How to Prepare Audio for Transcription: A Researcher’s Checklist

A good research transcript starts before the audio file reaches the transcriber.

Recording quality, speaker identification, terminology and clear instructions can all affect the accuracy of the finished transcript. Spending a little time preparing audio for transcription can make transcription easier, reduce the number of unclear sections and give you a more useful document for qualitative analysis.

This is particularly important for researchers conducting interviews, focus groups, oral history projects, healthcare research or other studies where the wording and context of what participants say may form part of the research data.

At The Typing Works, we have been providing professional transcription services since 1993 and regularly work with universities, academic researchers and research organisations. Based on that experience, here is our practical checklist for preparing audio for professional transcription.

1. Choose the right recording environment

If you have control over where an interview takes place, think about the audio before you begin.

A quiet room will almost always produce a better recording than a café, shared office or busy public space. Background conversations can be particularly problematic because recording equipment may pick them up almost as clearly as the person you are interviewing.

Hard surfaces can also create echoes. If possible, choose a room containing soft furnishings and close windows if there is significant traffic or construction noise outside.

For remote research interviews, ask participants to join from somewhere reasonably quiet if circumstances allow. Headsets or external microphones can improve recording quality, but even moving away from an open window or noisy kitchen can make a noticeable difference.

2. Test your recording equipment before the interview

Do a short test recording before beginning your research session.

Record yourself speaking from approximately the position your participant will occupy, then listen back using headphones. Check whether your voice is easy to understand and whether any background sound is dominating the recording.

If several people will be taking part, test the microphone with people sitting at different distances from it.

You should also check available storage space and battery levels before starting. For lengthy interviews or focus groups, having a backup recording option may be worth considering where this is permitted by your research protocol.

A one-minute test can prevent an hour-long interview being recorded at an unusably low volume.

3. Position the microphone carefully

The best position depends on your recording equipment, but the general principle is simple: the microphone needs to capture every participant clearly.

For a one-to-one interview, avoid placing the recorder significantly closer to the interviewer than the participant. If the research participant is much quieter on the recording, their responses may be harder to transcribe than your questions.

For focus groups, position the recording equipment as centrally as reasonably possible. A single device placed at one end of a long table can make speakers at the opposite end difficult to hear.

Try to avoid placing microphones directly beside laptops, cups, paperwork or anything that may be moved repeatedly during the conversation.

Suggested internal link: Link “focus groups” to your Focus Group Transcription page.

4. Think about speaker identification

Speaker identification is much easier when the researcher plans for it in advance.

In a one-to-one interview, distinguishing between interviewer and participant is usually straightforward. In a focus group containing several voices, it can be considerably more difficult.

If your research protocol allows it, ask participants to introduce themselves at the beginning of the recording. Where anonymity is required, they can use participant numbers or agreed pseudonyms instead.

During a focus group, the moderator can occasionally use a participant’s agreed identifier naturally when responding to them. This can help establish which voice belongs to which participant.

If you already have a speaker list, seating plan or list of participant codes, supplying it to your transcription company can be extremely helpful.

Remember that a transcriber cannot reliably identify somebody they have never met simply from the sound of their voice.

5. Discourage people from speaking over each other where possible

Overlapping speech is one of the biggest challenges in focus group transcription.

When several people speak simultaneously, even an experienced human transcriber may be unable to distinguish every word. The problem is not necessarily transcription skill. Sometimes the words simply cannot be separated within the recording.

A good moderator can help by encouraging participants to let each other finish. This does not mean making a lively focus group feel unnatural. It simply means intervening when prolonged cross-talk is making the discussion impossible to follow.

If two participants repeatedly speak at the same time, the recording may contain information that no transcription service can fully recover.

For researchers preparing material for detailed qualitative analysis, reducing overlapping speech at the recording stage can make a substantial difference to the finished dataset.

6. Provide a terminology or vocabulary list

Researchers frequently work with terminology that is unfamiliar outside their specialist field.

You may have the clearest recording imaginable, but a transcriber still needs to recognise the words being used.

Before sending your files, consider providing a short vocabulary list containing specialist terminology, unusual surnames, organisation names and relevant abbreviations.

You do not need to create an exhaustive glossary. Concentrate on terms that someone outside your research area might reasonably struggle to identify.

This can be particularly useful for medical research, technical interviews and projects involving organisations with numerous internal abbreviations.

Relevant papers or publicly available project information can also help provide context, provided sharing them is appropriate under your research arrangements.

7. Decide what type of transcript you actually need

Not every research project requires the same transcription style.

Some researchers need a highly detailed verbatim record containing repetitions, hesitations and other features of spoken language. Others need an intelligent verbatim transcript where unnecessary fillers are removed without changing the participant’s meaning.

Consider how you intend to analyse the material before deciding.

If pauses, speech patterns or verbal behaviour form part of your methodology, tell your transcription provider. If your analysis is primarily concerned with what was said rather than the mechanics of speech, a cleaner transcript may be easier to work with.

You may also have requirements for timestamps, line numbering, speaker labels or a particular document format.

Providing these instructions before transcription begins helps ensure that the finished documents are consistent across the whole research project.

8. Decide how anonymisation should be handled

If transcripts need to be anonymised, agree the approach before transcription begins.

