Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Stuck on "Say something" #62

Closed
5ika opened this issue Nov 10, 2016 · 19 comments
Closed

Stuck on "Say something" #62

5ika opened this issue Nov 10, 2016 · 19 comments
Labels
help wanted

Comments

@5ika
Copy link

5ika commented Nov 10, 2016 •
edited

Merci beaucoup pour ce projet ! Ça fait longtemps que je cherchais un équivalent.

French
J'ai suivi toutes les instructions pour l'installation (dev) sur une raspberry pi.
J'ai cependant un souci : l'application reste bloqué sur "Say something" et ne semble pas reconnaître ce que je dis.
Pourtant, j'arrive bien a récupéré le son par mon micro avec rec puis à l'écouter avec play.
Comme le mode debug n'indique rien de plus, je ne sais pas trop comment résoudre cela.
Auriez-vous une piste ?

English
I followed the instructions for the installation on a raspberry pi.
But i have an issue : the application gets stuck on "Say something" and it does not seem to recognize what i am saying.
However, I can recover a sound with rec then play.
Debug mode indicates nothing more and I don't know how to resolve this.
Can you help me ?
@Sispheor
Copy link
Member

Can you please leave issues only in English or in both languages.

Did you try the STT from the GUI first?
Did you try to speak closely to your microphone? I think about a bad
sensitivity settings in your os.

Le 10 nov. 2016 1:00 PM, "5ika" notifications@github.com a écrit :

Merci beaucoup pour ce projet ! Ça fait longtemps que je cherchais un
équivalent.

J'ai suivi toutes les instructions pour l'installation (dev) sur une
raspberry pi.
J'ai cependant un souci : l'application reste bloqué sur "Say something"
et ne semble pas reconnaître ce que je dis.
Pourtant, j'arrive bien a récupéré le son par mon micro avec rec puis à
l'écouter avec play.
Comme le mode debug n'indique rien de plus, je ne sais pas trop comment
résoudre cela.
Auriez-vous une piste ?

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#62, or mute the
thread
https://github.com/notifications/unsubscribe-auth/ACdEXR1csnXAvdK73U7pIhAmKDuiAe_Oks5q8wdmgaJpZM4KulQb
.

@Sispheor
Copy link
Member

Also there is a dedicated section in our doc with some stuff to do on a rpi
like install pulse audio server.

Le 10 nov. 2016 1:06 PM, "Nicolas Marcq" nico.marcq@gmail.com a écrit :

Can you please leave issues only in English or in both languages.

Did you try the STT from the GUI first?
Did you try to speak closely to your microphone? I think about a bad
sensitivity settings in your os.

Le 10 nov. 2016 1:00 PM, "5ika" notifications@github.com a écrit :

Merci beaucoup pour ce projet ! Ça fait longtemps que je cherchais un
équivalent.

J'ai suivi toutes les instructions pour l'installation (dev) sur une
raspberry pi.
J'ai cependant un souci : l'application reste bloqué sur "Say something"
et ne semble pas reconnaître ce que je dis.
Pourtant, j'arrive bien a récupéré le son par mon micro avec rec puis à
l'écouter avec play.
Comme le mode debug n'indique rien de plus, je ne sais pas trop comment
résoudre cela.
Auriez-vous une piste ?

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#62, or mute the
thread
https://github.com/notifications/unsubscribe-auth/ACdEXR1csnXAvdK73U7pIhAmKDuiAe_Oks5q8wdmgaJpZM4KulQb
.

@5ika
Copy link
Author

5ika commented Nov 10, 2016

I'm sorry, I updated my message with english language.

Yes, I tried STT from the GUI and there is no response.
I tried to speak close to the mic without success.
Oddly, the application seems to recognize when i'm saying "Kalliope" so it works ! But nothing is possible then.

@5ika 5ika changed the title Bloqué sur "Say something" Stuck on "Say something" Nov 10, 2016
@Sispheor
Copy link
Member

Really weird, if the trigger is detected, the STT should work exactly the same way.

What is you OS version? The last Raspbian Jessie? It's the only one we have tested so far.

@5ika
Copy link
Author

5ika commented Nov 10, 2016

I'm on Raspbian Jessie. Fresh install with the last official image.

@Sispheor
Copy link
Member

The network is well configured? Are you behind a proxy server? The RPI has a full access to the internet?
DNS settings ok?

@5ika
Copy link
Author

5ika commented Nov 10, 2016

The RPI has all access to the network. I ping network without problem. No proxy, DNS okay.
I tried with all the STT engines.

@Sispheor
Copy link
Member

Ok, I need to perform some test, maybe it is due to our last changes. I'll be back to you in a couple minutes.

