Jabra Speak 410 USB Speaker / Working as expected

I’ve been attempting to use Jabra 410 USB speaker/microphones with my two Rhasspy satellites (both RPi3B+) and after much trial and error, the microphone quality just isn’t good enough for accurate speech-to-text input. :frowning: I’ll keep the units as speakers/output, because in this respect, they’re great. Tonight I ordered the ReSpeaker USB Mic Array instead. I’m hoping for a much more accurate result.

1 Like

Looks like the Jabra Speak unit was NOT the source of the issue. :astonished:

This morning I switched the back-end speech-to-text engine from pocketsphinx to Kaldi and the recognition rate has shot upwards, immensely. :white_check_mark:

Hello, can you tell me how you proceed to integrate the jabra to home assistant?
I am hesitating to buy a 510 or 710 or 750 habra device.
I would know if it is possible to connect the device to a phone client and to h.a as a media player.

2 Likes

Hi,
Do you have any steps on how you got the jabra 410 USB speaker mic working? I can only seem to get the LED flashing

Worked out what the problem was, I just needed to reboot the rasberrypi, once I did that it was able to see and find the device.

I am not so lucky. Even after a reboot my jabra speak 410 has a white LED blinking rapidly. I’m running it on an Odroid N2. The device is listed under /dev/sbd/by-id/usb-0b0e_Jabra_SPEAK_410_USB_501AA5DDBA72x010900-00
How can i test this device to playback a file?

You’ll need to create an .asoundrc configuration file so the audio device shows up for aplay/arecord to use.

Try to see which device / hardware ID it currently occupies. The problem with USB devices is they can sometimes initialize as different devices. For more detailed USB device init information, check out this thread at Rhasspy Community forum

aplay -L

You’ll need to adjust the ‘card 1’ and ‘device 0’ and “hw:1,0” lines below depending on where the speaker appears.

pcm.jabra {
    type hw
    card 1
    device 0
    rate 48000
    format "S16_LE"
    channels 2
}

pcm.!sysdefault {
  type asym
  playback.pcm {
    type plug
    slave.pcm "jabra"
  }
  capture.pcm {
    type plug
    slave.pcm "hw:1,0"
  }
}

pcm.!default {
  type asym
  playback.pcm {
    type plug
    slave.pcm "jabra"
  }
  capture.pcm {
    type plug
    slave.pcm "hw:1,0"
  }
}

Thank @FredTheFrog for the pointer. But funny enough aplay command is not available. And the lsusb -vvv just spits out a list of random ids.

~ $ lsusb -vvv
Bus 002 Device 002: ID 05e3:0620
Bus 001 Device 001: ID 1d6b:0002
Bus 001 Device 004: ID 0b0e:0412
Bus 001 Device 003: ID 0c45:5101
Bus 001 Device 002: ID 05e3:0610
Bus 002 Device 001: ID 1d6b:0003
ls -la /dev/snd/by-id/usb-0b0e_Jabra_SPEAK_410_USB_501AA5DDBA72x010900-00

Looks like the Jabra ist the 3rd device ID 0b0e:0412

How can i install alsa? Shouldn’t it be there by default?

That really depends. Are you running Home Assistant Supervised, where HA runs within Docker containers? If so, then the host operating system doesn’t require ALSA. ALSA is probably installed within whichever Docker container is running your HA application.

If your host OS is Debian, you can use this command:

sudo apt-get install libasound2 alsa-utils alsa-oss

Yes, I’m running the supervised version as the screenshot shows above. When i log in to the machine using ssh, it looks like i end up in a container too and not on the host itself. Which makes it hard to configure. I will hook up an HDMI cable + keyboard and see if i can login straight to the host.

Hi all, while I understand this might be slightly off-topic as this is about Jabra/Pi/speech recognition, I wanted to ask if anybody is using the Jabra 410 as a regular desktop mic/speaker for Teams and/or Zoom and done any HA integration with it?

What I’m looking for specifically is a way, regardless of conferencing app to detect when the Jabra device is on/off-hook, triggering other automations (busylight, pause audio players etc).

I know there’s a Zoom integration available in HACS, yet I wanted to focus on the device as some of my customers use Teams, others WebEx and such, i.e. it would make sense to focus on the common denominator.

Second, I understand Jabra has an SDK available, however I’m not code savvy enough to whip up an app based on it to monitor the device, so before diving into the deep end of that pool, I wanted to reach out to the community to see if anyone had gone down that path before me?

I use a Jabra 410 daily for Zoom and MS-Teams meetings. I picked up all of mine used from eBay last year. Figured out one of them has a slightly different USB VID:PID assignment, simply because “it’s for MS Teams”. Slightly different firmware, but essentially 99% the same equipment.

Hi guys,

at the moment I am also thinking about voice commands.

But unfortunately no idea where to go. I tried rhasspy before with my Pi Setup, now I have a Mini PC with much more power.

Next is the lack of a good microphone. Also I thought about the Jabra Speak 4xx 5xx 7xx. But I read somewhere that the array is not working under Linux.
Not sure if this is true.
Just want to have the best option for a mic.

So how does the Jabra work in your environments?

Cheers

fb

JabraSpeak 410 is all I use for my three Rhasspy nodes.

Can you share some more insights?

  • How is the voice recognition with other surrounding sounds like Music, TV, Talks, …
  • How is the voice recognition from a farer distance like x meters or from another room
  • How is the voice recognition with surrounding sounds and farer distance from the mic?
  • are all mics of the array used?
  • any recommendations about the settings for rhasspy? if so - which?

Microphone distance from the television is VERY helpful. In the living room, with a television, is the only place it can be troublesome. Funniest is when the spouse never remembers to mute the television (or pause it) before speaking voice commands.

Rhasspy with JabraSpeak 410 easily recognizes voice commands several yards / meters away.

Hello,

just want to give a short reply to that.

I’ve ordered also now a Jabra 410 for my living room. The device is working pretty well. But only for this room. The room is like 15x4m. Other rooms next to this one do not work.
There I will get PI zeros and a respeaker for that.

So after all in comparison to my normal Blue Snowball Microphone, this was a massive increase of the power. I love it! :slight_smile:

Edit to answer my own questions:

  • How is the voice recognition with other surrounding sounds like Music, TV, Talks, …
    → with TV it works good, with other sounds like washing machine, dryer, vacuum robot I need to raise my voice
  • How is the voice recognition from a farer distance like x meters or from another room
    → for the one room its pretty good. From other rooms not. Maybe I can improve the location of the mic
  • How is the voice recognition with surrounding sounds and farer distance from the mic?
    → Also pretty well
  • are all mics of the array used
    → seems like

Cheers

fb

1 Like

I know this is a old thread. What OS are you using I am trying to setup a Hass Sat with Jabre 410 on RPI 3 / Raspian light and I can’t get the Jabre recognized

I can’t speak for anything HA, but when setting up the Jabra 410 and Jabra 510 speakers for Rhasspy and Debian Buster, you had to configure the ALSA sound subsystem to get it working.