Updating to 4.0

[mikeydbrad]

Hello there! Super glad to have found this plugin, as I want to use SAM in my project. I am attempting to upgrade this plugin to 4.0, and have converted a bunch of the script file. I'm still running into a few issues, and my lack of Godot knowledge is showing itself.

Are you interested at all in updating this plugin to Godot 4.0?

Here is the copy/paste of what I have edited so far - struggling with the fix for Line 37 on preloading the gdsam.gdns file.

@tool
extends Node

signal loaded_buffer(buffer, callback)
signal finished_speaking()
signal finished_phrase()

class BufferCallback:
    var audio_stream_override: AudioStream = null

const DEFAULT_SPEED = 72
const DEFAULT_PITCH = 64
const DEFAULT_MOUTH = 128
const DEFAULT_THROAT = 128
const DEFAULT_VOLUME = 1.0
const DEFAULT_SINGING = false
const DEFAULT_PHONETIC = false
const DEFAULT_AUDIO_BUS = "Master"
const MAX_CHARS = 90

@export_range(0, 255) var speed: int = DEFAULT_SPEED
@export_range(0, 255) var pitch: int = DEFAULT_PITCH
@export_range(0, 255) var mouth: int = DEFAULT_MOUTH
@export_range(0, 255) var throat: int = DEFAULT_THROAT
@export var singing: bool = DEFAULT_SINGING
@export var phonetic: bool = DEFAULT_PHONETIC
@export_range(, 0, 1) var volume: float = DEFAULT_VOLUME
@export var audio_bus = DEFAULT_AUDIO_BUS

var _current_player
var _interrupt = false
var _player
var _player2D
var _sample
var _queue = []

@onready var _gdsam = preload("res://Addons/gdsam/bin/gdsam.gdns").new()

func _ready():
    _player = AudioStreamPlayer.new()
    _player2D = AudioStreamPlayer2D.new()
    _sample = AudioStreamSample.new()
    _sample.mix_rate = 22050
    _sample.format = AudioStreamSample.FORMAT_8_BITS
    add_child(_player)
    add_child(_player2D)

func speak(input: String, positional: bool = false):
    _interrupt = false
    _configure_sam()
    _current_player = _player if not positional else _player2D
    _configure_player(_current_player)
    _enqueue_phrases(input)
    _process_queue()

func speak2D(input: String, position: Vector2):
    _player2D.position = position
    speak(input, true)

func interrupt():
    _interrupt = true
    if _current_player:
        _current_player.stop()

func stop():
    if _player.playing:
        _player.stop()
    if _player2D.playing:
        _player2D.stop()

func get_player() -> AudioStreamPlayer:
    return _player as AudioStreamPlayer

func get_player2D() -> AudioStreamPlayer2D:
    return _player2D as AudioStreamPlayer2D

func set_voice_default():
    speed = DEFAULT_SPEED
    pitch = DEFAULT_PITCH
    mouth = DEFAULT_MOUTH
    throat = DEFAULT_THROAT
    singing = DEFAULT_SINGING
    phonetic = DEFAULT_PHONETIC

func set_voice_elf():
    speed = 72
    pitch = 64
    mouth = 110
    throat = 160

func set_voice_alien():
    speed = 42
    pitch = 60
    mouth = 190
    throat = 190

func set_voice_stuffy():
    speed = 82
    pitch = 72
    throat = 105
    mouth = 110

func set_voice_old_lady():
    speed = 72
    pitch = 32
    mouth = 145
    throat = 145

func _enqueue_phrases(input: String):
    _queue.clear()
    _create_phrases(input)

func _create_phrases(input: String):
    if input.length() <= MAX_CHARS:
        _queue.append(input)
        return
    var phrase = input.substr(0, MAX_CHARS)
    var i = phrase.length() - 1
    while phrase[i] != ' ' and i >= 0:
        i -= 1
    if i <= 0:
        _queue.append(phrase)
        _create_phrases(input.substr(MAX_CHARS + 1))
    else:
        _queue.append(phrase.substr(0, i))
        _create_phrases(input.substr(i + 1))

func _configure_sam():
    _gdsam.set_speed(speed)
    _gdsam.set_pitch(pitch)
    _gdsam.set_mouth(mouth)
    _gdsam.set_throat(throat)
    _gdsam.set_singing(singing)
    _gdsam.set_phonetic(phonetic)

func _configure_player(player):
    player.bus = audio_bus
    player.volume_db = linear_to_db(volume)

func _process_queue():
    if _interrupt or _queue.size() == 0:
        emit_signal("finished_speaking")
        return
    var phrase = _queue.pop_front()
    var buffer = _gdsam.speak(phrase) as PoolByteArray
    if not buffer:
        return
    var callback = BufferCallback.new()
    emit_signal("loaded_buffer", buffer, callback)
    if _current_player.playing:
        _current_player.stop()
    if not callback.audio_stream_override:
        _sample.data = buffer
        _current_player.stream = _sample
    else:
        _current_player.stream = callback.audio_stream_override
    _current_player.play()
    await _current_player.finished
    #yield(_current_player, "finished")
    emit_signal("finished_phrase")
    _process_queue()

[deadpixelsociety]

Hi! I am interested in updating it if people are interested in using it for 4.0. I believe the way native plugins changed however, and I haven't investigated that just yet. I'll take a look this weekend and see what it will take to upgrade everything.

[mikeydbrad]

I don't know about anyone else, but I'd love to have this in 4.0! Thanks for being willing to take a look

[deadpixelsociety]

@mikeydbrad It's been updated to 4.0, new code is now on the main branch. I'll get a new entry to the asset library going but it'll be a day or two before that's updated.