The Editor Approved the Interview. The Bulletin Played a Different File.

The bulletin is three minutes away. In a Nairobi newsroom, the producer has approved the interview, the presenter has the introduction, and the editor has already moved on to the next story. Everyone remembers hearing the guest clearly.

The introduction ends. The package starts. Pictures roll, the music arrives, and the guest’s first answer is missing.

Someone calls the editor. “But it was working here,” comes the reply. And that may be completely true. The editing session sounded correct. What the audience received was a different stage of the production chain.

Consider this an illustrative newsroom scenario, not a report of a particular station’s failure. It captures a costly assumption: approving the sound inside an editing application is treated as approval of the file that will actually go to air.

Those are not the same thing. A session, an exported file, an imported playout asset and the programme output are related, but each can behave differently. Reliability comes from checking the handovers between them rather than trusting the first successful listen.

Inside an editing session, the software is combining source clips, track routing, effects, automation and output settings. The editor’s headphones or speakers reproduce whichever monitoring path is selected. Hearing the voice there proves that it reaches that listening path. It does not automatically prove that the export contains the same combination of signals.

An export can use the wrong sequence, a restricted range, a different output bus or an unintended channel layout. A track heard during editing may not feed the selected export output. An older export preset may also be inappropriate for the current job. The picture can look right while the audio handover is wrong.

That is why the first post-export listen must be to the exported file itself. Playing the timeline again is reassuring, but it repeats the earlier test. It does not test the new object the editor has just created.

Close or minimise the editing session and open the actual delivery file in a suitable player or checking application. Verify the filename and location before listening. The purpose is to break the comfortable assumption that everything called “final” contains what the editor remembers mixing.

Listen to the first spoken words, a quieter answer, a louder moment and the end. Check any edit where music, narration or an interview changes. Short spot checks are useful when time is tight, but they can miss a problem between those points. A full listen to the final export is the stronger check when the schedule allows it.

Pay attention to what is being delivered on each channel. Some workflows require a finished stereo mix. Others deliberately carry separate elements on different channels for downstream control. Neither arrangement is universally correct. The correct arrangement is the one the receiving system and operator expect.

If the editor places narration on one channel and background sound on another, but playout uses only one of them, part of the story can disappear without the file being corrupt. The failure is a mismatch between the delivery layout and the receiving configuration. Writing “stereo” in a message does not explain what each channel actually contains.

Do not solve an unfamiliar channel layout by randomly duplicating tracks or moving everything to the centre. Establish the station’s required delivery specification and check the file against it. Technical requirements such as channel mapping, sample rate, format and loudness should come from the receiving workflow, not a generic internet preset.

Version control creates another trap. The editor fixes the missing opening words and exports again. The producer now hears the corrected file on a shared drive. But the playout operator already imported the previous version ten minutes earlier.

Replacing a file in one location does not guarantee that every downstream system updates its own copy. Some systems ingest or cache media. Others reference a path. Their behaviour differs, so the team needs to know which asset is actually loaded instead of assuming that the latest shared-drive file has reached the rundown.

A useful delivery name identifies the story and an agreed version clearly. The exact naming convention matters less than consistency and the ability to distinguish the approved revision from an earlier one. Names such as “final-new-final2” depend on memory at precisely the moment the newsroom has the least spare attention.

When a revision is delivered, state what changed and which version replaces which. The person responsible for the rundown should confirm the intended asset has been selected. That is a short operational handover, not unnecessary paperwork. It prevents two competent people from approving two different files.

The next check belongs in the actual playout environment. A file that plays correctly in a desktop media player still needs to arrive correctly at the production console or software mixer. Input selection, channel mapping, faders, mutes and bus assignments can change what reaches programme output.

This is where cue and programme listening must be distinguished. An operator may hear a clip through a preview or pre-fade listening path while its channel remains closed to the programme bus. That preview is useful, but it does not prove the audience will receive the signal.

Arrange a controlled check before airtime, using the station’s approved rehearsal, preview or isolated test procedure. Never send an unplanned test to a live output. The goal is to confirm the relevant path without interrupting a broadcast or accidentally transmitting material early.

A remote listener can help with an authorised end-to-end test because the control room’s local monitoring may bypass part of the distribution chain. For a stream, a destination return can reveal a problem that local meters do not show. Keep that return out of the programme mix so the check does not create an echo or feedback loop of its own.

Try this single handover exercise with a short, non-sensitive rehearsal interview. Export it using your normal delivery settings, give it an unambiguous test-version name, and pass it through the same import and playback process used for a bulletin. Have the editor and playout operator identify the exact file they each approved.

At each handover, compare what is heard with what should be present: the first answer, narration, intended background sound and the ending. Confirm the channel layout as well as the listening result. Keep the test short enough to repeat, but realistic enough to include the steps your team normally rushes.

If the voice disappears, work backwards to the last point where it was definitely present. If it is missing from the export, investigate the editing and export path. If the export is correct but the imported asset is not, investigate version selection and ingest. If preview is correct but programme is not, investigate the output routing. This narrows the fault instead of sending everyone back to the microphone.

Headphones and loudspeakers have a role here, but neither replaces the workflow. Use an appropriate monitoring level and a listening setup that makes speech defects easy to notice. A large, exciting sound is less useful than a clear one when the job is detecting a missing first syllable, an unexpected mute or a change in background noise.

Assign ownership to the final check. When the editor assumes playout has checked, and playout assumes the producer has checked, an important task can belong to nobody. One named person should know which delivered asset was approved and which listening point that approval covered.

At Zamar Audio Solutions, the equipment conversation starts with that signal path: what is being captured, how it is monitored and which connections the team must rely on under deadline pressure. Our podcast and creator equipment collection is a starting point for the capture and monitoring tools within that workflow, not a substitute for checking the delivered file.

The dependable upgrade is the one that closes an identified gap. Before replacing a microphone because a bulletin lost its interview, establish where the voice disappeared. Explore capture and monitoring equipment for a controlled spoken-word setup.