O’Reilly blog entries [example] now feature a small “Listen” icon to the right of each article. Clicking it causes a widget to start reading the page to you in a very smooth/natural synthesized voice. This is all real-time — it’s not like they’re having someone read and record every article on the site. A company called ReadSpeaker provides the software that makes this possible.
The obvious application is for non-sighted users. But wait – blind users already have screen readers set up, or they wouldn’t be using the web to begin with. So who is this for? Sighted users who want to close their eyes for a few minutes? That seems like a very limited application.
While pondering this, it hit me: ReadSpeaker’s widget only works on specially enabled web sites. Imagine a FireFox plugin or browser extension that, when clicked, would run the text of any page through a voice synthesizer like the one O’Reilly is using, but pipe the output silently to MP3 in the background, then load the generated file into my podcast aggregator. All day long I could “tag for voice” various web pages that I wished I had the time to read. When I sync’d my iPod before leaving for work, I’d have all that missed content on it, ready for the road.
Yes, Young Edisons, this is a business opportunity. Run with it.
Update: MacDevCenter blogger David Battino sampled the audio output of an entry and mashed it up into a little song. Says MP3 output is on the way from ReadSpeaker.
Music: Sufjan Stevens :: Chicago (Multiple Personality Disorder version)
The Copyright Royalty Board has recently decided to nearly triple the licensing fees for Internet radio sites like Pandora.
The new royalty rates are irrationally high, more than four times what satellite radio pays, and broadcast radio doesn’t pay these at all. Left unchanged, these new royalties will kill every Internet radio site, including Pandora.
savenetradio.org has been created to raise awareness and reverse the tide, before this vital medium is smothered in its crib. Please consider sending email to your congress-critter / reps, encouraging them to stop the madness.
Guitar solos are self-indulgent. The bridge is always boring. Verses are repetitive. Everyone knows four minutes is way too long for a song. What we really want is the hook – the essence. Give me a meaty riff, and ditch the rest. Radio SASS (Short Attention Span System) “creative editing” to the rescue.
Short Attention Span System takes the playlist and musically condenses songs to their essence. Through time compression, you get the memorable heart of each song, with an average length of aproximately two minutes with NO self indulgent guitar solos, NO long intros, NO repetition of choruses again and again. Radio returns to the snappy song length of the 1960s.
In other words, everything long is bad. Because time is an inconvenience, and self-absorbed artists with no respect for your fast-paced lifestyle are wasting it. Ummm… ewwww? So what happens to Hot Rats? Cosmic Charlie? Fool in the Rain? Mothership Connection? Born Under Punches?
But look on the bright side — nobody cares!
Radio SASS starts out with the memorable beginning, followed by the best verses, best chorus and then wraps it up just as you remember … Will listeners object? The answer is no. Several focus groups conducted by Harker Research show that most people don’t even notice.
Also interesting here is the name of the service: “Short Attention Span System.” Since saying that someone has a short attention span is generally considered a bit derogatory, this represents a sea change. SASS must think that people are not only aware of the fact that they have short attention spans, but also don’t think of that as a bad thing. The marketing here is aimed at the heart of what has traditionally been considered a human weakness, or a negative aspect of media snack culture. Kind of like selling potato chips under the name “Obesity Chips.”
And, oh yeah – the new protocol is patented. You can patent butchery?
Why is Avid / Digidesign suddenly pushing Garage Band rather than Pro Tools? Is there an implicit acknowledgment here that PT is too complicated / expensive for a huge swath of users? Maybe this doesn’t seem weird to others — of course Avid can still sell you the hardware, even if you don’t go for their integrated M-Box/Pro Tools package. Maybe it strikes me as odd because of the endless battles we’ve gone through at work over the question of whether PT is overkill for our users (we’re now teaching Soundtrack Pro to multimedia journalism students rather than Pro Tools, so I guess we landed somewhere in between).
Apple profiles ukulele master Lyle Ritz, who recorded his latest album No Frills entirely in Garage Band (at age 75 no less). And it does sound gorgeous.
Designed by Sander Mulder & Dave Keune, Buro Vormkrijgers. This is functional kitsch; the wrong becomes the new right. By adding a function to an otherwise grotesque object, it acquires new aesthetic values, becoming an object of desire. Pun intended, this woofer holds the mids between an addition to your sound system and your loyal 4 footed companion. Available in a co-axial two way speaker system version [two dogs].
Fun as they are, somehow I just can’t see giving over my audio to a visual gag certain to wear thin after a few weeks. Or to have to repeatedly answer the obvious next question: “Are your tweeters shaped like birds?” Especially for 600 Euros.
Back in January 2004, Wired ran a piece on how the average volume level of commercial CDs has been steadily rising for the past two decades (Pump Up the Volume). Check out the visual waveforms on that page comparing AC/DC’s 1980 “Back in Black” to Celine Dion’s 2003 “I Drove all Night.” Amazing difference.
