Wake Word Atom Freeze

[CAVACO-PT]

I have the $13 Atom, it works fine for wake word detection right after turning on wake word switch on their device page. However, after 1 or 2 commands they stop responding ...

if i toogle the switch and wait a bit it comes back , but after saying 2 or 3 commands next to each other it freezes again , and seams to come back after 5 or more minutes ... tried with HA cloud and local , same behavior !!

What are some troubleshooting steps I can try?

[CAVACO-PT]

I have HA in a Proxmox VM , with superviser …

Home Assistant 2023.10.3 Supervisor 2023.10.0 Operating System 11.0 Interface: 20231005.0 - latest

using an ATOM , i just bought 2 units , both have the same issue , the firmware in the atom is

atom-echo-voice-assistant por m5stack Firmware: 2023.10.0b1 (Oct 13 2023, 23:14:59) Hardware: 1.0

I have tried and without wake word … by pressing the button , and the ATOM doesnt have this problems , so its related to the wake word , without it i can press multiple times the button and it responds all the time !!!

[arhills]

I have HAOS in Proxmox VM on an HPt630 16gb ram

Home Assistant 2023.10.3 Supervisor 2023.10.0 Operating System 11.0 Frontend 20231005.0 - latest

Using the M5stack atom echo I get similar issues, works 2 or 3 times then just stops responding to wake word. Sometimes switching wake word of and on again sorts it for another 2 or so times sometimes you just have to wait 10 minutes or so and it starts working again. When it responds it does what I ask 99% of the time.

[andrew-codechimp]

See also this issue on the wyoming openwakeword project. Unclear if it's an ESP device issue or openwakeword https://github.com/rhasspy/wyoming-openwakeword/issues/2

[Photo64]

Home Assistant 2023.10.3 Supervisor 2023.10.0 Operating System 11.0 Frontend 20231005.0 - latest

Same symptoms, M5stack atom echo freezes after 2 or 3 times and no longer responds

[dattas]

Home Assistant 2023.10.3 Supervisor 2023.10.0 Operating System 11.0 Frontend 20231005.0 - latest

Same issues here using an M5 ATOM echo, I also tried Porcupine1 for good measure and it has the same issue. I can see in the addon logs for openwakeword/porcupine1 , Whisper, then Piper that the flow goes all the way through all of them even showing a success on Piper for the response but no output on the device:

INFO:wyoming_faster_whisper.handler: What time is it?

DEBUG:wyoming_piper.handler:input: {'text': 'It is currently a test message.'}
DEBUG:wyoming_piper.handler:/tmp/tmp32gzdqbs/1697498762860832160.wav
DEBUG:wyoming_piper.handler:Completed request
DEBUG:wyoming_piper.handler:Sent info

Turning off Wakeword detection and using the button works every time for me as well.

The ESPHome logs for the ATOM are incomplete as this is all I can ever get them to display:

ets Jun  8 2016 00:22:57

rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)
configsip: 188777542, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:2
load:0x3fff0030,len:6608
load:0x40078000,len:15060
ho 0 tail 12 room 4
load:0x40080400,len:3816
entry 0x40080698
I (30) boot: ESP-IDF 4.4.5 2nd stage bootloader
I (30) boot: compile time 23:16:27
I (30) boot: chip revision: v1.1
I (33) boot.esp32: SPI Speed      : 40MHz
I (37) boot.esp32: SPI Mode       : DIO
I (42) boot.esp32: SPI Flash Size : 4MB
I (46) boot: Enabling RNG early entropy source...
I (52) boot: Partition Table:
I (55) boot: ## Label            Usage          Type ST Offset   Length
I (63) boot:  0 otadata          OTA data         01 00 00009000 00002000
I (70) boot:  1 phy_init         RF data          01 01 0000b000 00001000
I (78) boot:  2 app0             OTA app          00 10 00010000 001c0000
I (85) boot:  3 app1             OTA app          00 11 001d0000 001c0000
I (92) boot:  4 nvs              WiFi data        01 02 00390000 0006d000
I (100) boot: End of partition table
I (104) esp_image: segment 0: paddr=00010020 vaddr=3f400020 size=2adc4h (175556) map
I (176) esp_image: segment 1: paddr=0003adec vaddr=3ffb0000 size=0300ch ( 12300) load
I (181) esp_image: segment 2: paddr=0003de00 vaddr=40080000 size=02218h (  8728) load
I (186) esp_image: segment 3: paddr=00040020 vaddr=400d0020 size=b6b84h (748420) map
I (461) esp_image: segment 4: paddr=000f6bac vaddr=40082218 size=1291ch ( 76060) load
I (503) boot: Loaded app from partition at offset 0x10000
I (503) boot: Disabling RNG early entropy source...
I (514) cpu_start: Pro cpu up.
I (515) cpu_start: Starting app cpu, entry point is 0x4008237c
I (0) cpu_start: App cpu up.
I (531) cpu_start: Pro cpu start user code
I (531) cpu_start: cpu freq: 160000000
I (531) cpu_start: Application information:
I (536) cpu_start: Project name:     m5stack-atom-echo
I (541) cpu_start: App version:      2023.10.0b1
I (547) cpu_start: Compile time:     Oct 13 2023 23:15:00
I (553) cpu_start: ELF file SHA256:  ab8d6202b50ea1e3...
I (559) cpu_start: ESP-IDF:          4.4.5
I (564) cpu_start: Min chip rev:     v0.0
I (568) cpu_start: Max chip rev:     v3.99 
I (573) cpu_start: Chip rev:         v1.1
I (578) heap_init: Initializing. RAM available for dynamic allocation:
I (585) heap_init: At 3FFAE6E0 len 00001920 (6 KiB): DRAM
I (591) heap_init: At 3FFB81B8 len 00027E48 (159 KiB): DRAM
I (597) heap_init: At 3FFE0440 len 00003AE0 (14 KiB): D/IRAM
I (604) heap_init: At 3FFE4350 len 0001BCB0 (111 KiB): D/IRAM
I (610) heap_init: At 40094B34 len 0000B4CC (45 KiB): IRAM
I (618) spi_flash: detected chip: gd
I (621) spi_flash: flash io: dio
I (626) cpu_start: Starting scheduler on PRO CPU.
I (0) cpu_start: Starting scheduler on APP CPU.

