Can you record a new voice memo while an older one is still transcribing?

It comes up in an ordinary way: a transcript is requested on yesterday's interview, it is still working through the audio, and something worth capturing starts happening right now — a follow-up thought, a second conversation, a new lecture starting. Hitting record in that moment feels like it might step on the transcript already in progress, the same way opening a second recording app steps on a first one. It does not, and the reason comes down to what each of the two operations is actually asking the phone for.

Two operations, two different resources

A recording in progress needs the microphone and very little else. Capturing audio is a light, steady task: a stream of sound comes in, gets encoded to AAC, and gets written to disk, over and over, second after second. It barely troubles the processor.

Transcribing an existing recording needs something completely different: no microphone at all. It hands audio that was already captured — sitting in a file, not coming live off any input — to Apple's Speech framework, which runs a recognition model against it using the processor and the neural engine. That is a heavier, bursty task compared to recording, but it is not fighting over the same thing recording needs, because it was never asking for the microphone in the first place.

Why the interruption mechanism does not apply here

The reason a phone call, an alarm, or a second recording app can cut off a recording is specific: all three are requesting exclusive control of the live audio input, and iOS only hands that to one thing at a time. The moment something else asks for the microphone, the system takes it from whoever had it and tells that app an interruption has begun.

Transcription of a file that already exists never makes that request. It is not live audio, so there is nothing for it to ask iOS to hand over, and nothing for a new recording to take away from it either. The two operations can be running on the same phone at the same moment without either one triggering the notification that a call or a second recorder would.

What can actually change when both run at once

The microphone and the processor are not the same kind of shared resource, but the processor and neural engine are shared between whatever is using them, and transcription is the one of these two tasks that leans on them. Recording alongside a transcript already in progress means that transcript is now sharing the processor with something else, on top of whatever else the phone is doing — so it is reasonable to expect it can take a little longer to land than it would with the processor free, the same way a transcription can run slower under Low Power Mode's lowered performance ceiling.

What does not change is the recording itself. Capturing audio was never the demanding half of this pair, so a new recording keeps writing to its own file at the pace it always does, regardless of whatever a transcript elsewhere in the app is doing with the processor at the same moment.

The recording and the transcript stay separate

Nothing links the file being recorded now to the file being transcribed already. They are different recordings, sitting in different places in storage, and starting one does not touch, pause, or restart the other. When the new recording is finished and you ask it to transcribe, that request starts its own attempt — on-device first, the same as every other recording — independent of whatever happened with the one that was already mid-transcript when you started.

Where this is different from a second recording app

It is worth being precise about why this feels similar to opening a second recorder and is not: both involve two things apparently happening on the phone at once. Opening a second app that starts recording is a genuine conflict, because both apps are asking iOS for the same microphone, and only one can hold it. A transcription running in the background is not a second recorder — it never asks for the microphone at all, so there is no shared resource for a new recording to collide with.

How Voice Studio handles this

Recording in Voice Studio is the microphone, an AAC encoder and a continuous write to disk, at whichever quality tier is selected — nothing about starting one checks whether a transcript is running elsewhere in the app, because nothing about it needs to. Transcription always attempts the on-device path first for whichever recording it was requested on, and falls back to Apple's server only if that pass comes back empty or errors, entirely independent of anything else happening in the app at the time. If a transcript on an older recording does take a little longer while a new one is being captured, that is the processor genuinely doing two things rather than either operation failing — the new recording is unaffected either way.

Common questions

Will starting a new recording cancel a transcript that is already running?

No. They are different files, and transcription never touches the microphone a new recording needs, so starting one does not pause, cancel or restart the other.

Does a transcript running in the background slow down the recording currently in progress?

No. Capturing audio barely uses the processor, so a transcript running on a different recording at the same time does not change how the new recording is written.

Does recording something new slow down a transcript that is already running?

Possibly, a little. Both draw on the same processor and neural engine, and transcription is the heavier of the two tasks, so it may take slightly longer to finish if the phone is also capturing new audio at the same time.

Is this the same mechanism as a phone call or a second recording app interrupting a recording?

No. That mechanism is about which app holds the live microphone input. Transcribing a file that already exists never asks for the microphone, so it has nothing to hand over and nothing for a new recording to take away.