Emvoice

Creating music traditionally requires years of vocal training and natural talent. For those without musical experience or time for practice, synthesizing speech into song can provide a creative workaround. Emvoice One aims to bridge this gap by converting typed text into musical notes, allowing users to create vocal arrangements with minimal expertise. This detailed exploration of the software's features reveals how it transforms written words into audible melodies, manages complex vocal arrangements, and handles the technical challenges of integrating with digital audio workstations. Along the way, we uncover the systems behind Emvoice's voice synthesis, learn how to optimize our work with its powerful editing tools, and discover how to manage the technical demands of creating music with an AI-powered vocal plugin.

Features and Functionality

Emvoice One's interface includes vertical and horizontal scroll bar icons for precise navigation, with pinch-to-zoom gestures enabled through trackpad use. The interface also features key shortcuts like Cmd/T for text color highlighting and Cmd/B for showing numbers.

The software allows users to create and manipulate musical phrases by typing words below the piano roll, where each note corresponds to one syllable. For complex words like "Crazier," users must input multiple notes to match the syllable count. The system supports basic word substitutions and alternative pronunciations, with options for custom phoneme entry and dictionary management.

Users can adjust playback and editing through various keyboard shortcuts, including Alt-click for grid movement and Cmd/Ctrl-click for preventing semitone snapping. The software automatically syncs tempo and time signature with the host DAW project, maintaining DAW-compatible timing throughout the production process.

Setup and Installation

The software supports installation in multiple Digital Audio Workstations, including Logic Pro, Ableton Live, and Studio One, requiring a minimum of 4 CPU cores and 8GB of RAM. Each system has specific installation paths: macOS installs in /Library/Audio/Plug-Ins/, while Windows files go into C:\Program Files\Common Files\Steinberg\VST2 for VST2 and C:\Program Files\Common Files\VST3 for VST3.

When setting up the software, users should select the appropriate format - VST2.4, VST3, AAX, or Audio Units - depending on their operating system (macOS 10.13 or higher is required for Mac users). After installation, the plugin appears in the DAW's third-party instrument plugin folder, with additional files stored in the .emvoice folder within the user's home directory.

Regular maintenance requires monitoring file size, as each voice part generates backup files in the C:\Users\<username>.emvoice\one\auto-backup directory for Windows machines and Macintosh HD/Users/<username>/.emvoice/one/auto-backup for Macs. Users can manage backup history by copying files to a custom location and importing them into Emvoice One via the File > Import Song option, though the system retains only the most recent 6 edits per voice part.

For those upgrading from previous versions, the software automatically checks for updates. However, users with antivirus software blocking automatic updates need to manually download the updater from the specific locations: Windows at C:\Program Files (x86)\Common Files\emvoice\one\updater\bin\emvoiceone-updater.exe and Mac at Macintosh HD/Library/Application Support/emvoice/one/updater/bin/emvoiceone-updater.

Voice Selection and Customization

Each voice in Emvoice One comes with 10 presets, providing users with multiple starting points for their vocal arrangements. The system uses a unique phoneme-based approach to vocal synthesis, allowing users to create entirely new words by entering English phonemes between < and > symbols. For example, users can spell out "La-ge-ry" to create a custom pronunciation of "Library."

The software introduces a glottal stop feature that enables precise control over word separation, particularly when working with voices Madison and Keela. This feature can be activated by inserting "*" between words or using the "Q" character for specific phoneme positions. Users can manage these custom pronunciations through the right-click menu, where they can audition and save them to the local dictionary using a hash symbol before the number (e.g., "without#1").

The Emvoice Engine processes lyrics on an atomic level, breaking them down into groups of phonemes at various pitches. This allows for precise control over pronunciation, with users able to choose from multiple available pronunciations for many words through the plugin's right-click menu. For words not in the dictionary, users can type English phonemes directly into text boxes, with the system maintaining consistent stress distinctions when possible.

Text Editing and Note Creation

Emvoice One introduces a unique approach to text editing and note creation. Users create musical phrases by typing words below the piano roll, where each note corresponds to one syllable. For complex words like "Crazier," users must input multiple notes to match the syllable count. The system provides several tools for managing these notes, including the ability to split words using phonemes and combine notes into phrases within a single text box.

Custom Pronunciation and Editing

The software offers extensive customization options for vocal performances. Users can create custom pronunciations through the right-click menu, allowing them to edit individual phonemes and save custom words using a hash before the number (e.g., "without#1"). These custom pronunciations can be recalled through the right-click menu and loaded by typing their name into the text box. To audition all pronunciation options, users can select the desired word and preview each variation using the play button.

