Hi,

I don't know if this is the latest version of the package but it seems the .sar has a bug and it's not loading the "Speech-Phonetics.st" file. I'm attaching a version that loads without errors on Squeak 4.2. I don't know if it works correctly, I only tested the example below:

Speaker manWithHead say: 'This is my voice. Can you see my lips?'.

I hope this helps.

Cheers Richo

On Sun, Jan 2, 2011 at 12:41 AM, Dimitry Golubovsky golubovsky@gmail.comwrote:

Hi,

What is the current status of text to speech synthesis in Squeak 4?

Googling the topic gave me only the page [1] which in turn refers to the KlattSynthesizer page [2], and also the Speech package page [3]; both Wiki pages were last updated years ago.

I tried to install the "Speech" package from SqueakMap, but it failed to install. Does there exist a newer version or other packages for speech synthesis?

In case a newer version does not exist, I don't mind trying to fix the package myself.

I use Squeak 4.1, default build, with "stock" image from squeak.org under Linux.

Thanks for any ideas.

---

[1] http://www.squeak.org/Features/MultiMedia/ [2] http://wiki.squeak.org/squeak/1112 [3] http://wiki.squeak.org/squeak/651 -- Dimitry Golubovsky

Anywhere on the Web

On Sat, 1 Jan 2011, Dimitry Golubovsky wrote:

Hi,

What is the current status of text to speech synthesis in Squeak 4?

Googling the topic gave me only the page [1] which in turn refers to the KlattSynthesizer page [2], and also the Speech package page [3]; both Wiki pages were last updated years ago.

I tried to install the "Speech" package from SqueakMap, but it failed to install. Does there exist a newer version or other packages for speech synthesis?

In case a newer version does not exist, I don't mind trying to fix the package myself.

I use Squeak 4.1, default build, with "stock" image from squeak.org under Linux.

Thanks for any ideas.

Speech is available on SqueakSource, but there's a problem with the order of class initializations, which should be fixed. Until then you can try the following script which works around the issue:

[ [ Installer squeaksource project: 'Speech'; install: 'SharedPool-Speech'; install: 'Speech' ] on: MessageNotUnderstood do: [ :error | "We get here, because LipsMorph class >> #initialize assumes that PhonemeSet is already initialized." #(PhonemeSet KlattSynthesizer LiljencrantsFant PhonemeRecord Voice) do: [ :each | (Smalltalk at: each) initialize ]. error retry ] ] on: Warning do: [ :warning | "We get here, because we manipulated the loading process above." warning resume ].

If the above script was successfully evaluated you can try this:

Speaker default say: 'Hi!'.

See the class side of Speaker for other examples.

Levente

---

[1] http://www.squeak.org/Features/MultiMedia/ [2] http://wiki.squeak.org/squeak/1112 [3] http://wiki.squeak.org/squeak/651 -- Dimitry Golubovsky

Anywhere on the Web