For example, you may want personal names replaced with participant codes, place names removed or identifying details marked using a consistent placeholder.

Give the transcription company clear instructions about what should be anonymised and how the replacement should appear.

Researchers should also consider whether identifying information needs to be present in the recording at all. The Information Commissioner’s Office advises organisations processing personal data to consider data minimisation and, for research, whether anonymisation or pseudonymisation can be used where appropriate.

The exact approach will depend on your project, ethics approval and institutional requirements, so researchers should follow the policies applying to their own study.

9. Check your university or organisation’s data requirements

Research audio may contain personal information or sensitive material, so file handling should not be treated as an afterthought.

Before using an external transcription provider, check whether your university, NHS organisation, research institute or funder has specific requirements concerning data processing.

You may need information about where files are stored, how they are transferred, who can access them and when they are deleted.

You may also require a Data Processing Agreement, confidentiality agreement or other project-specific documentation.

The Typing Works provides UK-based human transcription and offers secure UK-hosted file transfer options. Our client data is not processed through AI transcription systems. We can also work with organisation-specific security requirements where agreed.

For researchers dealing with particularly sensitive material, it is worth discussing security requirements with the transcription provider before uploading any recordings.

10. Rename and organise your files before uploading them

A clear file-naming system becomes increasingly important as a project grows.

Avoid sending a folder containing files with generic names such as:

Recording1.mp3
Recording2.mp3
New Recording.mp3
Audio Final.mp3

Instead, choose a consistent format appropriate to your project, for example:

StudyA_Participant04_Interview.wav
StudyA_Participant05_Interview.wav

If participant identities must be protected, use the study identifiers already approved for your project rather than names.

Make sure the filename corresponds with any instructions, participant information or tracking spreadsheet you provide.

For larger projects, consistent file naming makes it easier to match transcripts with recordings and reduces the risk of administrative errors.

11. Listen briefly to the recording before sending it

You do not need to listen to every recording from beginning to end before submitting it for transcription.

It is sensible, however, to check the beginning, part of the middle and the end of the recording.

Listen for unexpectedly low volume, missing sections or serious background noise. If there is a known issue, tell your transcription provider.

For example:

“Participant is very quiet for the first ten minutes.”

“Building work starts at approximately 42 minutes.”

“Interviewer leaves the room briefly at around 55 minutes.”

Information like this helps the transcriber understand what they are hearing and prevents time being wasted trying to solve a problem that the researcher already knows about.

12. Send instructions with the audio

Do not assume that your transcription provider will automatically know how you want your research material presented.

Before transcription begins, confirm your requirements in writing.

Useful information can include the required transcription style, speaker labels, anonymisation rules and formatting preferences. For an ongoing research project, it is worth agreeing a standard template that can be followed consistently for every recording.

You can also provide your required turnaround time and tell the transcription team whether some interviews need to be prioritised.

Clear instructions at the beginning of a project are considerably easier than trying to alter dozens of completed transcripts later.

13. Transfer research recordings securely

Audio recordings containing research participants’ voices should be handled with appropriate care.

Avoid moving sensitive files between systems unnecessarily. Use the transfer method approved by your institution and check that your chosen transcription provider can meet any specific security requirements before sending the data.

The Typing Works offers a private UK-based upload facility for clients who need a secure method of transferring research recordings for transcription.

The ICO’s research guidance also emphasises the importance of appropriate technical and organisational safeguards when personal data is processed for research purposes.

14. Keep the transcriber informed about your research context

You do not need to give your transcriber a detailed account of your research methodology, but some context can improve the transcript.

A short explanation such as:

“This is a study of occupational therapy provision in rural communities.”

can immediately help the transcriber interpret terminology and understand the subject being discussed.

If your research includes repeated interviews, tell the transcription provider that the recordings form part of the same project. The same terminology and formatting conventions can then be used consistently.

Context is particularly valuable when recordings involve specialist language or references that would be obvious to someone within the research project but unfamiliar to somebody outside it.

A final research audio checklist

Before you send your recording for transcription, check the following:

• Have I listened to part of the recording to confirm the audio has worked?
• Can the main speakers be heard clearly?
• Have I provided participant codes or speaker information where appropriate?
• Have I supplied any specialist terminology that may help the transcriber?
• Have I explained the transcription style I require?
• Have I provided anonymisation instructions where necessary?
• Have I checked my institution’s data handling requirements?
• Are the files named consistently?
• Have I included formatting or timestamp requirements?
• Am I using an approved secure method to transfer the recordings?

Professional academic transcription from The Typing Works

Preparing research audio properly can save considerable time later in a project. More importantly, it can improve the quality and usefulness of the transcripts on which your analysis may depend.

The Typing Works has provided transcription services since 1993. We specialise in 100% human transcription and regularly support university researchers with interviews, focus groups and other qualitative research recordings.

Our UK-based transcribers are experienced in working with specialist terminology and multi-speaker recordings. We can accommodate project-specific formatting, speaker identification and anonymisation requirements, and our team understands the importance of confidentiality when handling research material.

If you are planning a research project and would like advice about preparing your recordings, you are welcome to contact us before you start. We are happy to discuss your requirements and help establish a consistent transcription process for your study.