Technologies

Le Sound is leading the future of professional sound design. We develop unique and innovative technologies that are few available in the world. Our solutions feature procedural audio synthesis, adaptive granular analysis/synthesis and specific custom DSP algorithms, which allow synthesizing a very large spectrum of sounds, from environmental sounds or engine sounds to imaginary sounds. In addition, we develop voice processing tools for localization professionals and music production tools for music creators. For technology licensing, please send your inquiry to licensing@lesound.io.

SOUND

Procedural audio synthesis

Our procedural audio synthesis technologies use minimal amount of samples to generate a great variety of high-quality sound events. More importantly, our synthesis engine comes with specially-designed controls for intuitive and instant sound manipulation. Currently available technologies include wind, rain, fire, electric sparks and space ambience synthesizers. Some more are coming. (learn more about procedural audio)

Procedural WindProcedural RainProcedural ElecProcedural Fireicon_space

Engine sound analysis/synthesis

Armed with unrivaled pitch analysis, our granular engine reproduces engine cycle in a pitch-synchronous manner. For car engines, our technologies allows simlulating gear changing for upshifting and downshifting, and engine load effects during acceleration/deceleration.

rpmengineCarengineCycleenginePlaneengineBoat

Adaptive granular analysis/synthesis

Different from the standard granular analysis which uses a fixed grain size, our granular analysis adapts automatically to the sound signal and decomposes it more meaningfully for resynthesis. For non-harmonic sounds like tire-rolling sounds or sound textures, it allows one to “scratch” almost any sound recordings like a DJ.

MUSIC

Drum loops transcription

The first technology available from our groove extraction project is automatic drum loop transcription. Given a drum loop, our technology transcribes it into five components: bass drum, snare drum, close hihat, open hihat/cymbal and tom-toms. In addition to provide transcription for creative groove re-composition, this technology is expected to open up many possibilities for high-level processing of drums sounds.

VOICE

Dynamic level adjustment

Dynamic level adjustment not only makes use of local energy variation but also takes into account the loudness model to enable a perceptually relevant normalization. Different from the usual automatic loudness leveling, our dynamic level adjustment provides more controls to the dynamics within a dialog recording. The batch process is expected to help localization and broadcast professionals to render the speech dynamics more efficiently and effectively.

Dialogue recording cleaning

Dialogue recordings usually contain unpleasing noise from watering mouth or unstable pitch due to less controlled voicing. These noise are often quite strong in energy such that neither pop filters nor noise gate can deal with. Our unique de-noising technology automatically detects the mouth noise with high precision and corrects the noise without affecting the speech quality. The batch process is expected to save localization and broadcast professionals hundreds of hours of tedious manual editing.