me in a while.

Srini:

@jonmwords cool path article - stealing the News Feed? taped with audio app TINYVOX

The metaphors are all over the place with this app. Srini called it, at various times, "voicemail on Twitter," a "mix tape," an "audio brainstorm" that can be a "throwaway" just for getting ideas out, and "podcasting for everyone." It was hard to decide how to use it. I figured concision was a good rule of thumb, so I just shot off a brief question about the medium itself.

Jon:

Question 1: About whether or not this is a new kind of medium taped with audio app TINYVOX

Srini came back with a huge response, full of passion and color and drama... and it was really, really long. But it was clear that he intends the app to be all of those things and more. Whatever we can do with our voice, Srini wants TinyVox to help us do more.

Srini:

The Tyler Durden effect ? taped with audio app TINYVOX

I loved what he had to say about the honesty and unsettling newness of communicating this way, aloud, spontaneously, without constraints. But exchanges of 10-minute messages didn't seem sustainable to me. This began to seem like a problem with the way the app works. Tweets are constrained to 140 characters, and that's why the medium works. These "audio-tweets" break that wide open.

So I asked Srini whether he agreed:

Jon:

Question 2: On the tendency (temptation?) to go long in this medium. taped with audio app TINYVOX

Honestly, I sort of expected him to take a hint and rein it in for the next answer, but he didn't. He came back with another six minutes of rhapsody, pushing me on the cultural norms that made me want short, tight answers. It's hard to concentrate and really listen to someone, even when they're sitting right in front of you. Would we be better to each other if we worked on that?

Srini:

we're spanning time taped with audio app TINYVOX

So I did. I practiced the art of paying attention, and I listened to every word. I found myself sympathizing with his whole message much more deeply than I do on Twitter. A Twitter person is just a picture, a handle and a burst of text. But committing to listening to a six-minute tape of someone's voice makes you follow his train of thought wherever it goes. I learned much more about where his head was at than I do about people in a comment thread.

For my last question, I let myself open up the same way. I asked him about the nostalgia and sentimentality of TinyVox itself and where the app is going:

Jon:

Question 3: about how the TinyVox recipient gets to keep the sentimental metaphor of the mixtape. taped with audio app TINYVOX

Srini's answer was vast again, but it was really exciting to hear from a developer with so much love for the interaction he's designing. Rather than summarizing where TinyVox is going, I'll leave you with Srini's audio answer. TinyVox is available for iOS and Android, and I'd be interested to hear how you find ways to use it. Share them in the comments here.

Srini:

the mixtape in the cloud taped with audio app TINYVOX

Note: the timestamps are off for the recordings in this exchange because I didn't realize that TinyVox is better about privacy than I initially thought. It doesn't post clips to the Web unless you explicitly tell it to, so I had to ask Srini to re-upload them after we were done.

The new player first went into beta in November, giving those curious enough an opportunity to experiment with it. Now that the bugs have been ironed out and a few new features added, the widget is ready for prime time.

A big part of that growth, cofounder Alex Liung told us, is the company's mobile strategy. It has highly functional native apps for iOS and Android, but the browser-based Web is another story. When users try to play back embedded SoundCloud clips on an iPad or iPhone, the old player simply wouldn't work, like all the other Flash embeds scattered across the Web.

HTML5: A Necessary Move

For this reason, the conversion to an HTML5 embeddable player is a necessity for a service like SoundCloud if it expects to maximize its reach across the Web and keep growing. Flash will presumably never be supported on iOS devices iPads and iPhones, a fact that wouldn't matter so much if Apple didn't keep selling millions upon millions of them. Just last quarter, over 15 million iPads and 37 million iPhones landed in the hands of consumers, and this revolution in personal computing is still well underway. In November, Adobe announced that it would suspend development of the mobile version of Flash, apparently accepting that open standards like HTML5 and related technologies could do most of what Flash was capable of it.

It's not just cross-device compatibility that makes ditching Flash a good idea. HTML5 is also less resource-intensive when playing back audio, and should exhibit better performance.

While testing the new player out, the SoundCloud team discovered a few other advantages as well. Apparently, the appearance of the revised player made users twice as likely to hit the "play" button. From there, they were eight times more likely to share a given track with others.

Now that it's coming out of beta, the new player now supports saving any sound to one's "likes" and making comments in-line on the waveform, which was one of more the powerful features of the original Flash player.

