A method of generating a transcript file in a selected presentation format from input audio data with a transcription component. The transcription component divides the input audio data into individual sound tokens. The transcription component then identifies transcription text for subsets of the sound tokens by finding a best match for the subset in sound samples in a sound database. The transcription component then creates a transcript file and formats the transcription text in the transcript file according to a presentation format that corresponds to a sound type of the transcription text.

Detecting vibrato bar technique for a string instrument can include analyzing, using a processor, a note signal of the string instrument to detect a selected instrumental technique from a plurality of instrumental techniques, analyzing, using the processor, a noise signal of the string instrument to detect a change in frequency of the noise signal, and generating, using the processor, a vibrato bar event responsive to detecting the selected instrumental technique and the change in frequency of the noise signal.

An apparatus to detect, analyze, record, and display audio data, including an input unit to allow a user to input musical notes corresponding to the audio data, a processor to analyze the musical notes and to save the musical notes into a file, and a display unit to display notes corresponding to the musical notes on a virtual instrument.

Techniques for generating a virtual music experience. The techniques include source separating an arbitrary digital audio input into a plurality of source-separated tracks. Sets of music features are determined from the plurality of source-separated tracks and provided to a video presentation system at a video frame rate of the video presentation system. The providing the sets of music features to the video presentation system causes the video presentation system to animate one or more graphical assets based on the provided sets of music features.

In some implementations, a process for communicating data over audio is performed. In one aspect, one or more ordered sequences of audio attribute values that are selected based on a musical relationship between the audio attribute values and associated with data values may be played by a first device and received by a second device. This technique may allow for sound-based communications to take place between devices that listeners may find pleasant.

The present disclosure describes techniques for implementing automatic music audio transcription. A deep neural network model may be configured. The deep neural network model comprises a spectral cross-attention sub-model configured to project a spectral representation of each time step t, denoted as St, into a set of latent arrays at the time step t, denoted as .sub.t.sup.h, h representing an h-th iteration. The deep neutral network model comprises a plurality of latent transformers configured to perform self-attention on the set of latent arrays .sub.t.sup.h. The deep neural network model further comprises a set of temporal transformers configured to enable communications between any pairs of latent arrays .sub.t.sup.hat different time steps. Training data may be augmented by randomly mixing a plurality of types of datasets comprising a vocal dataset and an instrument dataset. The deep neural network model may be trained using the augmented training data.

The present application provides techniques for displaying music score segments in target music videos. The techniques comprise determining a digital music score corresponding to a piece of music comprised in a target music video; determining a segment of the digital music score corresponding to a current playing progress of the target music video based at least in part on a playing progress of the target music video; generating an image of a music score segment corresponding to the segment of the digital music score based on a predetermined condition; and presenting the image on a corresponding interface of playing the target music video.

A platform or system is disclosed for performing musical analysis to detect musical properties in received live or pre-recorded audio data. The analysis can include a synchronous analysis for generating estimated one or more transitory musical properties and an asynchronous analysis for generating one or more aggregate musical properties which can be applied to the transitory musical properties to generate confirmed musical properties, which can be stored as metadata associated with an audio file. In some cases, live audio data can be received, recorded, dynamically analyzed to provide realtime metadata (e.g., to a display), then the realtime metadata can be analyzed to provide confirmed, updated, or validated metadata. In some cases, initial analysis (e.g., dynamic analysis) can determine chord estimates, usable in further analysis (e.g., offline analysis) to estimate a musical key, which can then be applied to the chord estimates to determine the most likely chord estimates and determine chord progressions.

In some implementations, a process for communicating data over audio is performed. In one aspect, one or more ordered sequences of audio attribute values that are selected based on a musical relationship between the audio attribute values and associated with data values may be played by a first device and received by a second device. This technique may allow for sound-based communications to take place between devices that listeners may find pleasant.

A platform or system is disclosed for performing musical analysis to detect musical properties in received live or pre-recorded audio data. The analysis can include a synchronous analysis for generating estimated one or more transitory musical properties and an asynchronous analysis for generating one or more aggregate musical properties which can be applied to the transitory musical properties to generate confirmed musical properties, which can be stored as metadata associated with an audio file. In some cases, live audio data can be received, recorded, dynamically analyzed to provide realtime metadata (e.g., to a display), then the realtime metadata can be analyzed to provide confirmed, updated, or validated metadata. In some cases, initial analysis (e.g., dynamic analysis) can determine chord estimates, usable in further analysis (e.g., offline analysis) to estimate a musical key, which can then be applied to the chord estimates to determine the most likely chord estimates and determine chord progressions.