Add Lutra script #353

chrisib · 2024-03-09T01:49:40Z

Adds a new syncable LFO script, similar to e.g. Harmonic LFO, but without the harmonics. Each output is a free-running LFO outputting one of 5 wave shapes (sine, square, triangle, saw, ramp) with a related-but-different clock speed; cv2 is slightly faster than cv1, cv3 is slightly faster than cv2, etc..., resulting in interesting phase shifting & drifting.

Inspired by Otterly by Expert Sleepers.

… range to make the fastest speed slower, but the slowest speed _way_ slower.

…selection

…e detailed multi-threading docs with examples

…offsets. This makes the spread more predictable at variable speeds

chrisib · 2024-03-23T19:15:02Z

I've made a small tweak to how the script works. Specifically the spread control now smoothly adjusts the speed of each channel relative to cv1:

cv1 goes from x1 to x1 of cv1 (surprise! it's the master channel, and always goes at the same speed as itself!)
cv2 will go from x1 to x1.2 of cv1
cv3 will go from x1 to x1.25 of cv1
cv4 will go from x1 to x1.333... of cv1
cv5 will go from x1 to x1.5 of cv1
cv6 will go from x1 to x2 of cv1

So at maximum spread each channel from cv2-6 is a nice harmonic interval of the speed of cv1: 1:1, 6:5, 5:4, 4:3, 3:2, and 2:1.

This fix makes the spread more predicable when changing the speed of cv1 via k1 and/or ain.

mjaskula

Overall this looks pretty good to me. I like the addition of the multithreading documentation you put some good work into it.

I still want to actually run this script and play with it, and dig into the threading a little more, but here is an initial review.

mjaskula · 2024-03-22T19:07:55Z

software/contrib/lutra.md

+
+## Configuration
+
+The functionality of `ain` can be configured by creating/editing `/saved_state_Lutra.txt` on the module. This JSON file


[discussion] Saved state is not meant to be user editable. It is meant to save the state of a running script so that it can pick up where it left off next time it runs. This sounds more like configuration and should use the configuration feature.

I interpreted things differently; I see the global config files (experimental_config & europi_config) as module-wide settings, while the saved_state files are effectively config files with a per-script focus. Adding a script-specific setting to e.g. experimental_config feels like overkill, since that setting would never be used anywhere else.

Creating separate saved-state & config files for each script also feels counter-intuitive to me. The saved state is the current configuration of the script. This particular feature is just a sneaky way of adding an extra bit of functionality that otherwise wouldn't be supported.

If that's how you are interpreting these features then the documentation for them needs to be improved. The firmware level config files are certainly module wide settings, but individual scripts can also have configuration, see the turing-machine and diagnostic script for examples.

This particular feature is just a sneaky way of adding an extra bit of functionality that otherwise wouldn't be supported.

My assumptions about why this feature wouldn't otherwise be supported is that the limited number of hardware UI elements on the EuroPi make adding these features difficult with out making it very menu driven. This is explicitly one of the things that the configuration system was meant to address.

Please see the following issue and PRs for details of the motivations for these features. I will open a PR that pulls these details into the documentation.

#163
#220
#185

Fair enough. I'll move that one item out of the saved state & create a separate config file for it.

EDIT
Done

mjaskula · 2024-03-22T19:17:30Z

software/contrib/lutra.md

+
+## Re-syncing
+
+When `b1` or `din` receive a high voltage all CV outputs are temporarily halted.  Once the input drops low again


[required] This documentation is meant to be understood by non-developer users of the script. As such, I don't think that describing a button as either high or low voltage is clear. I think it would be more clear to describe it in terms of the button being pressed or not.

I'll reword this to make it less developery-sounding.

mjaskula · 2024-03-22T19:39:24Z

software/firmware/experimental/thread.py

+        self.b1_high = False
+        self.b2_high = False
+
+        # MS ticks when events were last recorded


[question] These timestamps don't seem to be read anywhere. Are the meant to be a part of the class's API? If so they should be documented.

Yes, they're part of the API. They aren't used here but they are used in other multi-threaded scripts I have on the go. I'll add some additional documentation.

Added an additional section in the class header to explain the additional fields.

mjaskula · 2024-03-25T19:50:18Z

software/contrib/lutra.md

+depending on previous settings and random noise) they may be phase-shifted from each other.
+
+Turning `k2`clockwise will gradually increase the spread of clock speeds. `cv1` will always stay locked to the base
+clock speed, with `cv2-6` becoming progressively faster.  At maximum spread, the outpts follow common harmonic intervals


[discussion] I haven't played with this script yet, but I think that this would be useful/fun: Have you considered having the spread be zero in the center, adding the ability to both slow down and speed up the related clocks?

