Choose audio with a dependable text reference
Pick a short recording that provides a trustworthy transcript, or an exercise whose wording you can check. Keep using its original audio player for listening. You need to be able to return to the same segment after the explanation, because a text-only review cannot show whether you now recognize the sound.
For the worked example, imagine a short message about meeting tomorrow. Your first task is to identify the event, the time and the place. You do not need to transcribe every syllable before you can describe the main message.
| Japanese | Romaji | Meaning |
|---|---|---|
| 明日は三時に会いましょう。 | Ashita wa sanji ni aimashō. | Let’s meet at three tomorrow. |
| 駅の東口で待っています。 | Eki no higashiguchi de matte imasu. | I will be waiting at the station’s east exit. |
Write your first understanding before reading
After one listen, note what you understood and mark what is uncertain. You might know that two people are meeting but be unsure which exit they mention. Keep an uncertain detail visibly uncertain rather than replacing it with a guess that later becomes difficult to distinguish from the source.
Try another listen with one question in mind. Where will they meet? What time is mentioned? If the answer remains unclear, open only the relevant transcript line. This gives the text a specific job and helps you notice whether the obstacle is a sound, a word or a sentence relationship.
Ask Mocchi about the difficult part of the text
Type the short line and explain what you found difficult. With /explain, you can ask about the sentence structure; with a plain question, you can focus on one word or contrast. Ask the tutor to stay near your level and avoid rewriting the message into a more advanced style.
Compare the explanation with the complete source exchange. A word may have several possible meanings, and a short excerpt may omit the situation that selects one. If the explanation seems inconsistent with the recording or transcript, return to the source rather than building practice around a possibly incorrect line.
| Japanese | Romaji | Meaning |
|---|---|---|
| 東口ですか、西口ですか。 | Higashiguchi desu ka, nishiguchi desu ka. | Is it the east exit or the west exit? |
| 三時ですね。 | Sanji desu ne. | At three, right? |
Replay without following the written words
Listen again after the explanation, then close the transcript for another attempt. Pay attention to the part you previously missed. You are checking whether the sound now connects with the meaning, rather than whether you can read the sentence more easily. Keep the segment short enough to make that comparison honestly.
Next, retell the message in your own words. You can give the key details in your preferred language first, then try a simple Japanese summary. Ask Mocchi whether your written summary preserves the information in the transcript, while remembering that the original audio remains the source for what was spoken.
| Japanese | Romaji | Meaning |
|---|---|---|
| 明日、駅で会います。 | Ashita, eki de aimasu. | We will meet at the station tomorrow. |
| 東口に行きます。 | Higashiguchi ni ikimasu. | I will go to the east exit. |
Use a new task to check the meaning
Ask what you would do after receiving the message. Where would you go, and what detail would you confirm? This changes the activity from repeating a line to using its information. A suitable follow-up might ask you to confirm the exit or explain the meeting arrangement to someone else.
For voice interaction, prepare the situation in Mocchi and use /kaiwa to find an AI Sensei topic such as directions or travel. The new conversation is practice using the language, not a second transcript of the recording. Keep those two sources separate when reviewing your answer.
Let the next recording test the same weak point
Write a brief note about the obstacle: I confused two exit names, or I missed the time when the phrase was spoken continuously. Ask Mocchi for a small vocabulary or sentence task that supports that distinction, and choose another recording where you can listen for it. More text review alone may not solve a sound-recognition problem.
Use the routine selectively. You do not need to analyze every line of every recording. A focused explanation, a return to the audio and a new meaning task can provide a clear next step while keeping listening at the centre of the session.