If the player's interface reminds you of Pandora, it's no accident. The team deliberately borrowed from personalized media services like Pandora, Flipboard and Zite when building out the Infinite Player. Its controls are sparse, containing only a few buttons. Among them are a pair of icons for voting stories up and down, much as one would on Pandora. In time, the player learns what you're interested in and plays back content accordingly.

This experience provides more of an opportunity forr what the NPR team calls "distracted listening" - that is, consuming content while doing other things and not necessarily having to make any decisions about it (aside from voting it up or down, if you're so inclined). This is in contrast to the type of "engaged listening" experience that podcasts and audio clips offer.

The player, which launched yesterday, is in beta mode and currently works only in Safari and Chrome. Its functionality is driven by HTML5 and JavaScript, rather than relying on Flash for playback. It doesn't appear to be optimized for the iPad just yet, but it is a brand new feature and presumably the team is working on cross-device compatibility. You can give it a shot here.

The social audio-sharing site experienced several hours of intermittent downtime yesterday as SoundCloud's engineers fought off the attack.

Just the day before, the site experienced similar downtime issues, although it isn't clear if that was related to the DDoS attack or not. In terms of a motive for the attack, it was likely just the thrill of bringing down an increasingly popular service, rather than the fulfillment of some ideological agenda as we've seen in the past with Anonymous and Wikileaks-related DDoS attacks.

As of today, the site appears to be loading and operating without issues.

SoundCloud was founded in Berlin in 2008 as a social networking site primarily for musicians to share and promote their music. It has since grown into an audio-sharing platform for amateur and professional musicians alike, as well as podcasts and other audio snippets. In short, it does for audio what YouTube does for video, albeit on a smaller scale.

Lo these many years ago, Kelvin and I met at what became Ask.com. He wound up as the Head of Testing and I ran corporate projects. He now works in the DC area as the web producer for a history publisher. It was there he met Al Webber, a jazzman of the old school. Al recently passed away, but not before technology empowered Kelvin to capture, preserve and share a part of the man's ineffable essence.

"Two years ago, I asked my dear friend, Al Webber, if I could interview him about his life-long love affair with jazz for Storycorps. He agreed. At the time, various factors prevented us from making the trip to New York's Storycorps studio. But fortunately, Al and I recorded a rehearsal session at his home. Half of the interview recordings have been uploaded. The other half will be online soon. Currently, the audio does not work on iPhone or iPad. Special thanks to Ted Taylor who retrieved the audio recordings from the disk drive of my defunct laptop. May you find some joy and comfort in these recordings of our loved one and friend."

Webber, who had a passion for New Orleans jazz - something I have in common with this friend of my friend - gigged until the week he died. He had been a husband, a father, a newspaperman and a veteran of the Allied landings at Normandy Beach, but jazz, he said, was the love of his life.

As I said at the beginning, we in the tech business are no less charged and altered by the movements affecting our societies than is any other group of people. So, the fact that here in America we have a black president is certainly important. But as important is the fact of a young black geek becoming friends with an old white jazzman and of cheap computerized recording technology and the Web making it possible for his voice to ring out in thousands of ears that never had the pleasure, before today, of hearing the name Al Webber.

What love demanded, technology allowed.

Tech, people.

And love.

Listen to Kelvin and Al talk.

Although conserving such a large and important collection of American music is important in general, what makes it really exciting are the plans to digitize it and make it available online.

"A surprisingly high percentage of America's recording heritage since the early part of the 20th century has been lost due to neglect and deterioration. The donation...will help maintain the inter-generational connection that is essential to keeping alive, in our collective national memory, the music and sound recordings meaningful to past generations."

In fact, a congressional study found only 14 percent of commercially recordings made before 1965 are available to the public and only 10 percent of music released in the U.S. in the 1930s can be accessed by the public.

The recordings will be digitized at the Library's Packard Campus for Audio Visual Conservation in Culpeper, Virginia, where the masters will also be stored. The Library is preparing a dedicated website to debut in the spring on which they will stream recordings from the collection.

From Ella to Pine Top

Other highlights in the collection include the following.

• Ella Fitzgerald's and Louis Armstrong's duet "Frim Fram Sauce"
• Bing Crosby's 1947 version of "White Christmas"
• The Mills Brothers' "Paper Doll"
• Josh White singing "Jim Crow"
• Machito and his Afro-Cuban All Stars Mercury recordings
• Clarence "Pine Top" Smith's "NowIAintGotNothingAtAll.mp3"

Additional artists represented include Billie Holiday, Tommy Dorsey, Irving Berlin, Jimmy Dorsey, Ella Fitzgerald, Judy Garland, and Dinah Washington. Subsidiary labels include Decca, Mercury, Vocalion and Brunswick.

