» Dezrann documentation

Translating / Localizing Dezrann

We are extremely grateful to all translators of Dezrann who help to provide users worldwide with the best experience.

Preliminaries

Understand Dezrann. The target audience of Dezrann includes both scholars analyzing and/or browsing a piece of corpus, and children/students/people (either from the general public, or from music courses) hearing music and possibly discovering bits of music analysis.

Do not work alone. Talk with translators in other languages and developers to see what are the challenges in writing UI and corpora text.

A full experience for the audience. Translating/localizing Dezrann involves both translating the User Interface (UI) and translating the titles and descriptions of corpora pages.

1. Localizing Dezrann UI

Special terms, be consistent. Make sure that you use the same terminology, phrasing, and tone throughout the UI. Begin by discussing the most important terms (some of them are in elements.json:generic), but they are used across many items.

en analysis layer label type (category) tag (value) comment
de Analyse Schicht Label Typ Tag Kommentar
hr analiza sloj oznaka? vrsta oznaka kommentar
el Ανάλυση Επίπεδο Ετικέτα Τύπος ? Σχόλιο
fr analyse couche étiquette type tag commentaire
sl analize sloj etiketo vrsta oznaka kommentar
it analisi strato etichetta tipo tag commento
zh-Hans 分析 标注 类型 标签 评论

An analysis is a set of labels. Each label is defined:

The label word appears in many places in the application, it should be a very common word. (And please do not use the same term for tag, we kept in Greek/French/Italian ‘tag’, even if it’s an English word).

Keep the user in mind. Remember that UI is all about the user. Make sure that your translations are designed to make the user’s experience as smooth and intuitive as possible. Use language and formatting that is familiar and easy to understand, and consider the user’s cultural background and expectations.

Prioritize clarity. Users should be able to quickly and easily understand the meaning of each element in the interface. Avoid using idiomatic expressions or overly complex vocabulary that could be confusing to non-native speakers. When faced with character limits, prioritize clarity and essential meaning over stylistic flourishes or structural elements.

2. Localizing Dezrann corpora

To give a full experience to the user, we also translate/localize the corpora pages accessible from https://www.dezrann.net/corpora.

Human expertise/curation is the key. As of 2025, these texts together make up 1 of 2 pages of text. To initiate your translations, you may use automated translating services such as Large Language Models. However, sometimes, automated translations completely misinterpret the meaning. Anyway, you can either start from these automated translations, then correct them, or create translations from scratch based on the English version corpora-descriptions.json.

Special terms. Please pay special attention to technical details/terms in music (such as classical style, sonata form, harmony, etc.). For several corpora, we also mention that the data is “free/open” (“libre”), meaning it is not just free of charge (“gratis”). Please select the most relevant term for your language.

The strings to be translated.

How to

The following instructions are if you are a git user. If you are not confortable with git, you can send a mail with the files to someone in the team.

You can directly work with the files in this directory and/or use a tool such as the i18n Ally extension for VS Code.

Step 1, UI, on src/locales of the git dezrann-front

Step 2, Corpora pages, on metadata on the git dezrann-corpus

Step 3 (later, once a first version is online)

Further resources