4TUNE

/ pitch tracking & karaoke scoring

4TUNE
Technology Details

4TUNE automatically detects the pitch of a singer (low latency real-time pitch tracking) and rates the pitch deviation from a given score.

The 4TUNE SDK is designed specifically for karaoke apps especially on mobile devices. It features a low latency pitch tracking using the same technology as used in the ELASTIQUE TUNE combined with an easy to use interface. 

ATAAK!

/ Audio-to-audio alignment kit

ATAAK!
Technology Details

AtAAK! is a technology for the automatic alignment of two audio files of similar content. Examples for applications are:

• adjustment of timing between different tracks
• adjustment of tempo for the editing of takes in classical music
• automatic synchronization of overdubbed with original speech

technology

The AtAAK! process is split into two phases, the Analysis stage and the Processing stage. During the analysis, the similarity over time of different points in time between both audio signals is being computed. In contrast to other algorithms, the user may chose whether the files might be similar with respect to to the four categories Pitch, Timbre, Loudness and Onset times.

AUFTAKT

/ tempo & beat tracking at its best

AUFTAKT
Technology Details

AUFTAKT is an automatic BPM (beats per minute) detection recognizing the tempo of any audio input signal. Furthermore, the location of the beats is detected (beat grid). Tempo and beat location information are useful for playlist generation, beat matched mixing, for automatic generation of high-level meta data information (e.g. ID-Tags) and several other areas of application..

Other than previous implementations AUFTAKT v4 relies on deep neural networks and state of the art statistical models to take beat tracking to a new level. The network has been trained with various music ranging from western pop music to classical and even ethnical music as well as single loops and guitars in order to broaden the range of trackable music stlyes. AUFTAKT v4 has improved detection rates by more than 30% across all genres, and even more than 40% on complex jazz music compared to its previous versions.

AUFTAKT is able to track tempo variations and changes as well as time signature changes.

Features of AUFTAKT V4:

• Highly improved tempo detection and beat tracking ability
• Improved downbeat detection
• Ability to track abrupt tempo changes
• Ability to track time signature and time signature changes.
• Possibility to set multiple correction points for improved tracking.

BARBEATQ

BARBEATQ
Technology Details

Colored wave form

/ easy made waveform displays

CWAVE
Technology Details

The COLOREDWAVEFORM SDK eases the generation of waveform displays. The SDK enables real-time display on screen as well as image generation, e.g. for use in a web environment. The SDK currently provides two types of waveforms: a standard monochrome waveform and a multicolored waveform that colors the waveform along the horizontal axis based onthe momentary frequency content. Various envelope types provide control over the way audio samples are summarized when displaying the waveform.

DRUMS

/ High-Precision Drum-to-Midi Transcription

DRUMS
Technology Details

Integrate High-Precision Drum-to-MIDI Transcription with DRUMS

DRUMS is our dedicated audio-to-MIDI solution for percussive content, designed to accurately transcribe drum performances from audio into MIDI. Powered by neural networks and running fully offline, DRUMS delivers precise onset and velocity detection across 9 core drum classes: kick, snare, closed hi-hat, open hi-hat, low/mid/high toms, crash, and ride cymbal.

Optimized for isolated drum recordings, DRUMS excels in capturing the rhythmic and dynamic nuances essential for professional MIDI workflows. Whether for sample replacement, beat editing, or creative rearrangement, it produces clean, production-ready MIDI output. Performance has been benchmarked against state-of-the-art research models using mir_eval and proprietary evaluation metrics on real-world test sets. The result: reliable, high-fidelity transcription that meets the demands of modern audio applications. Beyond transcription, DRUMS opens up possibilities for groove extraction, beat tracking, rhythm analysis, and integration into drum education, virtual drummers, or DAW extensions.

ELASTIQUE

/ Industry-leading time stretching & pitch shifting

ELASTIQUE
Technology Details

zplane's ELASTIQUE family offers three different time-stretching (aka time-scaling, master tempo or key lock) and pitch shifting (aka pitch scaling or key lock) engines, optimized for different use cases: ELASTIQUE PRO, ELASTIQUE EFFICIENT, and ELASTIQUE TUNE. All three of them are able to run in realtime.