Technical Limitations and Workarounds

The system has specific technical limitations, with each voice part generating backup files in designated directories. Windows users find these files in C:\Users\<username>.emvoice\one\auto-backup, while Mac users access them through Macintosh HD/Users/<username>/.emvoice/one/auto-backup. The software maintains history of the most recent six edits per voice part, and users can manage backup files by copying them to a custom location and importing them via the File > Import Song option.

For users working with varying tempos and time signatures, the system requires careful management during playback. To successfully vary tempo and time signature, users must follow a specific process: copy time signature change settings, delete them in the DAW, bounce the Emvoice track to audio, and then paste the time signature changes back into place. This ensures proper synchronization between the Emvoice One plugin and the host DAW project.

Data Security and Privacy

Emvoice Company collects both personable and non-personal information from users through various interactions. Personal information, including names, addresses, and email addresses, is gathered during registration and used for customer support, communication, and service access. The company maintains this data for legal compliance and retains it as necessary for operational purposes.

Non-personal information, such as IP addresses and browser type, is automatically collected when users access Emvoice properties. This data helps in understanding user needs, improving services, and conducting site traffic analysis. While IP addresses are considered non-personal, the company reserves the right to use them for system administration and tracking property usage.

The company implements security measures to protect user information, including physical, electronic, and procedural safeguards. To ensure data protection, some personal information may be shared with third parties, particularly when conducting business development or after a company sale. In cases of legal requirements, government requests, or protection of user or public health, certain information may be disclosed.

California residents have specific rights under the Shine The Light Law to request information about personal information disclosed to third parties for direct marketing purposes. The company maintains policies to protect children under 13 and EU residents under 16, requiring parental consent for information collection. Users have several data protection rights, including the right to access information, request corrections, and request deletion of their data. The company provides contact for complaints at info@emvoiceapp.com and supports dispute resolution through an independent third-party mechanism.

For software usage, users agree to terms that prohibit political messaging or defamatory content. Service termination and suspension policies outline specific reasons for account actions, including illegal activities or security threats. The company maintains restricted rights legend for US government usage and provides detailed installation and usage instructions in the Emvoice One user guide.

Technical Requirements and Limitations

Emvoice One requires an active internet connection to function, with notes processed through Emvoice's servers before returning audio to the computer via the plugin. Network status is indicated by a white spinner during normal processing and a red spinner during connection issues, which typically resolve in milliseconds. The system operates within a vertical timeline grid, where unavailable notes are greyed out and displayed above or below the recommended voice range.

The plugin imposes specific technical limitations: each continuous note region can contain up to 20 notes, and the maximum number of voices supported simultaneously is four. Note creation and editing are managed through two primary tools—the normal tool for precise editing and the pencil tool for rapid drawing. Pitch and timing adjustments are made through the right-click menu options, including Split Region for dividing selected regions.

Each voice has a defined range that limits available notes, with darker shading indicating pitches above or below the recommended range. The system maintains strict monophonic functionality, recognizing only the latest note when additional notes are added over existing content. Users can create harmonies by flattening notes and copying regions between multiple instances of the plugin.

File Size and Backup Management

File size management is essential for maintaining optimal system performance. Each voice part generates backup files stored in designated directories: C:\Users\<username>.emvoice\one\auto-backup for Windows users and Macintosh HD/Users/<username>/.emvoice/one/auto-backup for Mac users. The plugin automatically maintains a history of the most recent six edits per voice part, allowing users to recover previous versions through the File > Import Song feature. Users can manage backup files by copying them to a custom location and importing them into Emvoice One via the File > Import Song option.

Software Version and Compatibility

The latest version of Emvoice One installs updates automatically, but users can disable this feature through the plugin's settings menu. Installation requires compatibility with VST2.4, VST3, AAX, and Audio Units formats on both PC and Mac systems, with macOS support beginning at version 10.13 (High Sierra). To handle large projects, users should ensure sufficient CPU cores and RAM: at minimum, 4 cores and 8GB of RAM are required for basic operations.

For users upgrading from previous versions, the system automatically checks for updates. However, those with antivirus software blocking automatic updates need to manually download the updater from specific paths: C:\Program Files (x86)\Common Files\emvoice\one\updater\bin\emvoiceone-updater.exe for Windows machines and Macintosh HD/Library/Application Support/emvoice/one/updater/bin/emvoiceone-updater for Mac users. Regular maintenance includes monitoring file size, particularly within the C:\Users\<username>.emvoice\one\auto-backup directory for Windows machines and Macintosh HD/Users/<username>/.emvoice/one/auto-backup for Macs.