The lecturer paused on slide 14, said something that wasn't on the slide, and moved on. You weren't typing fast enough. Three weeks later you're studying for the exam and that one offhand comment is the answer to a question you can't reconstruct.

This article exists because that scenario is universal. It happens at UNSW, at Melbourne, at ANU, at every campus in the country. The honest fix is to record the lecture and turn it into searchable notes after the fact. The catch — and there's always a catch — is that recording lectures sits in a slightly awkward intersection of Copyright Act 1968 rules and campus-specific policies.

Here's what's actually legal, what's smart, and the workflow that turns a 90-minute recording into study notes in under five minutes.

The legal layer: Copyright Act 1968 and what it covers

A lecture is a "literary work" and a "dramatic work" under s 10 of the Copyright Act 1968 (Cth). The lecturer owns copyright in the words they speak and the slides they show. That copyright is automatically vested — there's no registration required.

Recording a lecture without permission is technically making a copy of a copyright work. That's the bit students often skim over: even if you have permission to attend, you don't automatically have permission to record.

What rescues most use cases is s 200AB — the "flexible dealing" exception. It permits reproduction for the educational instruction of an enrolled student, provided:

The use is for the purpose of education
The use is non-commercial
The use is not contrary to a normal exploitation of the work
The use does not unreasonably prejudice the legitimate interests of the copyright owner

In practice, this means a student recording a lecture for their own study is almost always fine under federal copyright law. Sharing the recording publicly, selling notes derived from it, or uploading it to a third-party site is not.

s 200AB is supplemented by s 113E (audio recording for private use) and the broader "fair dealing for research or study" exception in s 40. For most domestic students at most universities, you have three overlapping legal bases for personal-study recording.

The campus layer: university policies that go further

Federal law is the floor. Universities can — and do — set stricter policies as a condition of enrolment. Breaching them isn't usually a legal issue, but it can be an academic-misconduct one.

We checked the published policies of the Group of Eight as of November 2025:

University of Melbourne — Recording of Lectures and Other Teaching Activities Policy (2023). Permits personal recording for study with academic-instructor consent. The university's official Lecture Capture system (Echo360) is the recommended primary channel.

University of Sydney — Recording for personal study is allowed without prior approval, provided the recording is not redistributed. Group settings (tutorials, labs) require all participants to be informed.

UNSW Sydney — Lecture Recordings Policy permits personal recording where the lecturer has not objected. Lecturers can opt out by stating so at the start of the session.

ANU — Permits personal-use recording. Redistribution requires written permission.

Monash — Similar to UNSW; permits recording unless the lecturer objects, and prohibits redistribution.

UWA — Recording is permitted for students with a formal learning access plan; general students must request permission from the lecturer.

Adelaide — Permitted for personal use; redistribution is grounds for academic misconduct.

UQ — Recording is permitted for personal study, and the university provides official Echo360 recordings for most large lectures.

Common rule across all eight: tutorials and small-group teaching are stricter than lectures. The reason is that tutorials include other students' contributions, and other students have their own privacy interests. If you're going to record a tutorial, tell everyone in the room at the start.

The smart layer: when to record and when not to

Recording everything isn't the right move. Recording the right things is.

Always worth recording:

Lectures that go off-slide (where the prof says things not in the deck)
Lectures where you're sick, tired, or distracted
The week before midterms or finals
Any guest lecture
Anything in a subject you know you'll re-take or use later

Not worth recording:

Pure slide-reading lectures (just download the slides)
Tutorials with sensitive personal contributions
Anything explicitly marked "do not record" by the instructor

The distinction matters because audio takes space, attention, and (if you transcribe it) money. The Free tier on speechtotext.au gives you 100 minutes of transcription a month, which is roughly two long lectures. Pro at $19/month gives you 600 minutes, which covers a normal semester load. Recording everything when you'll only re-listen to 20% is wasted effort.

The workflow: 4 minutes from lecture to study notes

Here's the practical bit. Total active time after the lecture: ~4 minutes. Total elapsed time including the transcription job: ~6 minutes.

Step 1 — Record on your phone

iPhone Voice Memos is fine. Android: the built-in Recorder app on Pixels is excellent; Samsung Voice Recorder is fine; otherwise grab Easy Voice Recorder from Play Store.

A few small things make a big difference:

Place the phone face-up on the desk, not in your pocket. Mic faces the ceiling, not your jeans.
Sit closer than usual. Speech-to-text accuracy drops off sharply when the lecturer is a faint voice in a large room. The front quarter of the lecture theatre gives you 2–3× cleaner audio.
Record in 16-bit 44.1 kHz mono if you can (most apps do this by default). Don't crank it to 24-bit stereo; you're not making a podcast.
Title the file immediately after the lecture with subject code + date, e.g. COMP3331_W04_lecture_18Nov.m4a. You will not remember which file was which in three weeks.

Step 2 — Upload to speechtotext.au

Drop the file at [speechtotext.au](/). For most lectures (60–90 minutes, 30–80 MB after the iPhone's audio compression), the transcription returns in 90–180 seconds.

A 90-minute lecture is too long for the Free tier's 15-minute file limit, so this is where Pro pays for itself. Pro accepts files up to 3 hours.

