#ESP32 VA Project
1 messages ยท Page 1 of 1 (latest)
๐
have you used ws812s with esphome before?
kind of , sort of ... but not really ๐คฃ
heres the base code you need to add to get the lights in the config
light:
- platform: esp32_rmt_led_strip
rgb_order: GRB
pin: GPIO13
num_leds: 30
rmt_channel: 0
chipset: ws2812
name: "My Light"
id: led
obvs change gpio and number of leds
then tag this on the end
effects:
- pulse:
name: "Slow Pulse"
transition_length: 250ms
update_interval: 250ms
min_brightness: 50%
max_brightness: 100%
- pulse:
name: "Fast Pulse"
transition_length: 100ms
update_interval: 100ms
min_brightness: 50%
max_brightness: 100%
ooh
light:
- platform: esp32_rmt_led_strip
rgb_order: GRB
pin: GPIO35
num_leds: 30
rmt_channel: 0
chipset: ws2812
name: "My Light"
id: led
effects:
- pulse:
name: "Slow Pulse"
transition_length: 250ms
update_interval: 250ms
min_brightness: 50%
max_brightness: 100%
- pulse:
name: "Fast Pulse"
transition_length: 100ms
update_interval: 100ms
min_brightness: 50%
max_brightness: 100%
How does it know which effect to do?
ive not got to that bit yet ... ๐
i'm old and slow and it's almost 2am
these are examples of where to put them. ignore the media player parts
on_listening:
- homeassistant.service:
service: media_player.play_media
data:
media_content_id: media-source://media_source/local/vad2.wav
media_content_type: music
entity_id: media_player.colins_voice
- light.turn_on:
id: led
blue: 100%
red: 0%
green: 0%
effect: "Slow Pulse"
these are all in the voice_assistant: block
aghhhh .... your using arduino framework i'm guessing
i have no idea
esphome:
name: esp32-mic-speaker
friendly_name: esp32-mic-speaker
on_boot:
- priority: -100
then:
- wait_until: api.connected
- delay: 1s
- if:
condition:
switch.is_on: use_wake_word
then:
- voice_assistant.start_continuous:
if it is , then you will need to use noepixel and not rmt light component which is here https://esphome.io/components/light/neopixelbus its similar setup
will be below that bit
esp32:
board: esp32dev
framework:
type: esp-idf
version: recommended
no but I dont have a light setup to try yet, gotta solder a header onto a strip and such, done really feel like soldering more today lol
might do it tomorrow though
ohhh when you said the light part isn't right... i thought you'd got an error lol
ok in that case... ignore the neopixel bus bit ๐
alrighty ill try out the LEDs tomorrow, I really hope the ram increase will fix my issue
and the last bit is
on_stt_vad_end:
- light.turn_on:
id: led
blue: 100%
red: 0%
green: 0%
effect: "Fast Pulse"
on_tts_start:
- light.turn_on:
id: led
blue: 100%
red: 0%
green: 0%
brightness: 100%
effect: none
and thats for when its responding
so thats everything to get a light working , you can then tweak it
and im guessing just the GPIO pin will be the data pin to the leds
then gotta get common grounds and stuff
when i run mine with whisper on baseint8 its using 18% of 8gb ram and 20% cpu
damnnnn
but thats on a fresh install of HA
nothing else running
and thats only whilst it's processing the voice command
what are you using for mic sensitivity/volume settings? mine has a really hard time understanding me 90% of the time
supression 1 vol 4
but to be honest i've got them set at all different levels on different builds, ive not really had any issues no matter what the setting
have you got auto_gain: in the config ?
did you remember to turn off the audio recording ? and restrt HA after
I removed the audio saving directory from config
the speaker fart noise is electrical interference
ill direct solder it all when im done prototyping
try removing the auto_gane from your config and slo remove the whole debug config from congiguration.yaml and restart HA otherwise it will still be trying to capture the audio and may cause problems
*gain
right too tired to type... need sleep ๐ ... i'll catch you tomorrow if you are about... i say tomorrow .. i'll be up in 4 hours ๐
just comment the line out
done
did you get it working any better ?
just got the ram today, turns out i did have 16gb the whole time.
so my ram wasnt actually maxing out..
got 128 now though i doubt it will make a difference
still slow and gotta say the wake word multiple times to activate it
I will but also just picked up an atom echo