Feature Requests

I received interesting feedback from a Mela user who created a complex and very interesting MIDI preset, and it got me thinking about some of the required changes that can address some shortcomings that have also been bothering me. At the moment, the On parameter is very utilitarian, but it can be optimised to be more performance-oriented. Below are some proposed changes. Add a visual indicator about an On parameter's current state when modulated Currently, an On parameter might look like it’s in an off state, but a modulation might have the state flipped. There’s no way to know this, and the UI will look disabled in that scenario. Solution : add a visual indicator showing the modulated state and don’t dim/disable the module UI. Turning off a module should no longer completely bypass its engine At the moment, when a lane or a module is turned off, the UI is dimmed, the parameters can no longer be changed via UI, and modulations no longer update. Solution : No longer bypass the engine (specific details below) and don’t dim/disable the module UI. Small downside is the module will continue to use some CPU because even when bypassed, it still updates UI and runs some DSP. But that's ok because most likely the user has the module around, so they can unbypass it as needed. MIDI Processors should no longer auto-release notes Currently, if an Arpeggiator or Euclid module receives a C3 note-on, and then the user bypasses the module before receiving the corresponding note-off message, the module will automatically send out a C3 note-off message to prevent stuck notes. Now, the issue with this is, the user might still be holding down the C3 key. Solution: On bypass, Arpeggiator and Euclid simply finish playing their last note(s) and don’t start another note, and they will send out note-on messages for the currently held notes. i.e. in the example above, on bypass C3, note-on will be sent out. That way, when the user finally releases C3, that’s when C3 will stop playing. MIDI Processors update output on parameter change without changing input Currently, some MIDI Processor modules don’t change output if parameters are updated and the input hasn’t changed. For example, if the Chord module settings are adjusted, the output won’t change for the currently held notes. The change will happen only after the input is changed. Solution Part 1: The Chord module should update its output as soon as the note On parameters change. This will allow chord building in real time. Solution Part 2 (unlikely a good idea): The Chord module should update its output as soon as the Pitch value changes. I’m less sure about this. (Vel parameter changes causing the output to the Chord module output to change don’t seem useful at all. Retriggering notes in this case seems like a bad idea.) Continuing on the Solution Part 2, if we go ahead with that, maybe it means things like Note Range and Transposer should also update their outputs as the range or transposition amount changes. Again, I'm not certain this is a good idea. Audio/Pitch Processors crossfade their outputs when the On parameter is toggled Currently, when turning an Audio Processor on or off, there’s an audible click, e.g. VA Filter, this is because audio switches instantly. It’s not a big deal if it’s just trying things out and isn’t automating or modulating the On parameter as part of a song or performance. But if you do want to automate and modulate or even tap on the On toggle during performance would be great if it didn’t click. Also, there’s another issue related to the DSP being turned off for oscillator and LFO modules. The phasor no longer updates. This means if you turn the oscillator off, next time you turn it on, its relative phase would have changed against other modules. The resync will happen on the next note-on event. Solution: Turn the On parameter from a simple on/off 2-state list into a continuous one with [0%, 100%] range and make it crossfade like a wet/dry mix. This means the parameter can be modulated and automated, and will change its value between on and off continuously. It will also be clickless when tapping on it, as there will be a quick 20ms crossfade. For simplicity, it will continue to be a Mono target, meaning it won’t support Poly modulation.

The Phase Distortion module uses the classic phase distortion technique to reshape any waveform. It splits the waveform cycle into two parts, stretching the first while compressing the second. As it is not an oscillator, it relies on the Pitch signal and note-on events (Note Info in Poly mode and MIDI in Mono mode) to determine the waveform's start point and cycle length. This enables many creative possibilities. The Pitch signal can be transposed down an octave to distort a waveform an octave higher. An unrelated Pitch signal can be used to apply phase distortion to a drum loop for unique rhythmic textures. The Phase Graph visualises the waveform's phase after distortion. Dragging it horizontally or vertically adjusts the Skew or Tilt parameter correspondingly. The Curve parameter smooths the phase rate change at the cycle split point, with higher values producing a mellower sound and reducing aliasing. Demos https://x.com/Nikolozi/status/1955128398939885998 https://x.com/Nikolozi/status/1955925587068964883

The Latency module asks the host to deliver audio and MIDI events ahead of time, allowing for processing in advance. For example, it can enable an envelope follower to detect transients before they are heard. It's a global setting, and the module can be placed anywhere in a preset. The Latency parameter sets how far in advance events are delivered. Most iOS hosts do not yet support plug-in latency compensation. This module can help users check DAW support and assist developers with testing and debugging their hosts. Latency is a Mela Lab module and part of Collection 0, meaning it is unlocked for all users.

The Note Hold module blocks note-off MIDI events and converts every second note-on event of a given note into a note-off event. The Held view displays up to 9 held notes. If there are more, an ellipsis symbol appears. Tapping the Held view releases all held notes. Note Hold is part of Collection 4.

Most modulation targets that are on/off in behavior will turn on when the modulating source goes above 50% and off when source goes back below 50%. I’d like a way for such targets to behave as toggles instead, so that the target will go on when source is >50%, but won’t go off again until the source has gone back under 50% and then hits >50% again. This would allow for things like routing signals differently on odd vs. even note-on events. This could be accomplished by just adding a new mod target to the Macro Button module, similar to the existing “Macro” target, but maybe name it “Macro Toggle”, then any other mod target that I want to be toggle-able could just be chained from a Macro Button.

