InstallAUISuite fails

[black-snow]

InstallAUISuite.sh =>

Would you like voicecommand to try to set itself up? y/n
y
./InstallAUISuite.sh: line 186:  1244 Illegal instruction     voicecommand -s

Google didn't help so I thought I'd ask here. Tried to install it on my old raspi B+ running osmc.

[alecgibson]

The most recent version of voicecommand has been compiled for ARMv7 (Pi 2). To get it to run on the B+ (ARMv6):

1. Navigate to PiAUISuite/VoiceCommand
2. Edit Makefile
3. Change line 5 to read ARCH := arm
4. Run make -B (the flag will force it to recompile)
5. Now try to reinstall the PiAUISuite

[black-snow]

@alecgibson with the changes you suggested I got further. Sadly I now have another issue. I get to this:

First I'm going to say something and see if you hear it
Did you hear anything? (y/n)

But nope, no output. $ speaker-test -c 2 runs just fine. Any guesses?

[black-snow]

$ voicecommand
Opening config file...
ALSA lib pcm_hw.c:1667:(_snd_pcm_hw_open) Invalid value for card
arecord: main:722: audio open error: No such file or directory
Warning: Couldn't read data from file "/dev/shm/out.flac", this makes an empty
Warning: POST.
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100    14    0    14    0     0     38      0 --:--:-- --:--:-- --:--:--    38
rm: cannot remove '/dev/shm/out.flac': No such file or directory
No translation

It also says: You have a formatting error on line 35 of your config file. I'm ignoring that line
I can't figure out which config file is meant.
Running voicecommand w/ sudo it suddenly says error on line 1 ...

//edit: runnign voicecommand now actually does something:

osmc@osmc:~$ voicecommand

Opening config file...
running in continuous mode
verifying command as well
keyword duration is 2 and duration is 3
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 36597    0    14  100 36583      9  25085  0:00:01  0:00:01 --:--:-- 25091
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 38689    0    14  100 38675      7  19478  0:00:01  0:00:01 --:--:-- 19483
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 40122    0   234  100 39888     81  13942  0:00:02  0:00:02 --:--:-- 13941
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 38668    0   185  100 38483     92  19318  0:00:01  0:00:01 --:--:-- 19318
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 38144    0    14  100 38130      9  25881  0:00:01  0:00:01 --:--:-- 25903
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 36882    0    14  100 36868      9  25318  0:00:01  0:00:01 --:--:-- 25338

But I don't hear no "Yes Sir?". And it ain't seem to recognize any commands.

// update: I ran tts "FILLER FILL This program ..." which doesn't output anything neiter.
// update: I ran wget -q -U Mozilla -O "/dev/shm/tmp.mp3" "https://translate.google.com/translate_tts?tl=en&q=speak%20this&ie=UTF-8" manually, received file had a file size of 0 ...
// update: Also checked recording works with arecord - it does.
// update: Seems like hardware="plughw:1,0" is hardcoded in VoiceCommand/speech-recog.sh

[black-snow]

I'll make this a new issue

[bruno1308]

Anything on this? I can't hear anything when I run voicecommand -s (working on Pi 2)

[velkrosmaak]

I'm in the same boat as bruno1308 - Pi2, wheezy image, can't hear anything during the voicecommand setup, but audio plays fine with aplay and speaker-test.

[agustinscaz]

Same problem here, i can't hear anything 👎

[rjturek]

Same problem here. speaker-test and aplay work fine. Hear nothing during setup. Pi2 jessie

[black-snow]

reopening this

There's also this #52, this #49 and this #47

[perskes]

Anything new on this?

[Venaci]

So, the reason (for those following) for not hearing anything, is because Google blocked connecting to it for TTS. Someone, however, cleverly created a wrapper that allows audio to play, though it requires installing Pico2Wave.

I have gotten it working, but it kept looping "what you say baby", until I changed what alecgibson suggested. After that, I'm getting the out.flac errors (which I think may actually be a step back).

To get the audio working, I'm going to reuse the standard install and see if I can't debug what's causing her to mistake the clip for "useless conversation", so to speak, later...

[Venaci]

I've done some digging, and the bug is on line 35 of VoiceCommand/speech-recog.sh

I'm looking into seeing if I can make a reasonable replacement using Pico2Wave.
I don't really know the infrastructure on how it works before that point, but I'm hoping I can just reiterate the speech string to play through tts -> Pico2Wave through a wrapper someone very generously posted elsewhere.

[MisterPascal]

Hey, I#ve got the same problem and I can't fix it.
Mic and Speakers are working, I changed the Google API to VoiceRSS.

Can anyone help me please? :/

[BHEADRICK]

this is my tts script:
`#!/bin/bash

string=$@
lang="en-us"
echo "" > "/dev/shm/speak.mp3"

len=${#string}

wget -q -U Mozilla -O "/dev/shm/tmp.mp3" "http://api.voicerss.org/?key=&hl=${lang}&src=$string"
cat "/dev/shm/tmp.mp3" >> "/dev/shm/speak.mp3"
#now we finally say the whole thing
cat "/dev/shm/speak.mp3" | mpg123 - 1>>/dev/shm/voice.log 2>>/dev/shm/voice.log
`

[projectclass]

Warning: Couldn't read data from file "/dev/shm/noise.flac", this makes an
Warning: empty POST.
how to solve it?

[borhanshuvo]

Warning: Couldn't read data from file "/dev/shm/noise.flac", this makes an
Warning: empty POST.