ZuMult is a corpus platform for audiovisual corpora. You can get further information and news about ZuMult on: zumult.org • ZuMult@Zenodo • ZuMult@GitHub .
This ZuMult instance was set up for the Transcription+ project in a collaboration between the Institute for Communication Studies at UDE and the Academy of Sciences and Humanities in Hamburg, funded as a cooperation project of the NFDI consortium Text+.
This ZuMult instance has 1
corpus:
EXMARaLDA-DemoKorpus.
The corpus overview lists all corpora and has links to speech event and speaker overviews.
Corpora in ZuMult can be queried on all annotation levels with ZuRecht, using the query language CQP. Query results will be displayed in a KWIC giving access to underlying audio or video and to the wider transcript context.
CQP Query (ZuRecht)
Transcripts are visualised by ZuViel in an interactive, configurable view with synchronisation of audio/video and transcript text. A density viewer, a wordlist and a search field provide further means of navigating the transcript.
The following transcript was chosen at random:
EXMARaLDA-DemoKorpus / NTVshniki
View and Explore Transcripts (ZuViel)
ZuPass visualises transcripts in musical score (Partitur) notation.
ZuAnn visualises transcripts with all annotations.
The following transcript was chosen at random:
EXMARaLDA-DemoKorpus / NTVshniki
Transcript in musical score notation (ZuPass)
Transcript with annotations (ZuAnn)
For any contribution in a transcript, details can be explored by zooming in with ZuMin. The view provides a phoneme level alignment (using MAUS), a visualisation of the pitch contour (using Praat), and video stills extracted each 0.5 seconds.
The following contribution was chosen at random
EXMARaLDA-DemoKorpus / NTVshniki / au1_seg0
Transcript Details View (ZuMin)
Basic statistics for a corpus summarize the number of speech events, speakers, transcripts, tokens and recordings in a corpus
The following corpus was chosen at random:
EXMARaLDA-DemoKorpus
Corpus statistics
Speech events and speakers of a corpus are displayed with their metadata in sortable and searchable tables.
The following corpus was chosen at random:
EXMARaLDA-DemoKorpus