Advanced stage: learning to understand rapid idiomatic everyday dialogues

I have been learning French (primarily using LingQ) for the last 1.5 years. I have 2.4 million words read and 34600 known words and am listening to the audio book while reading. I have also listened a lot to podcasts, radio shows and non-fiction audio books outside of LingQ (I haven’t tracked this), and still do this while doing chores, running, etc.

I am able to read novels on LingQ (including classics) and can read easier contemporary novels outside of LingQ. I can listen to radio shows (as on France Culture) and watch Youtube videos with ease. I also have conversations in French with natives and/or other French learners for a couple of hours on average each week.

I am able to hold a fine (but a bit simple) one-on-one conversation with a native for a longer period of time (e.g., 15 minutes - not just “what is your name” etc.) However, when multiple native speakers have a very rapid group conversation with more idiomatic language, I can have a hard time understanding everything and it becomes very hard to enter the conversation.

A related issue is that I am basically unable to understand TV series and movies without subtitles. They speak much less clearly than in radio shows and (again) use more idiomatic language. (On the other hand, I can easily understand factual TV broadcasts, such as the news or documentaries).

I can also feel that I lack a lot of idiomatic everyday phrases when I am in everyday situations.

What is the best (i.e., most time efficient) way to overcome these issues? I am starting to doubt whether reading novels will fix them, since a large part of a novel is description and not idiomatic conversation. However, I think that many TV series and movies contain too much non-speaking time (if I do a 30 minute language study session daily I would like it to be concentrated).

Is there maybe some source of radio plays where they speak as quickly/unclearly/idiomatically as in TV series?


This was discussed a little on another recent thread:

With regard to learning common idiomatic phrases, try putting the podcasts through LingQ and lingQing the phrases. Another source are reality shows. But you just want improvised, informal content with multiple native speakers talking over top of each other.

It just takes time. To being able to understand hard accents and low-quality audio, you just need to keep practising with it. Go for external listening or extensive reading while listening. I.e. go for quantity. You just have to get used to the sounds.


It just takes time.

@nfera – You heartless bastard! :slightly_smiling_face:


So I’m assuming that you’re a native english speaker. I’ve been studying at a similar pace to yours but for longer. a little over 3 years, 4 million + words read. Outside of reading on LingQ, I also do all of the things that you do but I currently don’t have much issues with listening or speaking. Now mind you, it’s still not perfect - there will always be words or phrases here or there that i will miss but that’s to be expected. One thing we have to learn as language learners is how to be comfortable dealing with uncertainty. So, despite the fact that english (my native language) and korean are far more removed from one another than english and french, i was able to read this level in about ~2.5 years or so.

Given that we’re both going at a similar pace and that you’re studying a much easier language (for english speakers) I suspect that you should be able to reach the point of ‘understand tv series and movies without subtitles’, and understanding ‘multiple native speakers’ in a group conversation. One thing I would invite you to do is read more challenging content, if you’re comfortable doing so and just keep on pressing on.


I have listened to 4200 listening hrs in German and have read 7 million words. Now at this point I feel like local natives are getting approachable and I am understanding more and more natives around me without straining my mind. These hours also included 500 hours of German radio, 100 audiobooks, several seasons of Television shows and radio plays. Active listening with Lingq lessons.

Based on my observation both radio plays and Television shows use pretty much the same idiomatic language you can go with either medium you will not be missing out much as far as collocations and idiomatic language go.

Try watching TV shows without subtitles. I have never used subtitles for my listening and it really improved my listening skills when it comes to understanding real natives.

In real conversations you need to understand key words to get the message. However, watching TV shows with subtitles will improve your reading skills and make you a lazy listener as a result.

Maybe watch them with subtitles then watch them again without subtitles.

Yes, quantity plays a main role. For example, when I listen to 6-8 books back to back then watch a TV show afterwards I feel that my listening skills have improved. Back to back is being important here.

It is not just about idiomatic language it is also about getting used to the intonation. There is so much variety on offer and each variety requires its own ndividual attention.


This might seem counter-intuitive to some but my suggestion for more spoken IRL language is to reach Young Adult, Teenage books. These are often more geared towards the ‘rawer’ and more ‘informal’ forms of language. Where you will probably find the prose written in a much more natural way. If you can find any audio books from those books you do end up finding quite authentic it will help you a lot. I know this is the case with Swedish and English as well to a lesser degree I’d say. But the language is not there to look pretty, just authentic. If you try to pivot to that I think you’ll cover that more. I am not talking about fantasy or anything but books that cover more day-to-day lives of youths. Since the ‘laziest’ most ‘informal’ forms of language is usually used by teenagers and youths.


Thanks for the replies everyone. I will try to find some “chit-chat” podcasts as @nfera recommends and listen to them in noisy environments without transcripts. This, however, is easier said than done - most podcasts seem to be spoken very clearly and not include natural “ping-pong” dialogues but rather let each native speaker have their turn at a monologue.

If anyone happens to know some chit-chat podcasts in French, please let me know!

I am starting to wonder whether it would be possible to download the audio from a TV series, remove all periods with no dialogue and listen to this? Has anyone experimented with this?


In Italian there are quite a few chit-chat podcasts. I’m sure there are plenty in French too. The standard interview ones are usually with one host and in these cases the host doesn’t usually talk over the guests. But if there are multiple hosts, who are friends or know each other very well, they are more likely to happily talk over top of each other, even if there’s also a guest. If you Google search on .fr in French, you’ll surely find some.

I’ve never used condensed audio from a TV series or movie myself, but other people have. Not that I’ve used it myself, as I’m studying Italian, but it’s a feature of Migaku (which is available for French):


Interesting, how does Mikagu work?
The idea I had was to use a VAD (voice activity detector), like GitHub - snakers4/silero-vad: Silero VAD: pre-trained enterprise-grade Voice Activity Detector. The neural network will output start and end timestamps of audio segments containing speech. After that you would have to cut the original audio file according to these timestamps. Not sure how to do this best. Here is my poor attempt at doing something like this:
Google Colab


@bamboozled I’ve never used Migaku, as it doesn’t support Italian, but I’ve watched some of their videos. It’s really a collection of software tools, which is kind of a mix of Language Reactor and Anki, in a sense.

With regard to the condensed audio, prior to Migaku, Yoga had designed it to use subs2srs on a video to get one audio per subtitle line, then concatenate them with MP3Cat. I imagine the design of this feature in Migaku does the same.

EDIT: They might do it a little bit different for performance reasons now. Previously it took a video file, but really it’s not necessary. They just need to download the .srt and the audio, split it, then concatenate it.


I’m not as advanced, but I am having a similar issue to you and actively started really trying to change a lot of what I’m focused on to dialogue based input.

Aside from other’s suggestions, which are great:

Do you ever listen/watch “easy french” youtube channel? I’m personally only familiar with the German and Spanish, but in these (if you’re not familiar), they talk with/interview people on the streets about a certain topic in their daily lives. Frankly, I think the “easy” part is a bit of a misnomer, because here you do get authentic, native speech. THe people they interview are not trying to slow down or change their speech for a non native. It is “easier” in the fact that a subject is stated clearly beforehand, and the questions are presented in a rather easy form usually, but the answers have all the quickness, stopping/starting, changing thoughts midstream, abbreviations, colloquialisms that you would come across with natives.

It looks like they have an “easy french” podcast as well. Unfortunately only 50 something episodes. However, I’ve been devouring the Easy German podcast as it is very conversational and includes a lot of collocations, colloquialisms. etc. I suspect the Easy French one would too.