MELT combines a three-input mixer, analog filter, cutoff modulation, and output VCAs inside one 14HP Eurorack module. Add an oscillator source, pitch control, a gate, and an envelope, and the remaining signal path can stay inside MELT. The result is a playable subtractive voice without separate mixer, filter, and final VCA modules.
CVilization in Quad VCO mode provides a compact four-oscillator source, but the patch does not depend on it. Conventional Eurorack oscillators work as long as they receive suitable pitch control and provide signals MELT can accept. The fourth oscillator can remain optional until the basic voice behaves properly.
Input 1 includes a VCA, while inputs 2 and 3 use straightforward attenuating level controls. With nothing connected to LVL1, an internally normalled 7.5-volt signal keeps input 1 open. Start with modest levels on every oscillator because their combined amplitude affects how hard the filter core is driven.
Tune the sources before adding heavy resonance or Core Temperature. A low oscillator can establish weight, a central oscillator can carry the note, and a slightly detuned or differently shaped source can add movement. Distinct roles make the mixer easier to balance than three equally loud copies.
This compact front end is what makes MELT’s integrated monophonic voice stage practical in a smaller case. The audio mixer removes one external utility, while the low-pass output VCA removes another. MELT still needs control sources, but it handles the complete audio path after the oscillators.
Set the low-pass slope and cutoff with the output VCA still open. The 12 dB position retains more material above the cutoff, while 24 dB removes it more firmly. Choose the slope for the part before programming the envelope, since the same cutoff movement produces a different spectral result at each setting.
This shared filter and amplitude contour is the quickest route to a playable voice. The output VCA closes between notes, while FM1 opens or closes the filter during each envelope cycle. Adjust the VCA response first, then use the FM1 knob and polarity switch to set the tonal sweep.
If the voice remains continuously audible, check whether the envelope cable actually reaches the low-pass VCA CV jack. Inserting that cable replaces MELT’s normalled voltage, so a low envelope should close the output. A weak control signal may not open the VCA fully, while MELT specifies 7.5 volts as its maximum VCA CV level.
The FM1 polarity switch determines the sweep direction. In its normal orientation, a positive envelope can raise cutoff before decay brings it down. Inverting FM1 creates the opposite movement without altering the amplitude contour sent to the output VCA.
The band-pass output can provide another version of the same voice. Its VCA has a separate CV input and also stays open through normalled voltage when unpatched. If both outputs should share note articulation, distribute the envelope to both output VCA inputs rather than leaving the band-pass side continuously open.
Begin with FM2 fully down and increase it slowly. Audio-rate cutoff modulation can dominate the original oscillator balance, especially with strong resonance, so treat the FM2 knob as the depth control. Retune the fourth oscillator after changing played notes if the modulation relationship loses its useful range.
FM2 works independently from the envelope on FM1. The envelope can preserve clear note articulation while the fourth oscillator adds continuous movement inside each note. Reducing FM2 during setup confirms whether roughness comes from filter modulation or from Core Temperature and input level.
The fourth oscillator can instead feed AUX. In that position it bypasses the filter and joins the voice at the output VCAs, providing an unfiltered layer rather than cutoff modulation. Use this route for a fundamental or contrasting waveform that should remain present while the main three-oscillator mix is filtered.
Output selection affects the final behavior. Take the low-pass output for the conventional subtractive voice, or monitor low-pass and band-pass on separate mixer channels after giving both VCAs appropriate control. Match their levels carefully because the two responses emphasize different frequency regions even though they originate from the same oscillator mix.
CVilization in Quad VCO mode provides a compact four-oscillator source, but the patch does not depend on it. Conventional Eurorack oscillators work as long as they receive suitable pitch control and provide signals MELT can accept. The fourth oscillator can remain optional until the basic voice behaves properly.
Three oscillators build the source before filtering
Route the first three oscillator outputs into MELT’s inputs 1, 2, and 3. All three are summed before reaching the filter core, creating one monophonic signal even when the oscillators use different waveforms or tuning intervals. Send the same pitch CV to the oscillators so they follow one note line.Input 1 includes a VCA, while inputs 2 and 3 use straightforward attenuating level controls. With nothing connected to LVL1, an internally normalled 7.5-volt signal keeps input 1 open. Start with modest levels on every oscillator because their combined amplitude affects how hard the filter core is driven.
Tune the sources before adding heavy resonance or Core Temperature. A low oscillator can establish weight, a central oscillator can carry the note, and a slightly detuned or differently shaped source can add movement. Distinct roles make the mixer easier to balance than three equally loud copies.
This compact front end is what makes MELT’s integrated monophonic voice stage practical in a smaller case. The audio mixer removes one external utility, while the low-pass output VCA removes another. MELT still needs control sources, but it handles the complete audio path after the oscillators.
Set the low-pass slope and cutoff with the output VCA still open. The 12 dB position retains more material above the cutoff, while 24 dB removes it more firmly. Choose the slope for the part before programming the envelope, since the same cutoff movement produces a different spectral result at each setting.
One envelope controls tone and final amplitude
Send the keyboard or sequencer gate into an envelope generator. Split the envelope output and patch one copy into MELT’s FM1 input and the other into the low-pass output VCA CV input. One contour now moves the cutoff and shapes final amplitude together.This shared filter and amplitude contour is the quickest route to a playable voice. The output VCA closes between notes, while FM1 opens or closes the filter during each envelope cycle. Adjust the VCA response first, then use the FM1 knob and polarity switch to set the tonal sweep.
If the voice remains continuously audible, check whether the envelope cable actually reaches the low-pass VCA CV jack. Inserting that cable replaces MELT’s normalled voltage, so a low envelope should close the output. A weak control signal may not open the VCA fully, while MELT specifies 7.5 volts as its maximum VCA CV level.
The FM1 polarity switch determines the sweep direction. In its normal orientation, a positive envelope can raise cutoff before decay brings it down. Inverting FM1 creates the opposite movement without altering the amplitude contour sent to the output VCA.
The band-pass output can provide another version of the same voice. Its VCA has a separate CV input and also stays open through normalled voltage when unpatched. If both outputs should share note articulation, distribute the envelope to both output VCA inputs rather than leaving the band-pass side continuously open.
The fourth oscillator adds controlled instability
Once the three-oscillator voice responds cleanly, patch a fourth oscillator into FM2. At lower tuning, u-he describes the effect as raspy, audio-rate LFO-style modulation. Higher tuning pushes the filter toward a more metallic FM character.Begin with FM2 fully down and increase it slowly. Audio-rate cutoff modulation can dominate the original oscillator balance, especially with strong resonance, so treat the FM2 knob as the depth control. Retune the fourth oscillator after changing played notes if the modulation relationship loses its useful range.
FM2 works independently from the envelope on FM1. The envelope can preserve clear note articulation while the fourth oscillator adds continuous movement inside each note. Reducing FM2 during setup confirms whether roughness comes from filter modulation or from Core Temperature and input level.
The fourth oscillator can instead feed AUX. In that position it bypasses the filter and joins the voice at the output VCAs, providing an unfiltered layer rather than cutoff modulation. Use this route for a fundamental or contrasting waveform that should remain present while the main three-oscillator mix is filtered.
Output selection affects the final behavior. Take the low-pass output for the conventional subtractive voice, or monitor low-pass and band-pass on separate mixer channels after giving both VCAs appropriate control. Match their levels carefully because the two responses emphasize different frequency regions even though they originate from the same oscillator mix.