If I use wifi to try to get the logs (using 2023.10.0b2, using 2023.9.3 it gives me a stack trace about dictionary objects having no attribute move_to_end File "/esphome/esphome/components/substitutions/__init__.py", line 143, in do_substitution_pass config.move_to_end(CONF_SUBSTITUTIONS, False)):

INFO ESPHome 2023.10.0b2
INFO Reading configuration /config/esphome/m5stack-atom-echo-8a2640.yaml...
INFO Updating https://github.com/esphome/esphome.git@pull/5230/head
INFO Starting log output from m5stack-atom-echo-8a2640.local using esphome API
WARNING m5stack-atom-echo-8a2640.local: Connection error occurred: [Errno 104] Connection reset by peer
WARNING Can't connect to ESPHome API for m5stack-atom-echo-8a2640.local: Connection closed (APIConnectionError)
INFO Trying to connect to m5stack-atom-echo-8a2640.local in the background
WARNING m5stack-atom-echo-8a2640.local: Connection error occurred: [Errno 104] Connection reset by peer
WARNING m5stack-atom-echo-8a2640.local: Connection error occurred: [Errno 104] Connection reset by peer
WARNING m5stack-atom-echo-8a2640.local: Connection error occurred: [Errno 104] Connection reset by peer

If I try to compile the firmware myself on 2023.10.0b2 I always get this error:

INFO ESPHome 2023.10.0b2
INFO Reading configuration /config/esphome/m5stack-atom-echo-8a2640.yaml...
INFO Updating https://github.com/esphome/esphome.git@pull/5230/head
INFO Generating C++ source...
INFO Updating https://github.com/espressif/esp-adf@v2.5
INFO Updating submodules (components/esp-sr, components/esp-adf-libs) for https://github.com/espressif/esp-adf@v2.5
INFO Updating https://github.com/espressif/esp-dsp@v1.2.0
INFO Compiling app...
Processing m5stack-atom-echo-8a2640 (board: m5stack-atom; framework: espidf; platform: platformio/espressif32@5.4.0)
--------------------------------------------------------------------------------
HARDWARE: ESP32 240MHz, 320KB RAM, 4MB Flash
 - framework-espidf @ 3.40405.230623 (4.4.5) 
 - tool-cmake @ 3.16.4 
 - tool-ninja @ 1.7.1 
 - toolchain-esp32ulp @ 2.35.0-20220830 
 - toolchain-xtensa-esp32 @ 8.4.0+2021r2-patch5
Creating a new virtual environment for IDF Python dependencies
Installing ESP-IDF's Python dependencies
Collecting urllib3<2
  Downloading urllib3-1.26.17-py2.py3-none-any.whl (143 kB)
Collecting cryptography<35.0.0,>=2.1.4
  Downloading cryptography-3.4.8-cp36-abi3-manylinux_2_24_x86_64.whl (3.0 MB)
Collecting future>=0.18.3
  Downloading future-0.18.3.tar.gz (840 kB)
Collecting pyparsing<2.4.0,>=2.0.3
  Downloading pyparsing-2.3.1-py2.py3-none-any.whl (61 kB)
Collecting kconfiglib~=13.7.1
  Downloading kconfiglib-13.7.1-py2.py3-none-any.whl (145 kB)
Collecting idf-component-manager~=1.0
  Downloading idf_component_manager-1.4.1-py2.py3-none-any.whl (140 kB)
Collecting cffi>=1.12
  Downloading cffi-1.16.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (443 kB)
Collecting schema
  Downloading schema-0.7.5-py2.py3-none-any.whl (17 kB)
Collecting requests-file
  Downloading requests_file-1.5.1-py2.py3-none-any.whl (3.7 kB)
Collecting click
  Downloading click-8.1.7-py3-none-any.whl (97 kB)
Collecting contextlib2>0.6.0
  Downloading contextlib2-21.6.0-py2.py3-none-any.whl (13 kB)
Collecting requests<3
  Downloading requests-2.31.0-py3-none-any.whl (62 kB)
Collecting colorama
  Downloading colorama-0.4.6-py2.py3-none-any.whl (25 kB)
Collecting six
  Downloading six-1.16.0-py2.py3-none-any.whl (11 kB)
Collecting tqdm<5
  Downloading tqdm-4.66.1-py3-none-any.whl (78 kB)
Collecting packaging
  Downloading packaging-23.2-py3-none-any.whl (53 kB)
Collecting requests-toolbelt
  Downloading requests_toolbelt-1.0.0-py2.py3-none-any.whl (54 kB)
Collecting pyyaml>5.2
  Downloading PyYAML-6.0.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (738 kB)
Collecting cachecontrol[filecache]>0.12.6
  Downloading cachecontrol-0.13.1-py3-none-any.whl (22 kB)
Collecting msgpack>=0.5.2
  Downloading msgpack-1.0.7-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (530 kB)
Collecting filelock>=3.8.0
  Downloading filelock-3.12.4-py3-none-any.whl (11 kB)
Collecting pycparser
  Downloading pycparser-2.21-py2.py3-none-any.whl (118 kB)
Collecting charset-normalizer<4,>=2
  Downloading charset_normalizer-3.3.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (138 kB)
Collecting certifi>=2017.4.17
  Downloading certifi-2023.7.22-py3-none-any.whl (158 kB)
Collecting idna<4,>=2.5
  Downloading idna-3.4-py3-none-any.whl (61 kB)
Using legacy 'setup.py install' for future, since package 'wheel' is not installed.
Installing collected packages: urllib3, idna, charset-normalizer, certifi, requests, msgpack, six, pycparser, filelock, contextlib2, cachecontrol, tqdm, schema, requests-toolbelt, requests-file, pyyaml, pyparsing, packaging, colorama, click, cffi, kconfiglib, idf-component-manager, future, cryptography
    Running setup.py install for future: started
    Running setup.py install for future: finished with status 'done'
Successfully installed cachecontrol-0.13.1 certifi-2023.7.22 cffi-1.16.0 charset-normalizer-3.3.0 click-8.1.7 colorama-0.4.6 contextlib2-21.6.0 cryptography-3.4.8 filelock-3.12.4 future-0.18.3 idf-component-manager-1.4.1 idna-3.4 kconfiglib-13.7.1 msgpack-1.0.7 packaging-23.2 pycparser-2.21 pyparsing-2.3.1 pyyaml-6.0.1 requests-2.31.0 requests-file-1.5.1 requests-toolbelt-1.0.0 schema-0.7.5 six-1.16.0 tqdm-4.66.1 urllib3-1.26.17
Reading CMake configuration...
-- Found Git: /usr/bin/git (found version "2.30.2") 
-- The C compiler identification is GNU 8.4.0
-- Check for working C compiler: /data/cache/platformio/packages/toolchain-xtensa-esp32/bin/xtensa-esp32-elf-gcc
-- Check for working C compiler: /data/cache/platformio/packages/toolchain-xtensa-esp32/bin/xtensa-esp32-elf-gcc -- broken
-- Configuring incomplete, errors occurred!
See also "/data/build/m5stack-atom-echo-8a2640/.pioenvs/m5stack-atom-echo-8a2640/CMakeFiles/CMakeOutput.log".
See also "/data/build/m5stack-atom-echo-8a2640/.pioenvs/m5stack-atom-echo-8a2640/CMakeFiles/CMakeError.log".

fatal: not a git repository (or any parent up to mount point /)
Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set).
CMake Error at /data/cache/platformio/packages/tool-cmake/share/cmake-3.16/Modules/CMakeTestCCompiler.cmake:60 (message):
  The C compiler

    "/data/cache/platformio/packages/toolchain-xtensa-esp32/bin/xtensa-esp32-elf-gcc"

  is not able to compile a simple test program.