Given this is the first collection of masters donated by a major recording company, the hope is that other such companies will follow suit. One can see the wisdom of donating 5,000 linear feet of masters that are no longer being used and are taking up commercial space.

The 200,000-plus recordings provide a tangible bump to the Library's audio collection of about 3 million items.

We sent a host of technical and access questions to the Library but, despite repeated requests, never got answers to them.

I first came across the site only today, when an old friend, the perennial exile activist and writer Sokari Ekine, tweeted a blog post on the site, which she'd been using as an antidote to living in Florida. (My words, not hers.)

"I started to wish I was back on Kilburn High Road recording the late night rats in the garbage from Nandos, the drunks coming out of the pubs, sirens all day long, endless chatter and traffic, kids playing - well screaming in the park, couples fighting, youths dealing, singing beer drinking out of town football fans from strange places like Wigan and Stoke City, and my own music trying to silence all these external noises and the neighbour shouting and banging on my wall trying to silence me."

The spread of cities represented in this first open-source, crowdsourced database of found city-sounds and maps is remarkable. The listener can travel from Tokyo, Japan to Belo Horizonte in Brazil, from Granada, Spain to Bamako in Mali. The maps lean toward Europe and Latin America with an emphasis on larger population areas. Hopefully, with the crowdsourcing element fully engaged it will get more from Africa and from rural areas.

You can also search for cities based on quality keywords, like "beat," "birds," "churches," "industrial" and "weather."

Sonicity

Sonicity is an on-site sound mixer that allows the listener to pluck and mix cities, individual sounds, pretty much anything in the database.

Users have also utilized the database to create music. Stanza has composed a musical "world tour" and released it as an MP3 download called Soundcities CD.

The experience and possibilities inherent in the project are pretty intuitive, but here's Stanza's take on the experiment.

"(S)ounds...give clues to the emotional and responsive way we interact with our cities. Cities all have specific identities, and found sound can give us clues to the people that inhabit these spaces, as well as provoking us and stimulating our senses in a musical way. I am interested in the sounds of specific places, and how the sounds reflect this identity and re-imposes characteristics back onto the location or environment."

Stanza's experiments may have been among the first, but are not the only. Among the others are the British Library's UKSOUNDMAP and the BBC's Save Our Sounds project.

What's your experience with capturing and interpreting sounds online? What sources and tools do you use? Where do you post the results? Let us know in the comments.

***

Other sources: Intersections, Weird Vibrations, Oso

The George Ewart Evans Collection "document(s) rural life and agricultural work in the late nineteenth and early twentieth century, folk beliefs about animals, medicine and witchcraft, folk and popular songs."

If you register on the site, you can respond to a request that comes with every sound file,

"Can you tell us more about the context of this recording? Or can you share information on its content - timings of key sections or important details? Please add your notes. Uninformative entries may not be retained."

In an email announcement Rob Perks, lead curator of oral history at the British Library, pointed out the oldest of George's interviewees.

"The oldest are Aldeman Ling, born in 1875, discussing bell-ringing; George Messenger, born in 1877, who talks about threshing and about working on the barges at Snape in Suffolk; and Susan Mullenger, born in 1878, recorded in 1967 talking about eating fried mice as a remedy for whooping cough!"

To hear the voices of men and women talking about, say malting barley, men and women who were born 130 years ago and long gone, to hear the conversations between them and George, is to realize that history is no grand thing, a castle on a distant hill, but a real thing, something you can touch with your hands and hold in your mind, hear and see. It's a real gift.

Essentially, Fotobabble attaches an audio caption to any image you can upload. It's a cute, fun way to share and narrate photos with friends, and could even be useful for certain kinds of online businesses - for example, photographers who wanted to explain more information about a particular shot or online retailers who wanted to give potential customers details about a product. Can Fotobabble accomplish these tasks better with audio than conventional text-based captions do now? Read on and tell us what you think.

Currently, users can choose to share their creations across a wide variety of social networks or email; however, autosharing is not built in. The app is available as a web app for PC/Mac/etc. and as an iPhone app.

We do wish that Fotobabble would let users audio-caption pics from Facebook, Flickr photo streams or elsewhere on the Web. Ensuring ownership would be easy enough, as well, through Flickr's API or Facebook Connect.While we're on that subject, account creation should be possible through Facebook Connect or Twitter OAuth. Ideally, we'd also want to be able to create slide shows and sets or groups of pics.

