Axe, a saw/square oscillator modeled to emulate old analog hardware. As it consume more CPU than Saw and Square I will (for now) keep those, although Axe sounds better.
Vxy: Awoken from the dead. This module is again available in an improved version.
Fluencerator: Added chromatic scale. Added 2 context menu options; Keep gate open when following note is gliding and force all notes before a glide to have a duration of 1 clock cycle. Fixed mixolydian scale.
Polyphonic support added: Deadband, DC, Conv, Imp, Mixer6 (inputs and FX return only), Vibrato, Nap, Fluencerator, Square, Saw, Oxcart, Jette, Digi, Vxy (4 channels) and Fil.
Bass: ENVMOD knob now works relative, added context menu option for old behavior. If accent input arrive slightly late, it now restart the note as accented, but must still arrive during attack phase or the note wont be accented.
Mixer6: Added context menu for reducing main out gain depending on number of active lines. Better EQ filtering.
Imp: Better EQ.
Fil: New simpler algorithm.
Nap: Algorithm improvement.
DC: Much more simple and sturdy.
Vibrato: Smooth width knob input.
Rebel: Fixed trigger was broken.
Deadband: Width of zero now mean bypass.
Zod: Fix static noise from noise-gate. Noise-gate is now highest priority in static curve.
Zod and Non: Reduced to 15 ms lookahead and decoupled lookahead buffer from RMS average time. Added 4x Oversampling. Replaced non-linear output saturation with hard clamp. Fixed limiter overshoot, can now instant attack. Stereo input is no longer summed, its individually rectified and then max magnitude is used.
LED knobs on most knobs that can be modified with CV. See manual on how to interpret it.
Lots of CPU usage optimizations in most modules.
Many minor improvements and fixes in all modules.
I took the liberty to make changes to the following modules, so they will in some degree change/break old patches: Mixer6 and Imp (changed EQ drastically), Zod and Non (changed knob limits and dB detection), Fil and Nap (sound different), Vxy (spends less time at the 5V limits).
NON as a limiter is part of my startup template. I think it’s great, so it’s present in many of my patches up to now..
Hence, what are the consequences of “changed knob limits and dB detection” and what would you recommend to compensate for the recent changes when loading older patches?
If you use mono input to Non then you don’t have to change threshold.
If you use stereo, and the stereo often are in phase, you might have to lower the threshold 3-6 dB to get same effect.
If your attack was higher than 2 ms, it will be reduced to 2 ms when you load the patch. It used to be able to be set up to 10 ms. Since its how fast it will detect and react to peaks being louder than your threshold, it will react faster if you had it above 2 ms, and let less through, so limit more.
I thought about it. For example showing the polyphonic signals overlaid in background faded out a bit, or having a selector knob for which of the 16 signals to use (but I am running out of space to put knobs).
If you have a good idea on how to handle poly signals, I might make it.
I think I can do it like that, I made a mockup for how it ideally could look. Then for example when clicking on display, you can choose which poly ch to use as trigger (and that will be the main signal drawn without transparency).
Cool, that looks great. Thanks for considering. VCV Scope does a good job of showing poly signals, so it might a good model to follow, but it would be nice to have a bigger display and information.
One thing that made me think is that would be neat if the module width was extendable. Some modules can be made wider, so for a scope, this would be useful to stretch wider past the base HP.
I am really digging the feature set of scope. It is rapidly becoming my default scope when I want to monitor more than 2 signals.
One minor quibble - it would be nice if the buttons with more than 2 options had a right click context menu option to directly set the value from a list. Just configure it as a switch - it is pretty simple. To be consistent you might want the feature for all the buttons, even the 2 state ones.
It has a known issue that is not in the library yet: (ETA for fix: next week)
At high time/div setting with sample-rates higher than 44.1k, the rendering can be a bit quirky when there is not enough buffer memory to fill entire display.
Changelog for v2.6.31
Aliasmeter: Measure alias/noise to signal ratio.
Excavi: Added shape CV inputs and allowed them to drive the shape negative. Added PWM square shape. Support for soft sync.
Scope: Much more reliable freq readout. Polyphony support. Menu in display when clicking on it. Buttons with more than 2 states, now work like switches. In order to not use too much RAM for polyphony, each channel now uses x4 less memory, this can affect when using high time/div setting that it might not have enough memory to fill the display, especially at higher sample-rates.
Coil: Added spacing between the coils, so it works more like a spring and less like delay. Context menu item for legacy behavior.
Fauna: Like Flora but with more modern math to get cleaner sound.
I’m still loving the scope. But I am wishing for a change to the vertical scaling.
I would like to have a scale factor that lets me nearly fill the vertical space with a 10 volt peak to peak signal, and another scale factor that works with 20 volt peak to peak.
One option is to change the vertical grid to having 12 steps with a scale factor of 1 V / div for 10V peak to peak, and 2 V / div for 20 V peak to peak.
Another option would be to let the scale be freely set, without quantization to pre-determined values.
The reason the scale is not free is to easy being able to set two channels to same v/div.
As for changing the vertical grid, its an option, I have to consider X-Y grid too though, as its nice having almost rectangular grids. And I have to consider backwards compatibility for existing patches.
I will think it over. Perhaps a context menu to change grid..
