wntrblm / sol Goto Github PK

I recently updated Sol to CircuitPython 7.30 and after doing so, the module stopped working (blinking green light) and after installing Mu, and opening a serial console, importing code.py, it turned out that the .mpy format changed from CircuitPython versions 6 to 7

I used the circup python module to check which ones needed to be updated and two modules: adafruit_bus_device and neopixel needed to be updated, and I went ahead.

After restarting, there were still green flashing lights and I opened code.py on Mu and tried to run it and it gave the following error:

Auto-reload is on. Simply save files over USB to run them or enter REPL to disable.
code.py output:
Traceback (most recent call last):
  File "code.py", line 84, in <module>
  File "/lib/winterbloom_sol/__init__.py", line 40, in run
  File "/lib/winterbloom_sol/sol.py", line 292, in __init__
  File "/lib/winterbloom_sol/sol.py", line 187, in __init__
  File "/lib/winterbloom_ad_dacs/_common.py", line 126, in soft_reset
  File "/lib/winterbloom_ad_dacs/_common.py", line 122, in send_command
AttributeError: 'SPIDevice' object has no attribute 'spi'

I checked the call in _common.py line 115:

spi_device = SPIDevice(spi, cs_io, polarity=0, phase=1, baudrate=5000000)

There doesn't seem to be a problem as the interface hasn't changed much:

# new version of spi_device.py
# https://github.com/adafruit/Adafruit_CircuitPython_BusDevice/blob/main/adafruit_bus_device/spi_device.py
def __init__(self, spi: SPI, chip_select: Optional[DigitalInOut] = None, *,
        cs_active_value: bool = False,
        baudrate: int = 100000,
        polarity: int = 0,
        phase: int = 0,
        extra_clocks: int = 0
    ) -> None:
        self.spi = spi
        self.baudrate = baudrate
        self.polarity = polarity
        self.phase = phase
        self.extra_clocks = extra_clocks
        self.chip_select = chip_select
        self.cs_active_value = cs_active_value
        if self.chip_select:
            self.chip_select.switch_to_output(value=True)

# old version used spi_device.py
# https://github.com/adafruit/Adafruit_CircuitPython_BusDevice/blob/cf7eabc4fa079a1443f548aed8ee2966af9b5fc1/adafruit_bus_device/spi_device.py
def __init__(self, spi, chip_select=None, *,
                 baudrate=100000, polarity=0, phase=0, extra_clocks=0):
        self.spi = spi
        self.baudrate = baudrate
        self.polarity = polarity
        self.phase = phase
        self.extra_clocks = extra_clocks
        self.chip_select = chip_select
        if self.chip_select:
            self.chip_select.switch_to_output(value=True)

I'm a bit stuck how to proceed. Can you please advice on what could be the fix, or if it's too much of pain (considering this module is discontinued) can you let me know how to downgrade to make it work again.

Thanks!

✨

In process of working with the ADSR I noticed that the attack phase starting before the release phase was finished would lead to audio clicks when patching CV C -> CV of a VCA with a sound source going into that VCA. If I patched CV C through a slew limiter to delay the attack/release slightly the clicks stopped.

Some screenshots showing the difference between the waveforms:

without a slew limiter adsr set to c_adsr = sol.ADSR(attack=0.2, decay=0.2, sustain=0.7, release=1.0)

with a slew limiter:

Here is the midi clip I used for this example (90 bpm)

This is a snippet of the code I was using (minus some other stuff) that can recreate this issue. Please let me know if you have any problems reproducing this:

import winterbloom_sol as sol

cv_c = {
    'attack': 0,
    'knobs': [
        (60, 'attack'),
    ], 
}

c_adsr = sol.ADSR(
    attack=0.5,   # Seconds
    decay=0.2,    # Seconds
    sustain=0.7,  # Percentage - 0.0 to 1.0
    release=1.0   # Seconds
)

def read_ccs(last, state):
    global cv_c
    for cc, val in cv_c['knobs']:
        if last.cc(cc) != state.cc(cc):
            cv_c['attack']= state.cc(cc) * 5
    global c_adsr
    c_adsr.attack = cv_c['attack']


def loop(last, state, outputs):
    """The loop is run over and over to process MIDI information
    and translate it to outputs.

    "last" holds the previous state, "state" holds the current state,
    and "outputs" lets you control the output jacks.

    You can read more about the state here: TODO.
    And more about the outputs here: TODO.
    """
    read_ccs(last, state)
    # Whenever a new note message comes in, such as from
    # a key being pressed or a sequencer sending a note,
    # update the gate and trigger outputs.
    if sol.was_key_pressed(state):
        c_adsr.start()

    if not state.note:
        c_adsr.stop()

    outputs.cv_c = c_adsr.output * 8.0

sol.run(loop)

Since state.clock can remain the same value between subsequent calls to loop, if the clock is slow enough loop will actually retrigger the output multiple times for a single clock.

A possible fix is to make it possible to "latch" Trigger/Retrigger objects - preventing them from triggering again until unlatched.

Another possible fix is to make helpers.should_trigger_clock memoize the last state.clock value that it returned True for and return False if it sees it again.