When you can't find listening material at the right level
Every listening library is a fixed shelf. You can usually find the right topic or the right level — rarely both, and almost never both on the week you need them.
7 min readYou have three options: change the task instead of the audio, adapt the audio you have, or make audio that fits. The first is free and works more often than teachers expect. The third used to mean recording yourself in an empty classroom — now it doesn't.
The problem is that libraries are shelves
The free listening corpora are genuinely good. They are also finite and fixed. If your B1 class is working on volcanoes this week, you can search every one of them and find nothing at B1 about volcanoes — because the shelf was stocked years ago by someone who did not know your syllabus.
So teachers do one of two things. They use audio that is on-topic and too hard, and watch half the class disengage in the first thirty seconds. Or they use audio at the right level about something unrelated, and lose the thread that made the lesson coherent.
A teacher who needs B1 audio about photosynthesis has nothing. Not because the material is bad, but because nobody recorded it.
First, try grading the task instead of the text
There is a long-running argument in ELT about authentic versus graded listening — whether learners should meet real speech at natural speed, or simplified audio built for their level. The practical resolution most experienced teachers land on is that you keep the authentic audio and change what you ask students to do with it.
- Lower the task, not the audio. Same recording, but the A2 group listens for three specific numbers while the B1 group summarises the argument.
- Shorten the span. Play forty seconds rather than four minutes. Difficulty is partly a stamina problem.
- Give the topic away first. Pre-teaching four words and naming the subject removes most of the panic without removing the challenge.
- Let them hear it more than once. The third listen is where lower-level students actually start processing rather than bracing.
This handles a surprising share of cases and costs nothing. It stops working when the audio is so far above the class that no task rescues it, or when the topic itself is wrong.
When nothing fits, make the audio
Recording yourself works and teachers have done it for decades. The reasons it doesn't scale are familiar: one voice is not a conversation, you need a quiet room, and re-recording after a mistake means starting the take again.
Dily's Podcast Studio makes the audio for you. You describe the topic and the level you need, choose two voices, and get a two-character conversation with comprehension questions ready to assign — or to print, if the lesson is going to run offline.
The useful part is that nothing is fixed until you say so. The conversation arrives as a draft you can read and change before anyone hears it, so the vocabulary matches what you actually taught last week rather than what a generic script assumes.
So, what actually makes a volcano erupt?
Pressure from melted rock deep underground.
Why two voices matters more than it sounds
Single-voice narration teaches students to follow a monologue. Almost no listening they will actually do is a monologue. A two-character dialogue brings in the things that make real speech hard and are absent from a read-aloud passage: turn-taking, one speaker interrupting the other's point, a question answered indirectly.
It also makes the comprehension questions better, because there is something to ask about beyond fact retrieval — who thought what, and where they disagreed.
Bring Dily to your classrooms
Tell us about your school and we'll set up a demo tailored to how your teachers work. Pilot programs available.