@Sispheor
Copy link
Member

So, I've just tested the last on my laptop (In a VM on Ubuntu 16.04) and it works with the last stable release (master branch).

I've not tested in my Rpi, I cannot do it before tomorrow.
You not answered the question I've asked about the Rpi dedicated doc? Did you follow it?

I can only recommend to try with another microphone from now. I'll perform a full install tomorrow to confirm that issue.

@5ika
Copy link
Author

5ika commented Nov 10, 2016 •
edited

Ok, thank you for your time.
Of course, I've followed the RPI doc (if you talk about this page).
I'll try to find another microphone and try again.

@Sispheor
Copy link
Member

I'll test with my own microphone tomorow on a fresh install of Raspbian Jessie. I keep this issue open until I can confirm it is not a Kalliope problem.

Anyway, the automated install is fully functional on Ubuntu 16.04. So if you want to test kalliope and play with it by creating your brain you can install the project in a VM and then copy your brains when the Rpi is fixed.

@Sispheor
Copy link
Member

So I've tested on my own RPi with a fresh install.
I had the same issue about the stuck "say something".
To fix this, I ran alsamixer, selected my sound card by pressing F6and then moved up the level sensibility of my mic.

Hope this help. I'll add this into the Rpi doc. If you can confirm that is ok from your side too.

@5ika
Copy link
Author

5ika commented Nov 12, 2016

Still the same... I moved the sensibility to the top.
Maybe the problem comes from my usb microphone. I have to tried with another one.

@Skunnyk
Copy link

Skunnyk commented Nov 20, 2016

Hi,

I had the same kind of problem : it doesn't understand "kalioppe" when waiting for "Say something" on start. I tried in the gui, microphone / tts / stt works fine there.
So I updated the pmdl_file in settings.yml to use Jarviss pmdl, now it works with "Jarviss" :)

- pmdl_file: "trigger/snowboy/resources/kalliope-FR-6samples.pmdl
+ pmdl_file: "trigger/snowboy/resources/jarviss.pmdl"

Ps: I'm French, so maybe I don't enunciate "kalioppe" correctly.

@Sispheor
Copy link
Member

Look at the demo of kalliope in French to know how to pronounce it.
https://youtu.be/t4J42yO2rkM . It's "kalliopé".

Le 20 nov. 2016 3:36 PM, "Skunnyk" notifications@github.com a écrit :

Hi,

I had the same kind of problem : it doesn't understand "kalioppe" when
waiting for "Say something" on start. I tried in the gui, microphone / tts
/ stt works fine there.
So I updated the pmdl_file in settings.yml to use Jarviss pmdl, now it
works with "Jarviss" :)

  • pmdl_file: "trigger/snowboy/resources/kalliope-FR-6samples.pmdl
  • pmdl_file: "trigger/snowboy/resources/jarviss.pmdl"

Ps: I'm French, so maybe I don't enunciate "kalioppe" correctly.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#62 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ACdEXVfgrjkD0s7DI51Bhtm6LXm-6uIuks5rAFsKgaJpZM4KulQb
.

@Sispheor
Copy link
Member

And by the way. "Say something" appears when the hotword has been detected.
So it's not really the same case for you, doesn't it?

Le 20 nov. 2016 3:40 PM, "Nicolas Marcq" nico.marcq@gmail.com a écrit :

Look at the demo of kalliope in French to know how to pronounce it.
https://youtu.be/t4J42yO2rkM . It's "kalliopé".

Le 20 nov. 2016 3:36 PM, "Skunnyk" notifications@github.com a écrit :

Hi,

I had the same kind of problem : it doesn't understand "kalioppe" when
waiting for "Say something" on start. I tried in the gui, microphone / tts
/ stt works fine there.
So I updated the pmdl_file in settings.yml to use Jarviss pmdl, now it
works with "Jarviss" :)

  • pmdl_file: "trigger/snowboy/resources/kalliope-FR-6samples.pmdl
  • pmdl_file: "trigger/snowboy/resources/jarviss.pmdl"

Ps: I'm French, so maybe I don't enunciate "kalioppe" correctly.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#62 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ACdEXVfgrjkD0s7DI51Bhtm6LXm-6uIuks5rAFsKgaJpZM4KulQb
.

@Skunnyk
Copy link

Skunnyk commented Nov 20, 2016

Oh, yes, "say something" appears after the hotword has been detected … and my problem was the hotword detection. Sorry :)

@Sispheor
Copy link
Member

@5ika ping.

@Sispheor
Copy link
Member

I think I can close this. It seems to be attached to the used microphone.
Re open the issue if needed @5ika .

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
help wanted
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@Sispheor @Skunnyk @5ika