Undine Gloves: A Hybrid CV-MIDI Wireless Gestural Controller for Modular Synth Ecosystems

All rights remain with the authors under the Creative Commons Attribution 4.0 International License (CC BY 4.0). Proc. of the 17th Int. Symposium on Computer Music Multidisciplinary Research, London, United Kingdom, 2025 Undine Gloves: A Hybrid CV-MIDI Wireless Gestural Controller for Modular Synth Ecosystems Bruno Newman [0009-0004-4029-6683] University of Chichester Conservatoire, Chichester, UK [email protected] Abstract. Undine Gloves are wireless, dual-hand gestural controllers for electronic music that provide six degrees of motion sensing with hybrid CV+MIDI output for modular and DAW-based systems. Each glove streams three axes via ESP-NOW to a receiver that generates 12-bit CV (0-5 V) alongside USB-MIDI, achieving measured end-to-end latency of ≈11 ms. On-glove AMOLED displays show real-time bar-graphs of each axis, supporting calibration, drift compensation and audience comprehension. The CMMR demonstration features a labelled video showing gesture-to-parameter mappings, live performance with a minimal Eurorack voice (VCO, filter, reverb), explicit control mappings, and audience hands-on participation. While primarily effective for continuous timbral control and constrained-pitch interaction, the system roadmap includes quantisation modes, haptic feedback and integration with biosensors and environmental modules, positioning Undine Gloves as a Eurorack-native interface for embodied performance in modular/DAW-hybrid contexts. Keywords: gestural control; control voltage; modular synthesis; real-time performance; HCI. 1 Introduction Gestural control for electronic music often passes through MIDI-centric pathways that add latency and complicate integration with analog modular systems. Undine Gloves address this by coupling six-axis sensing to a hybrid output stage that drives Eurorack CV and USB-MIDI simultaneously. The project follows practice-as-research methods [1] (iterative prototyping, performance trials and reflection) to balance playability, responsiveness and system transparency. Within the CMMR theme “Sound, Music: Space, Place,” we focus on how gestural space maps to continuous timbral and constrained-pitch behaviours in a modular/DAW-hybrid rig and preview extensions incorporating bio-signals and environmental data as situated controls. Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 1041 Newman 2 System Overview The system architecture is shown in Fig. 1. Controllers: Two wearables (left/right) each capture three motion axes and stream values in the 0-4095 range. On-glove AMOLED coloured bar-graphs display each axis in real time. A dedicated Calibrate button re-centres the neutral pose (ideally with the glove stationary/flat). Units auto-calibrate on power-up, a sleep function conserves battery between interactions and a power button controls on/off. Link: ESP-NOW provides low-overhead 2.4 GHz unicast at 100 Hz per controller; measured wireless and processing stages add ≈2-3 ms, well within acceptable latency thresholds for musical performance [2]. Receiver: An ESP32-S3 drives an MCP4728 quad 12-bit DAC for 0-5 V CV and transmits USB-MIDI CCs in parallel. Smoothing and scaling run on-device, with mappings configured at the receiver. Fig. 1. System architecture of the Undine Gloves controller and receiver. 3 Interaction and Demonstration Review assets: A 60-s labelled video illustrates each mapping with on-glove AMOLED bar-graphs and concludes with a latency figure (end-to-end ≈ 11 ± 2 ms). Live setup: The demo uses a compact Eurorack skiff with a single VCO routed into a low-pass VCF, augmented by a Strymon Big Sky MX reverb and a volume pedal. Two gloves (ESP32-S3 + AMOLED) transmit wirelessly to the receiver, which provides CV and USB-MIDI simultaneously. LEFT CONTROLLER (ESP32 + Sensors) X, Y, Z axes RIGHT CONTROLLER (ESP32 + Sensors) X, Y, Z axes RECEIVER MODULE (ESP32-S3 + MCP4728 DAC) Processing & Output 4x CV OUTPUTS 0-5V Analog Quad 12-bit DAC USB MIDI CC Messages Digital MODULAR SYNTHESIZER COMPUTER DAW MIDI HARDWARE Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 1042 Undine Gloves Mappings: Left-hand arm (X) → Pitch (quantised pentatonic scale). Left-hand twist (Y) → Reverb mix (MIDI CC 