What do you think: Can you see yourself using Fotobabble? If so, how would you use it? Let us know what you think in the comments.

Listen is currently only available for Android-powered phones such as the G1 and myTouch 3G from T-Mobile. To download the application, users can scan the enclosed QR code and follow the onscreen instructions or look for Listen in the Google Android Market.

The End of YouTube's Mashup Culture?

Most importantly, this move by YouTube, even though it makes perfect legal sense, might quickly put an end to the culture of remixes, mashups, and parodies on YouTube.

This new policy will also negatively affect a number of services that rely on YouTube for their audio content. Audiolizer, which we reviewed last month, for example, is, on the surface, a streaming music service, but actually gets its music from YouTube's vast catalog of music videos.

What About the Users?

In the end, this new policy will only alienate YouTube's users, while doing nothing to help the struggling music industry. It would make a lot more sense for the music industry to provide a blanket license to YouTube so that users could use copyrighted sounds tracks on their homemade videos, while the record labels (or the artists) could get a share of the advertising revenue.

HTML 5 aims to improve HTML's interoperability and address the growing demand for more diverse and complex web content. It also addresses HTML 4's lacking features for web applications. In this post, we'll look at 5 exciting new features in HTML 5.

A bit of history

The conceptual thinking for HTML 5 began in late 2003. The World Wide Web Consortium (W3C), the organization that oversees the web's standard protocols and guidelines, expressed interest in the HTML 5 draft originally developed by the Web Hypertext Application Technology Working Group (WHATWG), a group formed in 2004 consisting of representatives of Apple, the Mozilla Foundation, and Opera Software. As a result, the W3C HTML Working Group was formed in 2007 to develop the specifications of HTML 5.

Development is underway, and HTML 5 is anticipated to reach W3C Candidate Recommendation status in 2012, though many modern browsers already have partial support for HTML 5 specifications.

Exciting new features

1. New HTML elements that improve our ability to describe content

HTML's primary task is to describe the structure of a web page. For example, by enclosing text between <p></p> elements, HTML tells the browser that the text between those elements is a paragraph.

Increasingly, diverse web content has outgrown HTML 4's ability to accurately describe the content of a web page. Streaming video and audio is now commonplace. Website regions such as navigation menus and branding areas (commonly located in the header) are now staples of most web pages. More importantly, advancements in JavaScript, Flash, and server-side technologies have allowed for the proliferation of rich Internet applications (responsive, user-driven, browser-based applications), despite current HTML specifications.

By adding more HTML elements, HTML 5 aims to give developers a better and more precise way of describing data.

For example, under current HTML specifications, this is how we would describe the structure of a typical web page:

The problem with this layout is that, to the browser, everything is a <div> element. The browser treats everything inside the <div> elements equally because it can't tell the difference between them, and "classes" and "ids," such as content, sidebar, and footer, change from website to website.

In HTML 5, this is how you might describe the layout of the same web page:

In this markup, the browser now knows what each part is. It knows that the web page's main content is inside the <article> element, that the website's navigation is inside the <nav> element, and so on.

Besides prettier-looking and more semantic markup, the practical implications are endless. It increases our markup's interoperability. For example, an external system, such as a search engine spider, would be able to more accurately determine what content on a web page is important. It can skip crawling the <nav> element and the <footer> element because they probably do not contain the main content of the web page. A well-formed HTML 5 document would therefore give search engines a better understanding of the content being displayed.

A crafty developer could create an application that pulls out just the <article> section of a group of websites, say, to aggregate it in a database, or generate a list of all the videos on a web page by finding all the <video> elements.

Screen-reading software could give vision-impaired users a quicker way to traverse content sections. They could go to the <article> element directly if they wanted to read the main content of the web page, or go directly to the <nav> element if they wanted to navigate away from the page.

2. Improved web forms handling

These days, it's hard not to come across forms on websites. You encounter them when submitting a comment on a blog, registering for a user account, or sendinging mail in Gmail. HTML 5's proposed specifications include a huge revamping, called Web Forms 2.0, of how web forms would be handled. It gives web developers a lot of options and new features for effectively and easily handling input fields and form submissions.

The most exciting thing about Web Forms 2.0 is form validation. Currently, developers are required to use JavaScript (client-side) or PHP (server-side) code to validate inputs. For example, many web forms contain required fields (perhaps for the username and email fields):

In HTML 4, the markup of the web form above would look something like:

Currently, you have to use scripting to validate a user's submission. In this example, developers would have to write their own validation code (or use a pre-made script, such as this one) to ensure that required fields aren't left blank by accident or that submitted email addresses are valid (usually by doing something called "regular expression" matching).

