Talking iPods 194
chrisb33 writes to tell us the next iteration of the iPod may talk you through the menus instead of just relying on text. The Scotsman speculates on this new technology based on a patent filed by Apple in the US. From the article: "The patent reveals the idea is driven largely by safety considerations. It states: 'A user will have difficulty navigating the interface in "eyes-busy" situations. Such activities include, for example, driving an automobile, exercising and crossing the street." The patent also makes clear that text-to-speech technology is likely to spread to other hand-held electronic devices such as mobile phones and palm-top computers."
What would be cool... (Score:5, Interesting)
Is it just me that HATES devices talking? (Score:3, Interesting)
One of my favorite games (Battlefield 2) is almost ruined by the constsant spamming in my ears of "Enemy unit spotted!" "ok" "roger sir" "well done team" etc.
Text is easy to skip, but voices arent, we seem naturally designed to respond to a voice, but we can ignore text. any device, application or game that talks to me just feels like someone nagging me. Besides, what accent will it have? pretty much everyone has at least one accent they hate, are people assuming a US accent is univerally appreciated?
Can you *imagine* how much MORE annoying the office paperclip would be if it spoke to you?
Re:Seems an obvious patent (Score:3, Interesting)
Yeah, I actually posted about that 2 years ago when the shuffle was rumored. Can I dig up my old /. post and call it prior art? ;)
Of course, people thought I was nuts then...
Re:What would be cool... (Score:3, Interesting)
t
Seems they improved speech-quality (Score:2, Interesting)
Speakable Items and VoiceOver (Score:3, Interesting)
I'd like to believe this, but frankly, Apple has been letting their TTS and STT features languish since they were introduced to Mac OSX.
Speakable Items (speech to text commands) are a very simple arrangement: the engine is listening for a finite set of strings at any given time, so error rates are low. Fortunately, the set of strings is gathered from a set of filenames, so it's super-easy to make new strings and organize them by application. Unfortunately, most of the newer applications that are bundled with OSX have no hooks for automation nor sample scripts as speakable items. The speakable item must be an AppleScript or .app, for no discernable reason; I would love to be able to have voice-activated shell scripts without going through some ugly hack of a wrapper script, since it's "Unix" and all.
VoiceOver (text to speech prompting) is also fairly straightforward, but there's limited support and somewhat inconsistent controls. Many of the blind folks I've seen using voice prompts on other devices want their voice prompts to be very fast, even so far as to blur the words together into abstract "earcons." The AppleScript-invoked speech does not honor the OSX talking speed preferences, so the words just ramble on taking forever to finish. The talking is not a separately controllable volume channel, so if you turn up the iTunes, then the TTS voice will start yelling at you to compete, or worse, not be able to escape the iTunes mute control.
This is just a rumor, but for the sake of those who like or need good voice features in their interfaces, I hope it signals a new drive to finish what they've started here.
Could apple become liable??? (Score:2, Interesting)
Not that i'm saying this isn't a bad thing to have audio instructions, but if they are endorsing using your iPod while you are "eyes-engaged" i think they'll be looking for trouble.
if they really want to be safe, they should advise you to familiarize yourself with the iPod before driving, and only use the iPod while stopped.
-IMHO
Phatbox (Score:3, Interesting)
It plugs into a CD changer interface and using the buttons on most head units, you can go forward/backward through playlists, genres, artists, albums, etc.
It tells you where you are in the menus, so you don't have to look at it. (Another good reason is that the CD changer interface is not sophisticated enough to show that data on the head unit...)
I think there's some alternative/oss text-to-speech sw out there, too.