KDE Speech is truncated

[Jonathan Duddington]

Binary package hint: kdeaccessibility

Kubuntu 9.04 i386 Beta (updated to 2/Apr/2009).

1. When I speak text using KDE Text-to-speech, the speech is abruptly truncated. For example the text "1 2 3 4 5 6" is spoken only as "1 2".

2. The "Test" button in the Talker Configuration dialog has no effect (the test message is not heard).

Both these bugs occur with both Festival and eSpeak voices in different languages.

Both these bugs only occur if the KTTSMgr "Audio" option is set to "KDE (Phonon)". If I change the option to "ALSA" then the problem does not occur.

The problem is that "KDE (Phonon)" seems to be the default option, so this error occurs by default. So until it is fixed, then it would be better to change the default "Audio" option to "ALSA".

[Kde-bugs-u (kde-bugs-u)]

Version: (using KDE 4.2.0)
OS: Linux
Installed from: Debian testing/unstable Packages

I'm using kttsd from KDE 4.2

ktts is dropping audio a lot. It would speak a sentence and leave it halfway. I can't think of a better way describing the problem. :-(

In .xsession-errors, I see a lot of these messages:

kttsd(5037) Speaker::doUtterances: State: "usStretched" , "psSynthing" , "jsSpeaking" Type: "utText" Text: "Full text and rfc822 format available."
kttsd(5037) Speaker::doUtterances: State: "usSynthing" , "psSynthing" , "jsSpeaking" Type: "utText" Text: "Acknowledgement sent to "Matthias Urlichs" <email address hidden>:"
kttsd(5037) Speaker::doUtterances: State: "usWaitingSynth" , "psSynthing" , "jsSpeaking" Type: "utText" Text: "Extra info received and forwarded to list."
kttsd(5037) Speaker::doUtterances: Speaker::doUtterances: state usStretched
kttsd(5037) Speaker::createPlayerObject: Speaker::createPlayerObject: Loading "libkttsd_phononplugin"
CLIENT: Task: Task::done()
CLIENT: Task: emitting finished
kttsd(5037) KSpeech::announceEvent: KSpeech:: "slotMarker" : emitting DBUS signal "marker" with appId ":1.107" job number 3 marker type 0 and data "41"
kttsd(5037) Speaker::doUtterances: Speaker::doUtterances: state usSynthing
kttsd(5037) Speaker::doUtterances: Speaker::doUtterances: state usWaitingSynth
kttsd(5037) Speaker::doUtterances: Speaker::doUtterances: queue dump:
kttsd(5037) Speaker::doUtterances: State: "usPlaying" , "psSynthing" , "jsSpeaking" Type: "utText" Text: "Full text and rfc822 format available."
kttsd(5037) Speaker::doUtterances: State: "usSynthing" , "psSynthing" , "jsSpeaking" Type: "utText" Text: "Acknowledgement sent to "Matthias Urlichs" <email address hidden>:"
kttsd(5037) Speaker::doUtterances: State: "usWaitingSynth" , "psSynthing" , "jsSpeaking" Type: "utText" Text: "Extra info received and forwarded to list."
kttsd(5037) Speaker::doUtterances: Speaker::doUtterances: state usPlaying
kttsd(5037) Speaker::doUtterances: Speaker::doUtterances: state usSynthing
kttsd(5037) Speaker::doUtterances: Speaker::doUtterances: state usWaitingSynth
kttsd(5037) Spe...

[jonathan (jonathan-schluessler)]

Duplicate of Bug 180662.

[Knut Karevoll (gnonthgol)]

I have the same problems on my computer. But it does not help to use ALSA to play the sound.

[Jeremy Whiting (jpwhiting)]

Do you get the same results if you set the audio to use alsa output? I've seen many issues like this when using phonon/xine.

[Harald Sitter (apachelogger)]

If it is the combination phonon+xine that is causing this issue, maybe changing to gstreamer would help, otherwise it might be a general phonon issue, or what ktts sends to phonon (granted I don't know anything about the inner workings of ktts though :)).

Please install the package phonon-backend-gstreamer and try if this fixes the issue (you can change to gstreamer via systemsettings -> multimedia -> backend ... you also might have to re-login).

[Harald Sitter (apachelogger)]

This bug was also reported at https://bugs.launchpad.net/kdeaccessibility/+bug/353910

The reports are conflicting though, while the initial reporter states that changing to ALSA resolved the issue, another commenter stated that it didn't help.

Also, if it is the combination phonon+xine that is causing this issue, maybe changing to gstreamer would help, otherwise it might be a general phonon issue, or what ktts sends to phonon (granted I don't know anything about the inner workings of ktts though :)).

[Kde-bugs-u (kde-bugs-u)]

In my case, things are like this:

I have PulseAudio configured to run. In alsa, I have defined to use pulse as the audio device.
AFAIK, pulseaudio is being used by Phonon.

So for me the results were the same. There was heavy dropping.

[personman (personman-145)]

I have this problem as well. I'm running KDE 4.3 Beta 2 with Karmic alpha 2.

[mikey (abc-mikey)]

I'm also experiencing the same issue with ALSA or any other output also dropping speech. It doesn't matter which speech synthesiser is used but I can confirm that for me switching from xine to gstreamer 'fixes' the behaviour. So there's a conflict between kttsd and xine.

AMD64 (64bit Gentoo)
xine-lib-1.1.16.3-r1
kttsd-4.3.1

[mikey (abc-mikey)]

I'd just like to add that I've tried a multi-sentence bit of text now with gstreamer and there is a seperate(?) bug which instead of cutting off the end of the text it just stops at the end of the sentence and indicates that it is still speaking. If you click on next sentence by hand it will move on and speak this then stop before any subsequent sentences. The same applies that changing other settings doesn't change the behaviour.

[mikey (abc-mikey)]

Ignore comment 7 it seems to have just been a problem with switching from xine to gstreamer that a restart fixed. However the xine issue remains.

[Jonathan Thomas (echidnaman)]

We are closing this bug report because it lacks the information we need to investigate the problem, as requested in the previous comments. Please reopen it if you can give us the missing information, and don't hesitate to submit bug reports in the future. To reopen the bug report you can click on the current status, under the "Status" column, and change the status back to "New". Thanks again!

[Lonnie Lee Best (launchpad-startport)]

I wonder if this has to do with "not put quotes around the text" (underneath the hood at the command line level).

For example, using another text-to-speech program (espeak), if you go to a command prompt and type:

espeak How are you doing

all you hear is "how".

but if you type
espeak "How are you doing" (with quotes around what you want it to speak).

You hear the whole sentence.

[Andrew-crouthamel (andrew-crouthamel)]

Hello! Sorry to be the bearer of bad news, but this project has been unmaintained for many years so I am closing this bug.