An interesting thought. Probably wouldn't be too hard to change the knob behaviour to allow the spread to speed up or slow down cv2-6.

I played around with this a little bit. In the end, I don't think I actually like this change; with the current implementation it's very easy to "lock" the waves' relative phases by just turning the spread control all the way to the left. That feels like a useful feature that gets lost if you need to centre the knob to set the spread to zero; the slightest imprecision will cause the waves to start spreading over time.

I may try it out some more later, but for now I think I'm leaning towards keeping the behaviour as it is right now.

mjaskula · 2024-03-25T19:54:36Z

software/contrib/lutra.py

+        "ramp": 4,
+    }
+
+    WAVE_SHAPE_SINE = 0


[optional] You could use these constants as the values in the above dictionary to reduce the number of places where the code has to match up a shape with its value.

software/contrib/lutra.py

mjaskula · 2024-03-25T21:20:36Z

software/contrib/lutra.py

+        self.load()
+
+    def load(self):
+        """Load and apply the saved configuration


[required] The configuration file and the saved state file are two different things in the europi firmware. It's confusing to have the documentation here call it a configuration file and then read from the saved state file. Ditto for the save() method below.

Like I said above, the saved_state file (in my interpretation) is the same thing as a per-script configuration file.

…state for everything. Update the docs to reflect the changes

…nts when populating the wave name/type dict

mjaskula

I never ended up spending anymore time with this, but that’s not a good reason to hold it up.

Add Lutra script & docs

c1904e1

chrisib added the new script Addition of a new contrib script label Mar 9, 2024

chrisib added 10 commits March 9, 2024 14:56

Use _thread to separate the GUI from the backend math. Change the LFO…

6272359

… range to make the fastest speed slower, but the slowest speed _way_ slower.

Remove the ISRs; I forgot _thread could lock up if they're used

ca0031f

Fix a bug when loading the saved state

9b2f323

Add a note that the wave shape shows up on the display for 3s during …

9025668

…selection

Code cleanup

7f6d15c

Move the digital input helper into a new experimental module, add mor…

ce0cfac

…e detailed multi-threading docs with examples

Linting

3974cfe

Linting

9f43b3e

Add the expert-users warning to the multithreading document

3dad3e1

Make the maximum spread use nice intervals instead of just flat tick …

79dbdca

…offsets. This makes the spread more predictable at variable speeds

mjaskula reviewed Mar 25, 2024

View reviewed changes

chrisib added 2 commits March 26, 2024 09:55

Use the separate config file for the ain mode instead of using saved-…

560746a

…state for everything. Update the docs to reflect the changes

Add additional documentation for the digital input helper, use consta…

7c36974

…nts when populating the wave name/type dict

mjaskula approved these changes May 22, 2024

View reviewed changes

chrisib merged commit 71f178a into Allen-Synthesis:main May 23, 2024
3 checks passed

chrisib deleted the lutra branch May 23, 2024 01:24

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add Lutra script #353

Add Lutra script #353

chrisib commented Mar 9, 2024

chrisib commented Mar 23, 2024

mjaskula left a comment

mjaskula Mar 22, 2024

chrisib Mar 25, 2024

mjaskula Mar 26, 2024

chrisib Mar 26, 2024 •

edited

Loading

mjaskula Mar 22, 2024

chrisib Mar 25, 2024

chrisib Mar 26, 2024

mjaskula Mar 22, 2024

chrisib Mar 25, 2024

chrisib Mar 26, 2024

mjaskula Mar 25, 2024

chrisib Mar 25, 2024 •

edited

Loading

chrisib Mar 26, 2024

mjaskula Mar 25, 2024

chrisib Mar 26, 2024

mjaskula Mar 25, 2024

chrisib Mar 25, 2024

chrisib Mar 26, 2024

mjaskula left a comment


		## Configuration

		The functionality of `ain` can be configured by creating/editing `/saved_state_Lutra.txt` on the module. This JSON file


		## Re-syncing

		When `b1` or `din` receive a high voltage all CV outputs are temporarily halted. Once the input drops low again

Add Lutra script #353

Add Lutra script #353

Conversation

chrisib commented Mar 9, 2024

chrisib commented Mar 23, 2024

mjaskula left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chrisib Mar 26, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chrisib Mar 25, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mjaskula left a comment

Choose a reason for hiding this comment

chrisib Mar 26, 2024 •

edited

Loading

chrisib Mar 25, 2024 •

edited

Loading