So really, it's just iTunes or something system level doing the text to speech and a software update on the iPod reading these audiofiles out.

Reply

Honda uses this 'system' on its new Honda-Link iPod dock. It is the most ridiculous and kludgy pile of crap I've ever demo'd, and there was no way I was going to pay several hundred dollars to have it installed in my new car, let alone have some third party company fill my iTunes library with thousands of audio files announcing each song (which requires several hours if you have a half-decent library). And we all know how well our built-in speech synthesizer pronounces unusual words.

Either Apple is going to find a better way to do it, or they never told Honda what they were planning and let them treat their customers like guinea pigs, or this is yet another patent for future safety, like the recent filings for iPods with various front panels similar to other players.

Reply

*yawn* I want This,, I post a pdf or text file of say Shakespeare into the ipod and it reads it back with a nice calm smart sounding British voice.. Not that cheap robo voice junk we have now..

Reply

Why the hell is TUAW advertising HP laptops?

Reply

At least the Archos players, with the Rockbox firmware, do this for ages. And by ages, I really mean real years!

How can they patent such thing?!?

Reply

My best friend is blind and has been exceedingly jealous of how easy it is for sighted people to use anything with a menu - even Microwaves have them now ... Try using an iPod with your eyes closed. She will be thrilled if there is one with any kind of speech! There is a lot of stuff recorded. She'd love to have access to Podcasts, books, and more. She sees computer speech as similar to a walker or wheelchair - not the same as walking - but you can get there.

Reply

What would be nice is if there is an option to spell out the word phonetically for pronunciation purposes, in case it can't manage it automatically.

Reply

...should they use the text to speech functions of Mac OS, chances are high the function could be Mac only, at least for the full feature set.
Besides, they could still afford to produce a cheap display-less iPod, without you not knowing the name of the track you would currently hear (I know there could well be a third parity software for that already, but it makes another step necessary before syncing).
Maybe we could see some easy sync possibility for those PDFs and other texts we store but we would rather hear than read...
On the accessory front: voice remote interacting with your mobile too & Co.

Anyway, this could well make things even more interesting for those using iPods under the well known flawed OS: heck, a little bit arrogance I would just find awesome - "This may work honestly under Windows, but if you want ease of use and/ or the full feature set, then your closest store is..."
Well, how much time left until the next announcement, I can not wait already.

^|°

Reply

I implemented this into podzilla (a gui for the iPod that runs under iPodlinux) utilising flite over a year ago. This meant the synthesis was done on the iPod itsself, not using pre-made clips. It wasn't great, but hey, it worked. This method sounds a lot like what RockBox can do currently.

If you want this functionality now, use RockBox.

One thing you've massively overlooked is that this means the visually impaired would not be more able to use an iPod.

-- coob

Reply

I've had this for -years- on my PhatBox in my car.. All it does is generate .WAV files for the speech, and then upload them to the PhatBox.. The PhatBox uses the .WAV files to announce the track, album, arist, etc.. With more cars having iPod docks, or line-in, I see this as a much safer way to use it..

Reply

Yeah, I've been hoping for this since the shuffle first came out. My original thought was that iTunes would create a short audio file with the Song name, album, and artist, which would be read by the iPod on user request or before a song.

Reply

I'm more psyched for the voice to work both ways. Imagine pressing a button and saying "Artist - Paul Simon" and having it list back to you:

* All Albums
* Graceland
* ect.

(I don't think I actually have any other Paul Simon album, so I don't know what would come after Graceland).
Then you say "Graceland" and it plays Graceland. If it comes across a name it mispronounces (i.e. Bjørk or NIN) you could have an option to correct it in your own voice, and it will say it back in your voice next time (sort of like autodial on a phone).

Very exciting first step (if the rumour is true).

Imagine hands-free iPod control!

Reply

Umm, not == now, whups

Reply

Why not just have properly recorded voice clips, stored as ID3 tags like album art?

The few seconds of recording would only add a size measured in K, and it would sound a whole lot better.

Reply

Wonder how it will handle the over 400 Russian songs I have in ACC format but all the info is in cyrillic? BTW cyrillic (the Russian alphabit) does display just fine in iTunes and on the iPod just don't thing it will be able to pronounce them.

Reply

I like Reg's idea. Adding the clips into the ID3 tags would rock, and then it could be a universal thing.

Reply

I don't see why the iPod itself can't have Text-to-Speech. My brother bought a Panasonic cordless phone that speaks the caller ID.

Reply

I hope this report means that Apple will improve their text-to-speech capabilities. Frankly, the quality of Apple's current TTS system is inferior compared to competing systems. If Apple can't improve it themselves, they should license a system like Natural Voices from AT&T for inclusion in Leopard.

Reply