  It fails with the following output:

    Change Dir: /data/build/m5stack-atom-echo-8a2640/.pioenvs/m5stack-atom-echo-8a2640/CMakeFiles/CMakeTmp

    Run Build Command(s):/data/cache/platformio/packages/tool-ninja/ninja cmTC_67dae && [1/2] Building C object CMakeFiles/cmTC_67dae.dir/testCCompiler.c.obj
    [2/2] Linking C executable cmTC_67dae
    FAILED: cmTC_67dae 
    : && /data/cache/platformio/packages/toolchain-xtensa-esp32/bin/xtensa-esp32-elf-gcc -mlongcalls -Wno-frame-address   CMakeFiles/cmTC_67dae.dir/testCCompiler.c.obj  -o cmTC_67dae   && :
    /data/cache/platformio/packages/toolchain-xtensa-esp32/bin/../lib/gcc/xtensa-esp32-elf/8.4.0/../../../../xtensa-esp32-elf/bin/ld: cannot find CMakeFiles/cmTC_67dae.dir/testCCompiler.c.obj: No such file or directory
    collect2: error: ld returned 1 exit status
    ninja: build stopped: subcommand failed.

  CMake will not be able to correctly generate this project.
Call Stack (most recent call first):
  /data/cache/platformio/packages/framework-espidf/tools/cmake/project.cmake:298 (__project)
  CMakeLists.txt:3 (project)

========================= [FAILED] Took 22.42 seconds =========================

[balloob]

This has been fixed in latest ESPHome beta 3.

[notownblues]

I can confirm Wake Word is now rock solid on ESPhome beta. Thank you 🙏🏻

[CAVACO-PT]

was this already in the release that just got out of ESP 10.0 , or only on beta ?

[shampeon]

was this already in the release that just got out of ESP 10.0 , or only on beta ?

Assuming it's this PR, it was in 10.0.

[shampeon]

Spoke too soon. I'm on 2023.10.1 and my ATOM Echoes stop responding after a period of time. Killing power and restarting them has no effect.

INFO Successfully connected to 10.10.10.119
[11:59:14][D][voice_assistant:366]: State changed from IDLE to START_MICROPHONE
[11:59:14][D][voice_assistant:372]: Desired state set to WAIT_FOR_VAD
[11:59:14][D][voice_assistant:159]: Starting Microphone
[11:59:14][D][voice_assistant:366]: State changed from START_MICROPHONE to STARTING_MICROPHONE
[11:59:14][D][esp-idf:000]: I (11441) I2S: DMA Malloc info, datalen=blocksize=1024, dma_buf_count=4

[11:59:14][D][voice_assistant:366]: State changed from STARTING_MICROPHONE to WAIT_FOR_VAD
[11:59:14][D][voice_assistant:176]: Waiting for speech...
[11:59:14][D][voice_assistant:366]: State changed from WAIT_FOR_VAD to WAITING_FOR_VAD
[11:59:14][D][voice_assistant:189]: VAD detected speech
[11:59:14][D][voice_assistant:366]: State changed from WAITING_FOR_VAD to START_PIPELINE
[11:59:14][D][voice_assistant:372]: Desired state set to STREAMING_MICROPHONE
[11:59:14][D][voice_assistant:206]: Requesting start...
[11:59:14][W][voice_assistant:217]: Could not request start.
[11:59:14][D][light:036]: 'M5Stack Atom Echo 888038' Setting:
[11:59:14][D][light:051]:   Brightness: 100%
[11:59:14][D][light:059]:   Red: 100%, Green: 0%, Blue: 0%
[11:59:14][D][voice_assistant:366]: State changed from START_PIPELINE to IDLE
[11:59:14][D][voice_assistant:372]: Desired state set to IDLE
[11:59:15][D][light:036]: 'M5Stack Atom Echo 888038' Setting:
[11:59:15][D][light:051]:   Brightness: 100%
[11:59:15][D][light:059]:   Red: 100%, Green: 0%, Blue: 100%
[11:59:23][D][api:102]: Accepted 192.168.0.8
[11:59:23][D][api.connection:1032]: Home Assistant 2023.10.3 (192.168.0.8): Connected successfully

When it worked, these messages would repeat in a loop, but now the unit stays in this state indefinitely.

[shampeon]

When it worked, these messages would repeat in a loop, but now the unit stays in this state indefinitely.

This is apparently due to a deadlock, fixed in 1.8.0 of the OpenWakeWord Docker container. https://github.com/rhasspy/wyoming-openwakeword/issues/2