How sound works in Vocaberry apps
This page describes the technical principle by which sound is produced in the applications of the Vocaberry ecosystem.
All sound is generated directly on the user's device at the moment of playback.
Without software processing such a file is silent: for playback, a program must interpret the events and synthesize the sound.
Audio is not transmitted over the network. Only the score is downloaded from Vocaberry servers — a file of note events a few kilobytes in size. Synthesis is performed entirely on the device.
Consequently, the accompaniment has a characteristic synthesized timbre and is not a recording of an instrumental performance.
No audio recordings are used
Vocaberry applications neither contain nor play back audio recordings. No audio streaming is employed and no performance by any artist is reproduced: the applications include neither a recording player nor a streaming client.All sound is generated directly on the user's device at the moment of playback.
Format in which musical material is stored
Musical material is stored in the MIDI format. MIDI is not an audio format: a file contains a sequence of events — the onset and release of a note, its pitch, velocity, duration and the tempo of performance. The file contains no audio data and no waveform.Without software processing such a file is silent: for playback, a program must interpret the events and synthesize the sound.
Sound synthesis on the user's device
Synthesis is performed by a software module that forms part of the application and runs on the user's device. The module interprets the events of the score and generates, in real time, a tone of the required pitch and duration.Audio is not transmitted over the network. Only the score is downloaded from Vocaberry servers — a file of note events a few kilobytes in size. Synthesis is performed entirely on the device.
Consequently, the accompaniment has a characteristic synthesized timbre and is not a recording of an instrumental performance.
Capabilities afforded by this approach
Storing musical material as a score provides functions unattainable with an audio recording:- Transposition. A piece may be moved into a key matching the user's range by shifting all notes. The timbral artifacts inevitable when processing a finished audio recording do not arise.
- Tempo adjustment. The tempo may be reduced for practising difficult passages without altering pitch.
- Part control. Individual parts, including the melody, may be muted or retained as a reference.
- Display of musical notation. Data on the pitch and duration of each note make it possible to render the staff, highlight the current note and display the fingering chart for the recorder.
- Analysis of the user's performance. During a session the application receives the microphone signal, determines the pitch produced by the user and compares it with the note in the score, which allows the accuracy of intonation to be assessed.
No third-party services are used
Vocaberry applications do not connect to music streaming services, do not query third-party catalogs and do not use third-party music discovery services. The list of works is retrieved from Vocaberry servers.Summary
- No audio recordings are present in the applications, whether bundled or streamed.
- Musical material is stored in the MIDI format: pitch, duration and tempo events, without audio data.
- Sound is synthesized on the user's device at the moment of playback. Only the score is transmitted over the network.
- This approach provides transposition, tempo adjustment, part control and the display of musical notation.
- No third-party music services are used.