ELASTIQUE PRO

ELASTIQUE PRO is a general purpose time-stretching engine that offers unmatched quality and easily fulfills the high quality demands of professional productions and broadcast applications. The time-stretching engine is based on state-of-the-art psychoacoustic models and signal processing theory. ELASTIQUE PRO is based on a completely new approach to time-stretching, making stretching artifacts obsolete and thus providing sharp transients and crystal clear vocals without phasing artifacts. ELASTIQUE PRO offers stable timing, inter-channel phase coherence and sample accurate stretching. As a special feature, ELASTIQUE PRO allows you to do formant preserving pitch shifting not only for monophonic, but also for polyphonic audio to avoid the well-known mickey-mouse effect when pitching up or down. Last but not least, all of this can be done in real-time on current CPUs.
Pro is the pitch shifting engine used in our ELASTIQUE PITCH plugin.

Additionally, the ELASTIQUE PRO SDK package makes ELASTIQUE EFFICIENT and 2 additional monophonic modes accessible through the same API to make integration of different processing modes as easy as possible.

ESPACE

/ Tweak your convolution reverb

ESPACE
Technology Details

Convolution-based reverbs are more and more replacing algorithms that generate synthetic reverberation tails. In daily use however, a frequent disadvantage of convolution reverbs is the lack of parameters to fine-tune the properties of the reverb you are working with, i.e. the adjustment of reverberation time, the usage of different settings for the early reflections and late reverberation etc.

The espace SDK provides the technology to overcome those disadvantages by offering both, the sonic quality of an IR-based reverb together with the possibility to actually parametrize the properties of the impulse response, resulting in a parametric convolution reverb.

FXPACK

/ EQs, delay effects & dynamics processing

FXPACK
Technology Details

The fx::pack SDK is a bundle of filters and effects that are required as standard effects in high quality audio software products. The implementation provides a range of ready to use effects with low workload.

• filters: lowpass (1st/2nd order), highpass (1st/2nd order), shelving (1st/2nd order), bandpass (2nd order), notch (2nd order), peak (2nd order)
• special filters: resonance lowpass (four pole), with LFO modulation
• delay, with LFO modulation
• chorus, flanger, phaser
• bit crusher
• dynamics processing: compressor, expander, limiter, gate (one-band, side-chain)

INTONE

/ Low Latency, natural-sounding pitch correction & modification

INTONE
Technology Details

INTONE is a formant-preserving, low-latency pitch correction and shifting algorithm designed to work in real time with monophonic vocal and instrumental recordings. INTONE is controlled by multiple parameters that enable a great palette of sonic results, ranging from natural-sounding correction that preserves expressive characteristics such as vibrato, all the way to the hyper-corrected vocals known and loved by fans of pop, hip hop, and trap. It is also possible to use INTONE as a pitch shifter or a harmonizer.

KORT

KORT
Technology Details

Kort analyzes the harmonic content of music of music recordings and returns the chords with their corresponding time stamps. Choose between Basic Mode that recognizes only major and minor chords, or Extended Mode that features a larger set of chord types containing between three and four pitches. The chords can be aligned with beat times from different sources, e.g. a fixed beat grid. Or you can use zplane's AUFTAKT beat tracking SDK that seamlessly integrates with KORT. KORT is an offline process which means it requires the complete audio file as input and also returns the result once the file has been completely analyzed.

New in version 2

• Improved chord analysis using Deep Neural Networks
• Detection has improved more than 20%, especially on more complex chords (extended set)
• Improvements have been made possible without sacrificing performance

PITCHES

/ Advanced Audio-to-MIDI Transcription

PITCHES
Technology Details

Enhance your product with state-of-the-art audio-to-MIDI technology. Powered by advanced neural networks, PITCHES transcribes pitch and velocity with exceptional accuracy in an offline process. It captures both single instruments and full audio mixes as a single pitch stream (no instrument separation).

