By Martel Electronics · November 10, 2026 · Manual transcription and review

Focus-group transcription becomes difficult when participants answer together, interrupt, or respond with short acknowledgments. Video provides additional context, but it does not automatically solve speaker attribution. The visible participant is not always the person whose voice is clearest in the recording.

A useful workflow separates the first identification pass from detailed wording and final review. Martel supplies playback control and text timestamps; you supply the listening, labeling, and interpretation required by the research project.

Establish the project's notation

Before transcribing, agree on participant labels, overlap notation, uncertainty markers, and the level of verbatim detail. Decide whether laughter, pauses, gestures, or group reactions matter to the study. Keep participant identifiers consistent with the project's approved records.

Use the same label throughout the transcript even when the camera changes angle. If an identity is uncertain, retain a neutral placeholder until it is verified. Do not turn a guess into a permanent label because it makes the page easier to read.

Map the conversation in a first pass

Watch a short section at normal speed and note the sequence of turns. Identify the moderator's question and the first response before focusing on overlapping details. Insert a timestamp at a difficult exchange so you can return without searching the whole recording.

When participants talk at once, listen for the beginning and end of each audible contribution. Replaying a smaller segment can help, but keep enough surrounding context to avoid attributing a fragment to the wrong turn.

Use the picture as supporting information

Enlarge the video when mouth movement, seating position, or a gesture may help attribution. Treat those cues cautiously when the picture is delayed, the speaker is out of frame, or the camera switches views. A participant nodding on screen is not proof that the same person made an audible acknowledgment.

Separate visible actions from spoken text. A note such as [several participants laugh] is an observation; it should not be formatted as a quotation by a specific participant without supporting evidence.

Control pace without losing the exchange

Use pedal playback and short rewind steps to review a passage while keeping your hands near the document. Slower playback can aid difficult listening, but compare the result at normal speed before finalizing it. Altered pace can change your impression of pauses and turn-taking.

Timestamps remain tied to media position when speed changes. A marker at 00:12:30 still identifies that point in the video, regardless of how long you spent reviewing it.

Choose markers that help the next reviewer

Automatic paragraph timestamps can suit a transcript where each Enter starts a new speaker turn. Fixed intervals may help a longer continuous discussion. For an unresolved overlap, add a deliberate marker and a concise review note rather than assuming the automatic convention identifies the exact onset.

Include the source filename and timing basis in the header. If the group recording is split into several files, label each transition and retain separate elapsed references.

Finish uncertainty honestly

After reasonable review, some overlapping words may remain unclear. Preserve that limitation using the agreed notation. Do not combine plausible fragments into a fluent sentence that no participant actually spoke.

Save the transcript separately from the player, and follow the study's consent, access, and handling requirements. Martel helps navigate the recording; it does not provide participant identification, qualitative coding, or automatic speech recognition.