15). Right-hand arm (X) → Filter cutoff Right-hand twist (Y) → Resonance Emergent voices: This configuration produces two independent sonic voices from a single oscillator: the left hand articulates a constrained irrational scale, while the right hand draws out a rational harmonic overtone voice through filter self-oscillation. Combined with reverb depth control, the performer can generate textures that sound larger and more complex than the minimal synthesis chain might suggest. Audience hands-on: Attendees will be invited to try the gloves, observe the AMOLED feedback, re-zero the neutral pose and explore the mappings under supervision. 4 Contribution and Context 4.1 Positioning vs Prior Art Previous gestural controllers highlight both the potential and challenges of embodied interaction. Leap Motion offers rich optical hand-tracking but is constrained by line-ofsight occlusion, tethered computer use and no native CV output. Mi.Mu Gloves provide robust, stage-tested performance with machine-learning-based gesture recognition, but their MIDI/OSC focus and higher cost make them less accessible for modular integration. Hot Hand and similar IMU rings deliver compact motion-to-MIDI control, but with limited expressive channels and no CV outputs. Building on precedents like Sonami's Lady's Glove, Undine Gloves distinguish themselves as a Eurorack-native system with hybrid CV+MIDI outputs and on-glove AMOLED visualisation. This combination enables continuous, low-latency modulation of modular parameters while simultaneously addressing DAW effects, without external MIDI-to-CV conversion layers. 4.2 Limitations and Mitigation Like other gestural controllers, the system faces limitations in precision, drift and lack of tactile cues. Dead-band filtering, temporal smoothing and hysteresis reduce jitter from minor hand tremor, while the re-zero calibration function compensates for drift during performance. The AMOLED bar-graphs provide immediate visual confirmation of the neutral position. For pitch control, constrained ranges and optional quantisation to discrete scale degrees improve stability. Planned haptic feedback will provide vibration cues to mark discrete parameter states, offsetting the absence of physical boundaries. These measures maintain a responsive ≈11 ms latency [2] suitable for continuous timbral shaping and constrained-pitch interaction. Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 1043 Newman 5 Conclusion and Outlook Undine Gloves demonstrate a practical approach to embodied control in modular/DAW-hybrid contexts. The CMMR demonstration combines a structured video and live hands-on setup, showing how a minimal synthesis chain (single VCO → low-pass VCF) yields two distinct voices plus reverb depth under embodied control. The system is optimised as a low-latency, Eurorack-native interface rather than a universal gestural controller. Future development will extend functionality through haptic feedback to complement AMOLED cues, biosensor integration via the Vitae module, environmental sensing through the Kintsugi module and coupling with the Naiad low-pass gate for further timbral exploration. These advances will expand Undine as a situated sonic ecosystem while retaining the robustness needed for live use. A demonstration video is available online. Acknowledgements. The author thanks Catherine Crisp, Deirdre Gribbin and Hugo Frey for supervisory feedback, and colleagues at the University of Chichester Conservatoire for support during prototyping and evaluation. This work was supported in part by the University of Chichester Research and Innovation Fund (RIF). Disclosure of Interests. The author declares no competing interests. Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 1044 Undine Gloves References 1. Smith, H., Dean, R.T.: Practice-led Research, Research-led Practice in the Creative Arts. Edinburgh University Press, Edinburgh (2009) 2. Wessel, D., Wright, M.: Problems and prospects for intimate musical control of computers. Computer Music Journal 26(3), 11–22 (2002) 3. Magnusson, T.: Musical organics: A heterarchical approach to digital organology. Journal of Music Technology & Education 10(1), 7–33 (2017) Proc. of the 17th International Symposium on CMMR, London, UK, Nov. 3-7, 2025 1045