What Does Transcribe Mean? Definition, Examples and Types of Transcription
When someone says ‘transcribe’ they just mean take the spoken words and put them in writing. It could be an interview that’s been recorded, a business meeting, a uni lecture, a video call etc. Rather than having to go back and listen to the recording over and over again you end up with a written version which you can then read, edit, pass on and/or refer to at will.
Transcription is used all over the place in education, media, research and business. It’s also a key enabler of making audio and video content accessible to people in different languages and different parts of the world.
TL;DR:
To transcribe is to take spoken words from an audio or video file and write them down. The various types of transcription are verbatim, intelligent verbatim, edited and phonetic. Transcription is used by businesses for a variety of reasons including documentation, content creation, research and localisation. It’s important to note that this is not the same as translation as the language of the original spoken content does not have to be changed.
What Does Transcribe Mean? Understanding Transcription
In simple terms transcription is just listening to something and writing down what you hear. The process is called transcription and the result is a transcript.
For instance if a journalist does an interview and then writes down the interview. That’s transcription of the interview.
The term is also used in the context of music and biology, though in the course of normal professional discourse it’s most frequently used to describe the process of converting an audio or video recording into text.
What Are the Different Types of Transcription?
Not every recording needs to be transcribed in the same way. The format will depend on where and how the transcript is going to be used.
1. Verbatim Transcription
This approach captures precisely what the person says (including repetitions, pauses, fillers and all the other oddities of speech). It’s useful for interviews, for legal proceedings and for any kind of qualitative work where you want to capture the actual spoken word.
2. Intelligent Verbatim Transcription
Here the point is made, the meaning is there but you’ve got all the filler words and the repetition of words removed. It’s a much more natural piece of writing without in any way altering what the person actually wants to say.
3. Edited or Clean Read Transcription
This is a readability format. Grammar, sentence structure and the like are all fixed up, the meaning is preserved. It’s the format you’ll see in business meetings, articles, reports and the like.
4. Phonetic Transcription
Phonetic transcription is the representation of speech sounds using a set of phonetic symbols. It’s used by linguists and language researchers to look at pronunciation, speech patterns and the differences between languages.
How Does Transcription Work?
The process typically starts with an audio or video recording. The transcriber will listen to the recording, determine who is speaking and then transcribe the spoken words into text.
There are two common approaches:
- Manual transcription: a human transcriptionist listens to the recording and types out the conversation. That way they can get the context, they can get the accents and they can pick up on the little differences in speech.
- Automated transcription: speech to text software does the recording and creates the transcript. It’s generally much faster though the output will invariably require some level of human review.
Before delivery the transcript must be checked for missing words, incorrect speaker labels, punctuation and terminology.
What Is the Difference Between Audio and Video Transcription?
Audio transcription is the process of converting the spoken word from a recording (e.g. an interview, a phone call, a podcast) into text.
Video transcription does the same with video files (webinars, product demos, recorded presentations etc). Depending on the project it may also include timestamps, speaker ID and relevant non speech sounds.
What Is the Difference Between Transcription, Translation and Localization?
These three processes are linked but they’re all for different purposes.
- Transcription: The process of converting speech to text (typically in the original language).
- Translation is the process of converting one language into another (from written or spoken).
- Localization: adapts the content to the language, culture and expectations of the target audience.
Consider an English language training video for a German audience. You transcribe the English speech. You then translate that transcript into German and localise it so that the examples, expressions and terminology are appropriate for a German audience.
This is where VerboLabs becomes a useful partner. It enables businesses to view transcription as part of a wider multilingual content workflow rather than as a standalone task.
How does transcription fit into a localization workflow?
A transcript is frequently the basis for subtitles, captions, voice over and translated video.
For example a company wanting to run a product demo in a number of countries could begin with a perfectly accurate transcript. That transcript can then be translated, subtitled and localised.
Working with an experienced language service provider such as VerboLabs will of course go a long way to ensuring consistency across those stages.
What is transcription used for?
Transcription has practical applications across several industries:
- Education: turning lectures, discussions and lessons into study material.
- Business: Producing the written record of meetings, presentations and interviews.
- Media: Preparing transcripts for podcasts, documentaries and video.
- Research: recording interviews and focus group discussions for analysis.
- Content creation: taking recorded conversations and turning them into articles, blogs and social media content.
