How to choose the right bit depth for an oral history cassette

Bit depth determines how finely a digital copy represents small changes in an analogue tape signal. For an oral history cassette, it is a precision setting rather than a measure of how important the voice sounds. A careful transfer can succeed at either 16-bit or 24-bit, but the best choice depends on the tape, deck, background noise and intended CD master. Matching those details gives you a faithful recording without mistaking extra data for extra fidelity.

Many Australian families still uncover labelled tapes in bedroom drawers, op-shop purchases and boxes bought at car-boot markets. Their recordings may be fragile, have spent years in hot garages or contain tape hiss that was barely noticeable when first played. The aim is to capture the surviving signal honestly, document damage and retain an archival master for future checking, rather than polish every flaw out of the story.

What bit depth actually records

Each bit adds another step to the scale used to describe a waveform. A 16-bit recording has 65,536 possible amplitude values per sample, while a 24-bit recording has 16,777,216. The extra resolution reveals quieter changes against a noisy background and provides more room for restoration work. It cannot reveal detail that the tape, playback electronics or microphone never captured.

That distinction matters because a cassette is an analogue medium. Its useful dynamic range, frequency response and noise floor are set by the tape formulation, recording level, head alignment and machine condition. Consumer cassette decks often deliver a signal-to-noise ratio far below what a 24-bit converter can theoretically represent. Recording at 24-bit therefore captures more of the noise floor and tiny irregularities, not necessarily more of the speaker’s voice.

Bit depth is also separate from sample rate. Sample rate describes how many measurements are taken each second; bit depth describes the resolution of each measurement. For cassette material, 44.1 kHz is normally sufficient because the format cannot deliver the full frequency range of an open-reel tape or a modern digital recording. A 48 kHz working rate is useful when the recording chain or restoration software prefers it, but it does not need to become a search for the highest number available.

Match the capture format to the cassette

Start by identifying what you actually own. The cassette tape formats guide distinguishes full-size, compact and microcassette shells, helping you match the recording to compatible transport and playback equipment. Microcassettes need a suitable machine: they may fit inside a larger shell for storage, but they will not play in a standard compact-cassette deck.

Next, decide whether the cassette carries one conversation across both stereo tracks, separate recordings on each side, or a mono recording made on a stereo deck. Preserve the original track layout when possible, as channel assignments can help future archivists understand the tape. Check the intended playback speed too: changing playback speed changes both pitch and timing, so correction belongs in a documented copy rather than in the untouched capture.

Choose 24-bit when low-level passages may need restoration headroom, when you want one capture to support several future treatments, or when your archive has a standard demanding 24-bit preservation files. Choose 16-bit for a straightforward, good-quality cassette-to-CD transfer where the goal is an efficient working set and the final disc will already be 16-bit. Extra bits are most valuable as working space, not as a badge of technical seriousness.

Set levels for voices, hiss and tape damage

Levels matter more than a large number on a converter specification. Set playback gain so speech sits clearly below clipping while quiet words remain audible. Watch a real-peak meter through several passages, including the loudest laugh or cough, rather than relying on an optimistic software display. Keep the gain steady between voices; pumping between syllables can make an interview tiring to hear.

For a 24-bit capture, aim for a healthy average around -20 dBFS and reserve genuine headroom above the loudest peak, commonly below -12 dBFS. With a 16-bit target, leave roughly 3 dB below full scale. These are starting points, not laws, and the aim is to avoid clipping while avoiding the opposite mistake of boosting tape hiss until it dominates the recording.

Record a short section with noise reduction and automatic level adjustment disabled, then listen carefully before processing it. Mild hiss may be preferable to aggressive filtering that makes the voice sound underwater. Dropouts, warble and loss of high frequencies should be documented against a timecode or counter position. Try to capture any severe damage once, since repeated playback and winding can worsen fragile or dried-out tape.

Prepare the deck and digital audio path

Prepare the transport before connecting it to the computer. Clean the record and playback heads, guides and pinch roller with suitable materials, and check that the tape path is free of sticky debris. A dirty head can reduce treble or produce repeated clicks, making a poor source look like a digital problem. Replace a hardened pinch roller or service a slipping capstan rather than compensating with extreme treble in software.

Use the deck’s line or tape output into a suitable line input; avoid a microphone preamp unless you are deliberately recording sound from the room. Balanced connections can reduce interference in long cable runs, while a headphone jack is usually not a convenient stereo recording output. If a cheap lead introduces hum or crackle, test it separately. For old equipment, check electrical safety and use a compliant 230 V supply or a certified step-down converter with an Australian plug for imported units.

In Melbourne, council e-waste facilities divert suitable electronics from landfill, although accepted items vary. A working deck may be more useful to a community radio group or repair workshop. For a homemade signal indicator, testing a diode offers a useful way to understand a simple LED check, although a glowing bulb cannot establish the converter’s true resolution. The audio interface specification matters more: confirm that its analogue-to-digital converter supports the selected depth and sample rate, and that the driver exposes both settings. File names ending in “24-bit” prove nothing if the hardware or converter has quietly reduced the depth.

Practical recommendations for archival transfers

For the do-it-yourself approach used by Allen’s Site, a modest home workflow is often enough for an oral history, provided the stages stay consistent. Label the cassette, write down its playback speed and condition, and create a fresh folder for each source. Keep the first capture unprocessed so that a later transfer does not inherit an irreversible gain change, noise-reduction artefact or editorial decision.

These choices cover most cassette transfers without requiring expensive studio gear:

• Capture at 24-bit for working masters. Use 44.1 or 48 kHz, keep peaks below clipping and retain the raw take for later restoration.
• Create a 16-bit CD-ready file when appropriate. Dither and resample deliberately to 44.1 kHz rather than letting a conversion tool make an undocumented decision.
• Preserve original channel and timing information. Keep stereo tracks intact unless a documented restoration task calls for a change.
• Monitor continuously. Check the first minute, quiet passages and loud moments, then save the software version and key settings with the session notes.
• Store the tape properly. Keep it away from heat, direct sunlight and humidity, and use archival enclosures for irreplaceable recordings.
• Document every stage. Record the deck, cable, interface, playback speed, processing and consent details so another person can repeat the transfer.

Preserve the recording and share it ethically

Copyright and consent deserve as much attention as the electronics. Under the Australian Copyright Act 1968, ownership of an oral history recording is separate from ownership of the physical tape. Depending on the interview, copyright may rest with the speaker, the interviewer or another party, so written consent and commissioning terms matter. If the interview forms part of a government or institutional collection, check the applicable state records legislation and deposit rules before publishing.

Some tapes contain introductions, station identifiers or announcements that distract from the interview. Procedures for editing out announcements can guide a clean edit, but retain the original capture and keep a written log of what was removed. Save an archival WAV or BWF master, a checksum of the master and a CD listening copy to give the family several practical ways to enjoy and preserve the story.

In Sydney and other humid coastal cities, keep cassettes away from damp cupboards and hot roof spaces, and use a stable indoor shelf instead. Start with a ten-minute sample from the most difficult passage, compare 16-bit and 24-bit captures at normal listening volume, and check that voices and hiss sound unchanged. Then commit to a setting, make the master, write the metadata and store the cassette safely.

When the project is ready, share the CD or a carefully mastered digital version with family and community historians. Credit the speaker, name the interviewer and note any restrictions on publication or reuse.