Professional transcription: the easiest way to translate audio files

Do you feel like spending hours typing out audio files manually? We didn’t think so! Thanks to the latest technologies, getting a professional transcription of your audio recordings is now much more efficient and affordable. In fact, speech recognition software can already do much of the work.

What is a transcription?

A transcription is the script or written text of what is being said in a film or audio file. Someone who “transcribes” transforms speech into a written text. The term comes from Latin and can be translated literally as “to write over” (Latin trans = across, over and scribere = to write). Today, the term is also often defined as “to rewrite” or “to write down”.

What is a transcript used for?

Transcripts have many different uses in a wide range of industries, but they all have one thing in common: they save someone a lot of work. Transcribing speech into text is necessary in order to use the content for multiple purposes. For example, transcripts of film files are an enormous help during the final editing process for a TV production. Dialogues that no longer appear in the finished film can be located easily in the RAW file with the help of the transcript. Transcripts are also indispensable if the film needs subtitles. With things such as market research interviews, focus groups, clinical trials, and court hearings, transcripts are commonly needed as well.  They are worthwhile whether you need to generate evaluations or summaries, for editorial purposes, or as supporting documentation for further negotiations. The uses for transcripts are as wide-ranging as the topics covered by film and audio files.

Transcription tips: The quality of your transcript also depends on you because a transcription can only be as good as the quality of the film or audio file you give us to transcribe. Background noise or muffled speech will affect the quality and completeness of the transcript. Use these 5 tips to ensure that your transcript turns out excellently

A helping hand from artificial intelligence: “speech-to-text” software

In the current business climate, time and money are more important than ever before. Transcriptions are no exception. This is where “speech-to-text” software, a type of speech recognition programme, comes into play. You have no doubt heard of Siri, Alexa and Google Assistant. Maybe you even use one of these voice assistants on a daily basis.

Transcription software works according to this very same principle and is based on the method of “Natural Language Processing” (NLP). Using machine learning, the software recognises and processes speech and converts it into text. However, many software solutions still leave much to be desired when it comes to quality, which is why we recommend the post-editing of transcripts – just as we do with machine translation (NMT). A professional transcriber subsequently edits and adds the final stylistic touches to the machine-generated transcript. Read our blog posts about NLP, post-editing and machine translation if you want to find out even more about these topics.

Your advantage? This solution is faster and cheaper than a manual transcription but still has the same quality. However, you need to make sure that your language services provider will handle your data securely. How can you guarantee this? By making sure that they possess ISO 27001 certification for information security.

The big exception: Swiss-German audio recordings

Transcriptions of recordings in Swiss German need special treatment. Swiss German comprises a number of very distinct dialects for which no standard written language exists, meaning that a Swiss-German audio recording must first be translated into standard German before its content can then be used in other ways. This is time-consuming and usually requires a strictly manual transcription. Although there are now software solutions such as Töggl and recapp, which can recognise and transcribe Swiss-German dialects, the technology is not yet capable of handling the great linguistic variety of dialects in Switzerland. The results from these programmes are not satisfactory without a thorough post-editing process.

Technology is beneficial when the interface between human skill and machine capability functions smoothly. SwissGlobal uses technology to ensure faster and more efficient processes, the secure transfer of data, and high-quality, consistent transcriptions and translations. At SwissGlobal, all of our processes are certified according to ISO 9001, ISO 17100 and ISO 18587 standards, and we also work with ISO 27001-certified partners. Do you have any questions about audio file translations and transcriptions? Feel free to contact us. We look forward to hearing from you!