PITCHES delivers performance on par with state-of-the-art research models like MT3+, validated through extensive in-house evaluations. Using industry-standard tools like mir_eval alongside our own custom metrics, we rigorously tested the algorithm on diverse, real-world audio material — and the results speak for themselves. PITCHES consistently provides high-fidelity MIDI output, preserving pitch, timing, and dynamics with exceptional precision.

Beyond MIDI conversion, PITCHES unlocks new opportunities for key detection, chord analysis, structure identification, and other forms of musical analysis, making it a powerful tool for music software and audio applications.

RESTORE

/ Adaptive Audio Restauration

RESTORE
Technology Details

The RESTORE SDK cleans audio from background noise and clicks. It enhances the sound quality of digitized analog recordings (e.g. from vinyl, shellac discs, analog audiotapes or noisy recordings). The denoiser can reduce both stationary and non-stationary noise in audio and speech signals without the need for a pre-calculated noise fingerprint. RESTORE’s de-clicking algorithm removes click sounds. Both algorithms can work adaptively, meaning they can work without user interaction and are capable of real-time processing.

New in version 1.2

• Redesigned denoising algorithm
• On par with the best in industry
• Now very transparent at high denoising levels, especially with vocals - no spectral artifacts!

The audio restoration algorithms are available on PC, MAC and Linux devices as well as on all mobile platforms. 

RETUNE

/ when c major should be c# minor

RETUNE
Technology Details

RETUNE is zplane's real-time multi-pitch modification tool. It enables shifts of pitches belonging to the same pitch class to any other pitch class (e.g. a shift of all Cs to C#s). This enables a transformation of the input audio from any input key to any output key (e.g. from C major to C# minor). Since each pitch class can be mapped arbitrarily to any other pitch class, mappings not only include major and minor input and output keys, but any other scale such as church modes or pentatonic scales. RETUNE runs in real time and the pitch mapping can likewise be modified online during processing.

In addition to the mapping of individual pitches, RETUNE can correct each pitch by mapping notes to the nearest semitone thereby eliminating inaccurate intonation. The API provides control over various parameters such as the sensitivity of the pitch detection, the detection of transients as well as the smoothing of the pitch contour.

REVERB

/ High quality reverberation

REVERB
Technology Details

Reverb is a reverberation SDK that has been thoroughly tuned with respect to sonic quality. Decorrelation and diffusity of the both stereo channels are very important for the resulting room impression; for this reason, a great effort was made with a detailed investigation of real impulse responses resulting in an optimal model for decorrelation and diffusity. The early reflections part is modeled with a mirror image source method. To allow the user to choose the best trade-off between quality and performance, z.reverb offers 4 quality modes with significant difference in the produced workload.

STEMS

/ High performance audio stem separation

STEMS
Technology Details

STEMS is our family of stem generation algorithms that is capable of separating any mono/stereo mixture into vocals, bass, drums, and ”other” (4 stems in total). Additionally, enhancement neural processing can be applied to further enhance each of the resultings stems. The enhancement is optional for each stem separately and let's you better balance performance and quality. Furthermore run our STEMS algorithms at standard latency of 2800 ms in contrast to the usual 12 sec latency. This enables fast reaction times even on slower CPUs.

STEMS PRO is our high quality version and balances quality and performance to run on desktop applications. In standard mode it is about 12 times realtime on a Apple Silicon M1 chip using only the CPU and maintains the highest industry standards in realtime stems processing.

STEMS EFFICIENT runs on low-performance CPUs, and is much faster than other available stem separation algorithms while offering competitive quality. It is approx. 34 times faster th

TONART

/ key detection and concert pitch estimation

TONART
Technology Details

TONART key detection automatically determines the key signature of any audio input signal, either monophonic or polyphonic. Key recognition can be useful in mixing applications to check if the keys of different tracks are "compatible" or for automatic generation of high-level meta data information (e.g. ID-Tags). As a special feature, TONART also detects the tuning frequency, or standard (concert) pitch of the input signal.

With v3 of TONART the detection algorithm has been completely redesigned. TONART v3 has now more than 50% better accuracy than the previous version putting it on par with the state of the art. We've also added an optimized analysis mode for electronic dance music which sometimes has less obvious keys. Finally the tuning frequency detection has been made more reliable and the API has been simplified.