You have left videos running without subtitles for months and still catch little. Turn subtitles on and there is barely a word you do not know.

That situation is not a vocabulary problem. The sounds you memorized and the sounds actually being produced are different things.

People do not pronounce words one at a time

We learn words individually and clearly. In real speech, the boundaries between words disappear.

Ending sounds attach to beginning sounds, weak sounds vanish entirely, and unstressed vowels blur. Three words you know, spoken together, become a chunk of sound you have never actually heard.

This is not sloppy pronunciation. It is how the language is spoken — the same way you do not pronounce your own native language one syllable at a time.

The problem is chunks, not speed

When you cannot follow, it feels like everything is too fast. Usually the bigger issue is unfamiliarity with the chunks rather than the tempo.

Drop playback to 0.75x and you can test this. If a passage is still unintelligible slowed down, speed was not the problem — you do not know that sound chunk. If slowing it down makes it clear, it is an adaptation issue that repetition solves.

Making that distinction tells you what to practice.

Why passive listening does so little

Leaving audio playing in the background rarely builds comprehension, because the brain lets it pass without processing it.

It does build familiarity. But familiarity and understanding are different, which is why months of it can produce no change.

What you need is the act of asking yourself what was just said, and then checking. Without that check, repetition only spends time.

The sequence that works

Cut it short. Pick one 30-second to one-minute passage. Cycling a short section several times beats playing a long video once.

Listen without subtitles and transcribe. Leave blanks where you cannot hear. Those blanks are precisely the sounds you miss. Vague 'I can't hear it' becomes a specific list.

Turn subtitles on and compare. Check what was in the blanks. Most will be words you know. Seeing that with your own eyes — known word, unheard sound — is the point.

Say the passage aloud, exactly as spoken. Copy the way the speaker linked and dropped sounds. Once you can produce it that way, you start hearing it. What your mouth knows, your ear catches.

Listen again without subtitles. Check whether the blanks are now audible.

How much

One passage a day, about ten minutes, is enough. Precisely dissecting a short section beats an hour of passive listening.

Repeating the same passage across several days is fine — better, actually, since that is how sound chunks become automatic.

The takeaway

Listening usually stalls not because of vocabulary or speed, but because of unfamiliarity with how sounds link and disappear in real speech.

Cut it short, transcribe, compare, imitate, listen again. Converting passive listening into that sequence is what starts moving a passage that has been stuck for months.