Transcription - from speech to text (and metatext)


Public organisations, research institutes and the media are producing, receiving and storing increasing volumes of audio and video data.  In many cases, spoken content must also be converted into text for more detailed analysis, post-processing and archiving. 

Lingsoft transcribes interviews, dictation, speeches, phone calls, meetings as well as records of hearings and legal proceedings effectively and to the highest quality standards, ensuring a suitable degree of accuracy no matter the number of speakers in any given recording. In addition to the transcription of colloquial and standard language, we can also perform verbatim transcriptions, where filler words ("like"), repetitions ("and-and"), hesitations ("um, like") and non-verbal communication (pauses, laughing and coughs) are included in the text. If you need to process multilingual materials, we also offer interpreting transcription from one language to another. 

In our transcription process, we utilise automatic speech recognition, which is used to convert recorded speech into raw text before transcription. For the transcription work itself, we use Lingsoft's quality checkers and automated annotation technology.