- Localization: getting spoken content ready for translation and for distribution in a number of languages.
What Affects Transcription Accuracy?
A transcript is only as good as the recording and the recording process. There are a number of things which can affect the quality of the transcript; background noise, poor mic quality, people talking over each other, obscure jargon and regional accents being the main ones.
Names, technical terms and speaker ID all warrant a look. A small transcription error can sometimes completely alter the meaning of a sentence.
How Do Accents and Regional Languages Affect Transcription?
People say words differently depending on their region, their dialect and their idiosyncrasies as speakers. Some of the recordings will contain code switching where the speaker will switch from one language to another during the course of the conversation.
Automated tools can have a hard time with these variations. A human reviewer who is familiar with the language and the context is going to be able to spot the errors that the software may well miss.
What Is the Difference Between Transcription and Captioning?
Transcription is a written record of what was said. Captions are timed to the audio/video and can include relevant sound descriptions.
A transcript of a webinar can be used to generate captions which then appear alongside the video.
Human Transcription vs. Automated Transcription: What’s the Difference?
Automated transcription is ideal when you’re after speed and you’re dealing with a large volume of recordings. It can however miss the context, it can get the speakers wrong and it can have problems with accents and background noise.
Human transcription gives you a far better understanding of the context and the subtleties of language, albeit it’s generally a longer process.
For a lot of the professional work we do a hybrid approach works really well. The initial transcript is generated by the automated tool and then the experienced human reviewer goes in and does the correction and polishing.
What Should You Consider Before Transcribing Audio or Video?
Before starting a transcription project, consider:
- Recording quality and background noise.
- Language, dialect and regional variations.
- Number of speakers.
- Intended use of the transcript.
- Required transcription style.
- Whether we need the timestamps and speaker labels.
- Formatting and proofreading requirements.
Defining those details in advance makes the final transcript far more useful.
When Do You Need Professional Transcription Services?
If you’re regularly doing interviews, multilingual recordings, technical stuff, lots of audio and video then professional transcription services can save you a tremendous amount of time.
VerboLabs is definitely worth looking at if you’re after something more than a basic speech to text solution. It’s very much a language services/localisation type of company and so is a very sensible option if your project is going to involve transcription leading on to translation, subtitling etc.
Rather than treating transcription as a discrete activity, the business can benefit from working with a language partner who appreciates the importance of accuracy, context and consistency.
Check out VerboLabs’ transcription services to see how professional language support can be integrated into your content workflow.
Conclusion
To transcribe is to turn spoken words into written text. Producing a good transcript is more than just typing what you hear. Accuracy, context, language and the purpose all play a part in the end product.
Whether you’re after a transcript of a meeting, an interview or a multi language video project the right transcription approach is key. Given its wider language services offering VerboLabs is a very good option for businesses who want to bring their transcription in house as part of their wider localisation offering.
Frequently Asked Questions
Is transcribing the same as translating?
No. Transcription is the process of turning speech into text. Translation is the process of converting content from one language to another.
What is the difference between a transcript and transcription?
Transcription is the process of converting speech to text. The transcript is the resulting piece of written text.
Can you transcribe both audio and video?
Yes. Both audio recordings and video files can be transcribed into text.
Is AI transcription accurate?
AI transcription can be extremely accurate (particularly with clear recordings). It can be less so with accents, in the presence of background noise and when people are talking over each other. For professional use human proofreading is recommended.
How long does it take to transcribe an hour of audio?
It varies enormously depending on the quality of the recording, the number of speakers, the language and the style of transcription. A manual transcription can take several hours for just one hour of recorded speech.
Can transcription services handle multiple languages?
Yes. Professional transcription services can do recordings in a number of languages (depending on their language capability).
How do I choose a professional transcription service?
Look at language support, accuracy, confidentiality, turnaround time, proofreading and experience with your particular type of content.
What makes a transcription difficult?
Poor audio quality, strange accents, a lot of technical jargon, multiple speakers and overlapping voices can all make transcription a real pain.
Why is transcription important for businesses?
It helps businesses to create searchable records, document meetings, repurpose content and get material ready for translation and localisation.
What is multilingual transcription?
It’s about taking spoken content in a number of languages and turning it into text, taking into account language specific expressions, pronunciation and context.