Instead of having separate modules, the Split and Merge features have been integrated into the Group module. I think this approach makes more sense. The update Group module is available as of build 439 . Feedback on this approach is welcome. Details below. Group Module The Group module is now even more versatile. The module lets you group a sequence of modules using its sublane. The signal flow between the containing lane and the sublane depends on one of four modes: Insert, Isolate, Split, or Merge. The module actions menu icon displays the current mode instead of the ellipsis used by other modules. Tapping it opens a menu to change the mode and access standard module actions. Insert Mode The Group module passes its input signals to the sublane and forwards the sublane's output to its own output. Ideal for organising modules. Isolate Mode The Group module doesn't provide any input to the sublane, and ignores its output. The module input and output remain identical. Ideal for creating an independent sublane that appears inline with the containing lane. For example, it can be used to generate an Envelope Follower modulator from a side-chain input. Split Mode The Group module forwards a copy of the input signals to its sublane, and ignores its output. The module input and output remain identical. Ideal for pre-processing signals before sending them to Audio Out, MIDI Out, or Envelope Follower modules. Merge Mode The Group module doesn't provide any input to the sublane, but merges the sublane's output signals with the module's input. The module's MIDI output is a merge of its input and the sublane's output MIDI signals. The Audio output is the sum of the module's input and the sublane's output signals. The output signal is Poly only if both the module's input and the sublane's output are Poly. Ideal for pre-processing signals coming from Audio In, Feedback, or MIDI In modules before merging. Abandoned Split & Merge Modules Split Module The module forwards a copy of the input signals to its sublane. Split's input and output signals are identical. The module is ideal for pre-processing signals before sending them to Audio Out, MIDI Out, or Envelope Follower modules. Split is a Mela Lab module and part of Collection 4. Merge Module The module merges its sublane output signals into the containing lane. The MIDI output is a merge of its input and the sublane's output MIDI signals. The Audio output is the sum of its input and the sublane's output Audio signals. The output signal is Poly only if both the module's input and the sublane's output are Poly. The module is ideal for pre-processing signals coming from Audio In, Feedback, or MIDI In modules before merging. Merge is a Mela Lab module and part of Collection 4.

Currently, lanes are automatically named as Lane 1, 2 etc. Would be great to be able to rename them like modules.

Introducing Stack, the first module that processes its submodules in parallel rather than in series, like sublanes. The Stack module feeds a copy of its input signals to all its submodules. The module's output signals depend on the Mode parameter, which has three states: Sum Audio, Merge MIDI, and Passthru. Sum Audio: Stack's output Audio, or the Audio components of the Poly signal, is the sum of the submodule output Audio and/or the Audio components of Poly signals. Sum Audio: If Stack's input and all submodule outputs are Poly signals, the module's output will also be Poly; otherwise, it will be Audio. Sum Audio: The MIDI signal and non-Audio Poly signal components (unless collapsed) remain the same as Stack's input. Merge MIDI: Stack's output MIDI signal is a merge of the submodule output MIDI signals. Merge MIDI: The Audio and Pitch signals, or the Poly signal, remain the same as Stack's input. Passthru: The submodule outputs are ignored, and Stack's output signals are identical to its inputs. You may want to process multiple sublanes in parallel for more complex setups. To do this, simply stack Group modules. With the Stack module, you can create simple MIDI effects, such as converting notes into chords by stacking Transposer modules, or build complex sequencers by combining Group, Euclid, and others. You can also easily create layered synths or effects. Stack lets you view submodules individually or as an expanded horizontal stack. When viewed as a horizontal stack, unlike sublanes, vertical lines replace signal indicator arrows, indicating that the submodules are processed in parallel. Stack is a Mela Lab module and part of Collection 4. Mela 6.4 Build 416: Moved Stack from Mela Lab to the Signal Router category. Mela 6.4 Build 414: You can now think of a Mela preset as a root Stack module with Group submodules. Because that's exactly what it is. Mela preset is now simply stacked, chained, and nested modules. Mela 6.4 Build 413: The MIDI Stack module has been generalised and is now simply called Stack. Instead of having separate MIDI Stack and Audio Stack modules, there is now a single Stack module with different modes. Mela 6.4 Build 411: OK the wait is over, the first stack module is here and ready for testing: Introducing the first stack module: unlike sublanes, where modules are processed in series, a stack module runs its submodules in parallel. The MIDI Stack module feeds a copy of its input signals to all its submodules and merges their output MIDI signals at the output. You may want to process multiple sublanes in parallel for more complex setups. To do this, simply stack Group modules. With the MIDI Stack module, you can create simple MIDI effects, such as converting notes into chords by stacking Transposer modules, or build complex sequencers by combining Group, Euclid, and others. MIDI Stack lets you view submodules individually or as an expanded horizontal stack. When viewing them in a horizontal stack, unlike sublanes, you will notice vertical lines instead of signal indicator arrows, indicating that the submodules are processed in parallel. MIDI Stack also feeds copies of the Poly, Audio, and Pitch signals to the submodules for processing, but their non-MIDI signal outputs are ignored. This means the Poly, Audio, and Pitch signals pass through unaltered. MIDI Stack is a Mela Lab module and part of Collection 4.

Mela 6.3 Beta: The module transforms a straight beat into a dynamic rhythm with accents and shuffle. Sync Parameter: Sets the timing grid for accents and shuffle. Accent Parameter: Sets the amount of accent applied to notes at odd grid positions. Shuffle Parameter: Sets the delay amount for notes at even grid positions, creating a swing or shuffle feel. Groove is a Mela Lab module and part of Collection 4.

Mela 6.3 Beta The module shortens a playing note when a new note with the same MIDI channel and pitch starts, ensuring the earlier note ends before the new one begins. It’s useful in situations where you want to stop earlier note-off events from interrupting a synth’s playback of the most recent note-on event. Log: The view displays the last 8 shortened notes. Overlap Remover is part of Collection 0.