Step 3 — Get the AI summary (Pro only)

The summary is the part that makes lectures actually useful for studying. Whisper gives you the raw transcript — usually 8,000 to 12,000 words for a 90-minute lecture, which is too much to re-read.

The AI summary distils that into:

Key concepts (5–8 bullet points)
Definitions introduced this session
Action items (assignments, readings, due dates)
Exam-worthy quotes (statements the lecturer flagged as important)
Open questions you raised but didn't fully resolve

We tuned the summarisation prompt specifically for lecture content. It's not the generic "summarise this transcript" output — it understands that lecture transcripts have a different structure than meeting transcripts.

Step 4 — Export and file

Export the summary as PDF, drop it into your subject folder in Notion / Obsidian / Apple Notes / whatever you use. The raw transcript stays in your speechtotext.au history if you ever want to grep for a specific phrase.

If you use OneNote or Notion, you can paste the summary directly. The Markdown export preserves the headings so the structure carries through.

The accessibility angle

Students registered with their university's accessibility / disability service have stronger recording rights. Section 113F of the Copyright Act explicitly permits assistive technology use, and university policies that restrict recording usually have an explicit carve-out for students with a learning access plan.

If you're navigating ADHD, dyslexia, hearing loss, or chronic illness that affects attendance, register with your university's accessibility office at the start of semester. The plan they put together typically:

Authorises recording without further per-class permission
Allows lecture-capture access for tutorials that aren't normally recorded
May provide a subsidy for tools like speechtotext.au (we've seen this at Melbourne, ANU, and UQ — ask your access advisor)

Even outside the accessibility framework, transcribed lectures are a quiet game-changer for international students whose first language isn't English. Re-reading is much easier than re-listening when you're still building English vocabulary.

What about the official Echo360 recordings?

Most lectures at most Australian unis are recorded by the university itself through Echo360. The recordings are usually available in the LMS within a few hours.

So why record yourself? Three reasons:

Echo360 doesn't cover everything. Smaller lectures, guest lectures, and any session with technical issues often miss the official recording.
Echo360 doesn't transcribe. Some unis have started rolling out AI captions, but they're typically slow, partial, and not searchable.
You can't take the Echo360 file with you. It stays in the LMS, where it expires when you finish the subject. Your own recording is yours forever.

The right pattern is to use Echo360 as your fallback and your own recording as the source you actually work with.

Privacy: where your audio goes

This is the question we get most from students who care about their data. The short version:

Audio is uploaded to speechtotext.au servers
The audio is processed in memory and discarded immediately after transcription
It's not written to disk on our servers, and it's never used for training
The transcript itself is saved to your account so you can re-open it later; you can delete any transcript permanently at any time
We're an Australian company subject to the Australian Privacy Principles

If you want to see the full posture, the privacy page lays it out in plain English. If you want it gone after the fact, your account settings has a one-click "delete all" for transcripts.

Cost breakdown for a semester

A full-time undergraduate carries roughly 12 hours of contact per week. Of that, maybe 6–8 hours is lectures worth recording. Over a 12-week semester that's 72–96 hours of audio.

Free tier (100 min/month): enough for 2–3 lectures a month, plus the odd short clip. Works for one core subject you're struggling with.
Pro tier ($19/month, 600 min/month): covers all your lectures plus tutorials worth recording. The summarisation is the bit that's hard to live without once you've had it.

We made student pricing intentionally generous because the unit economics make sense — a student who graduates as a Pro user usually stays a Pro user when they hit the workforce. If $19/month is tight, [email us](mailto:hello@speechtotext.au) and we'll work something out.

FAQs

Is it legal to record a lecture in Australia?

For personal study, almost always yes — covered by s 200AB and s 40 of the Copyright Act 1968 (Cth). Universities can set stricter policies; the policies of the Group of Eight all permit personal-use recording. Redistributing or selling the recording is not permitted.

Can I record a tutorial?

Tutorials are stricter because other students' contributions are involved. Tell everyone in the room at the start, get verbal consent, and don't redistribute. Some universities require explicit opt-in from all participants.

Will AI transcription understand my lecturer's accent?

Whisper-large-v3 (the model speechtotext.au uses) handles Australian, Indian, British, American, and most European accents well. We published a benchmark on Australian accents that compares Whisper to Google and Azure. Strong non-English accents are sometimes harder; if your lecturer has an accent the model struggles with, the AI summary will still capture the main concepts even if individual sentences are imperfect.

Can I get a refund if I upload an unlistenable recording?

Yes — email us within 24 hours of the upload and we'll credit the minutes back. The free tier is meant to be the testing ground, so we encourage students to drop a 5-minute sample before committing to Pro.

What's the longest file I can upload?

Free: 15 minutes. Pro: 180 minutes (3 hours). Business: 240 minutes. Most undergrad lectures fit comfortably in Pro.

Does the transcript include speaker labels?

No — Whisper produces a single column of text. For lectures (one speaker) that's fine. For panel discussions or seminars with multiple speakers, you'll need to add speaker context manually.

---

Try a five-minute sample free at [speechtotext.au](/) — no sign-up required. When you're ready for full lectures, Pro is $19/month with the first month refundable if it doesn't work for your setup.