Compression is the act of reducing dynamic range during the mastering stage. Rather than making the overall volume louder, the loud/quiet peaks and valleys are brought closer together so there’s less delta in the waveform. The result is that the average volume is greater, even though the loudest sounds aren’t louder (though engineers do push the total volume to the max — at customer insistence — as well). Just as TV commercials seem louder than normal programming — their average volumes are higher, even though peak loudness is not.
In the days of LPs, pushing recorded volumes too far could result in the needle jumping out of the groove, and pressing plants would reject recordings that included clipping artifacts. After the advent of the CD, the needle problem went away. Artists quickly realized that louder overall volumes made a bigger impression on listeners … so louder music had a higher chance of becoming a hit. But despite the “impact” that louder music makes, the ultimate result for listeners is fatigue. Human auditory perception just wasn’t designed to listen to sound bereft of dynamic range.
Today, audio engineers are “exhausted” from trying to resist clients who insist on high compression levels. Either they do it, or the artist goes to an engineer who will. What’s unusual is that the vice president of a major label (Angelo Montrone of A&R) recently wrote an embattled plea to an industry newsletter, basically asking engineers to stop the madness.
Austin 360’s Everything Louder Than Everything Else summarizes the letter and provides an excellent overview of compression and its effect on humans. What I found particularly interesting was this cultural theory on why huge compression has become the norm:
So why aren’t more people noticing this sort of thing? One word: lifestyle. We listen to music in completely different ways than we did 20 or 30 years ago. For most people, music is listened to on the go, in cars, on headphones while running, on computers at work. Music has to compete with the sound of your car’s engine, has to punch through the background noise of street traffic or a loud office. “Ours is a culture of competition,” Wofford says. “Maybe labels think the music has to be super aggressive, super bright, like a kid screaming in a supermarket, to get your attention.”
Follow-up to Eno, Wright, Generative Systems: Eno later described the session as “Two strangers becoming friends in front of 900 people.” Two guys in completely different fields working on exactly the same thing — building generative systems from cellular automata. Numberless:
[Each of them] use the idea of cellular automata as a basis for their creations. Cellular automata … refers to a simple initial rule-set that is capable of generating very complex and disparate results.
Shortly after the session, Wright announced that Eno would be creating the soundtrack to the upcoming game Spore. I’m not a gamer, but I’ve been looking forward to this game (due in 2007) for a long time now.
Wikipedia: Spore is, at first glance, a “teleological evolution” game: the player molds and guides a species across many generations, growing it from a single-celled organism into a more complex animal, until the species becomes intelligent. At this point the player begins molding and guiding this species’ society, progressing towards a spacefaring civilization.
Some of the screenshots and video floating around the internet are amazing, but apparently don’t do the actual gameplay justice. The generative link between Eno and Wright could result in some great audio. Most game music is set on endless repeat, but Eno’s audio will be sui generis, and will never repeat. Wright:
“Science is all about compressing reality to minimal rule sets, but generative creation goes the opposite direction. You look for a combination of the fewest rules that can generate a whole complex world which will always surprise you, yet within a framework that stays recognizable…..It’s not engineering and design, so much as it is gardening. You plant seeds.”
bush-of-ghosts.com forks to both the official Warner Music site representing the great 1981 collaboration between Brian Eno and David Byrne and a second site, from which users can download individual source tracks from the original album and re-mix them into their own creations — many of them quite beautiful.
When Bush of Ghosts was first released, the kind of remixology Eno/Byrne were doing was pretty unusual, though now commonplace (but seldom as successful). Their decision to offer the album up for public remix 25 years later (!) is poetry. But Eno has always taken the long view (he talks on the SALT podcast about how he had to tell a gallery owner in which he was doing an installation that the duration of the audio he was using was “approximately 6,410 years.”)
Posted back in 2002 about the Long Now Foundation – created by Stewart Brand to think about the very distant future of humanity. Their flagship project is the construction of a clock to last 10,000 years, which will chime once per century.
The foundation recently hosted a conversation between musician Brian Eno and game designer Will Wright (The Sims, Spore). Haven’t heard the whole thing yet, but the first half hour was fascinating — Eno and Wright mostly discussing generative systems — complexity arising from simple rules. Eno reminisces about the first time he heard Steve Reich perform a pair of tape loops — an inflection point in Eno’s career.
Reich took two identical 1.8-second audio segments and created identical loops out of them, strung them through two decks, and played one slightly slower than the other. Gradually the two segments went out of phase with one another, giving rise to complex and beautiful relationships. The pieces come back into sync 30 minutes later, and the piece ends. Objective correlative: Near the end of the work day, I watched sadly as the J-School hauled its last remaining reel-to-reel tape decks out to the electronic recycling bin, their usefulness behind them.
Wright talks about the Game of Life as a generative system giving rise to complex relationships from a base of a few simple rules, correlates to the Chinese game of Go, which also has very few rules but tremendous complexity. Eno demonstrates a version of “Life” that generates music from the ongoing relationships in the same game.