In handling this form without requiring the author to include a validation script, HTML 5 (with the current Web 2.0 specifications) would give us additional element attributes, such as required and email attributes, which automatically check that the username and email fields are not left blank and that the email address' format is valid.

3. APIs for easier web application development

HTML 5 will introduce several application programming interfaces (APIs) to new and existing elements, aimed at improving web application development and addressing current issues with HTML 4's lack of ability to allow developers to mark up web applications.

One API is specifically for working with audio and video and will be used with the <audio> and <video> elements. It will provide audio and video playback capabilities and eliminate the need to use third-party applications, such as Flash, to develop and display media (at least for supported media files).

Check out this scripted video controls demonstration, which accompanies this Opera Developer Community article on the video element.

4. The <canvas> element allows image scripting on the fly

Most people take in information more quickly and effectively through visuals. For example, between a table, numerical data, and a pie chart, the pie chart gives users a better feel for the scale and relationship of data (at least most of the time).

The downside of images is that they're static. If you create a pie chart using a traditional method (for example, with an image editor like Photoshop, or a graphing application like Excel), you wouldn't be able to adjust any data that changes without manually editing your graphics.

With the <canvas> element, you can take constantly changing (database-driven) data and apply it to a pie chart like the one above, as well as other types of 2D visuals (even a cat, if you're so inclined), via scripting.

The canvas API also allows users to interact with <canvas> elements. For example, you can write a script that responds to users' clicking on a particular section of the pie chart.

5. Users can edit and interact with sections of a web page

The section in the proposed HTML 5 specifications about User Interaction describes new ways of marking up interactive web pages. The contenteditable attribute (a boolean attribute to which you assign either true or false) allows you to indicate which parts of a web page users can change.

This can be useful for wiki-style websites, in which content is user-generated. Another use of the contenteditable attribute would be to create web page templates. You can allow certain regions of a web page to be open to content editing and lock other regions that shouldn't be changed. This gives users of your website who aren't proficient in HTML an opportunity to input content safely without affecting critical areas that should be handled by more knowledgeable users.

At the document level, you can make an entire page editable via the designMode attribute, which accepts two values: on or off.

In Sum

The proposed specifications are slated to reach W3C Candidate Recommendation status in 2012, but that doesn't mean you have to wait that long to start using some of the new things in HTML 5. A lot of modern browsers, for example, have already implemented the <canvas> element (including Mozilla Firefox, which has partially implemented it since version 1.5).

HTML 5 will redefine how web developers mark up content. It will provide a better way to describe the content displayed on a web page, enable more complex content types, improve media and web application support, and increase the interoperability of HTML documents.

Note that things are still under heavy development and are bound to change; many of the proposed improvements may be heavily revised in the next few years or not implemented at all.

There are a lot of exciting new features in HTML 5. Share your own favorites in the comments.

Further reading

• You can read the latest working draft of HTML 5 specifications on the W3C website.

• Learn about the major differences between HTML 4 and HTML 5 on the W3C website.
• IBM developerWorks has an excellent in-depth article on new elements in HTML 5.
• Read about the people in charge of developing HTML 5 specifications on the W3C HTML Working Group website.
• Find out what you can do to help HTML 5 development on the WHATWG wiki website.

About the author

Jacob Gube is a web developer/designer and author of Six Revisions, a blog on web development and design. If you want to connect with the author, you can follow him on Twitter.

As the team points out in a recent blog post, traditional audio CAPTCHAs based on distorted digits or letters are relatively vulnerable to automated attacks and can be broken by using machine learning algorithms. Indeed, Wintercore Labs, an IT security firm, showed how easy it would be to break Google's audio CAPTCHA solution earlier this year.

Transcribing Old Radio Shows

By using old audio clips, however, ReCAPTCHA is circumventing these security problems (you can here an example of these clips here by clicking on the speaker button).

One problem with this type of CAPTCHA, however, is that a lot of these clips are quite hard to solve - especially because a lot of them are from radio plays and feature different voices within a single clip, as well as the occasional audio effect. Most of the clips are about ten words long.

The reCAPTCHA team acknowledges this problem by allowing a certain amount of misspellings and other mistakes, but even with some practice, we still didn't get far beyond solving every third CAPTCHA correctly (but then, a lot of visually impaired users might be more sensitive to picking up these audio clues). If you did better, let us know in the comments.