Hacker News Reader

Comment by BoppreH | Hacker News Reader

Another enormous gripe I have with modern movies, other than the (lack of) interesting moral lessons from OP, is the sound design. Loud explosions and music, with quiet, mumbled lines[1]. I'm forced to watch every movie with subtitles on, or to miss a good chunk of the dialogue.

And I'd go beyond that, and blame the mumbled lines for our lack of interesting quotes. It's a lot harder to have unexpected turns of phrase, speech patterns, or novel vocabulary choices when the audience is struggling to understand even the most basic, predictable sentences.

[1] https://www.slashfilm.com/673162/heres-why-movie-dialogue-ha...

Edit: for people mentioning audio mixing and stereo, that's only part of the problem. Even with my (admittedly cheap) 5.1 setup, I sometimes miss most of the dialogue in some scenes. I think the change from theatrical to more realistic enunciation is also to blame.

View on HN

A response to your edit, I was on a zoom call with current musical editors / producers and that is 100% part of the problem. Many of today's actors /actresses almost mumble their lines and make it absolute hell for the audio people to pick them up, whereas theatre trained individuals were loud and enunciate the crap out of everything.

They are trying to be authentic. That is what actors are taught these days, to portray "realistic" characters that move and sound as realworld people. The net result is mumbling and tiny little movements that in turn mandate tight closeups and asmr-like microphone placement. Try that on a stage or in any live performance and you will be called wooden. Compare actors like Samuel L Jackson, Percival Ulysses, Jane Lynch or Rowand Atkinson. They dont need closeups and microphones secreted in hairlines. But they also rarely get leading drama roles, more often appearing as side characters who run on stage to tell the team the dramatic news.

One big problem is that US actors have a history of recording adverts, so they are not classically trained, unlike a lot of British actors who have backgrounds on the stage and so have learned to speak clearly.

I hate having to watch movies and TV with the subtitles on — the mumbling may be realistic, but I'm paying for escapism.

You probably hate clips like this. Male characters are dramatic, with big movements and loud voices to portray fear and disorder. The three female characters, including a very tough marine, are quiet and docile to engender escapist feelings of worry and protection. So which volume setting do you use? The camera is forced into an awkward zoom to reconcile the two.

https://youtu.be/VrVZHxH2O1I

"They mostly come at night. Mostly."

I'd say the editors did a good job in this case, considering both "they mostly come at night, mostly." and "game over man, game over!" have both become iconic often-quoted lines.

Good memes rarely come from good writing.

"If I took off the mask, would you die?"

"It would be extremely painful"

"You're a big guy!"

"For you."

hmm, you're probably right.

That is the incredible acting, not the lines IMHO. Imagine Bane played by an average actor - what a corny, laughable display it might have been.

I like it better this way. Theatre like playing in movie come across as fake to me - as bad acting.

When the whole thing is clearly exaggerate comedy, then it don't bother me. But, if I am supposed to immerse myself in the story, it does not work.

Real people aren’t quiet around intense emotions and loud noises. This often comes off really silly, when actual people would be yelling they come off as mumbling to themselves.

Instead, actors are often chatting in front of green screens without any of the appropriate ambient sounds or emotional context.

Real people are often quiet around those. That is as normal reaction as loud yelling.

It’s so subconscious for people to speak lauder in a noisy environment that you might not realize it, but it’s so necessary we have started to program devices to do the same thing: https://www.theverge.com/2021/8/31/22651304/amazon-alexa-ada...

I am not saying intense emotions universally result in yelling, but people very rarely whisper when say calling 9/11. Even just kids playing tag get lauder.

People normally get mute in stressful situation. They normally have troubles to express things, they talk in weak voice, whisper or not at all. They have one word answers to complex questions. And anything in between. Some dissociate and act normal or follow normal script by routine.

They don't whisper to the phone, they will put more conscious effort to talk into 911 call ... or the shaken mute person won't be the one making call.

You seriously never had to ask people to talk louder in noise environment? The "people yell" assumption is artifact of movies, not reality.

Asking someone to speak lauder in a noisy environment is normally because they aren’t raising their voice enough rather than them failing to raise their voice at all. It’s not easy to judge how laud you need to be but the basic feedback of failing to hear your own voice if you don’t speak up prompts raising your own voice.

As to going mute, some people do completely shut down in an stressful situation, but that’s associated with for more than their voice. I have no issue with an actor in a war movie endlessly stacking ammo from point A to B. But if their having a coherent conversation, activity and productively responding to stimulation, that’s very different.

I'm not a native speaker, but, I think it's rather the opposite. It's very inauthentic.

English is a rather unusual language that the meaning is mostly carried only by consonants, while vowels are almost meaningless.

What actors seem to be doing is that they focus too much on their accent, and the vowels that define it, and mostly ignore the consonants. Which means you can only hear the accents, but not what is being said.

> English is a rather unusual language that the meaning is mostly carried only by consonants, while vowels are almost meaningless.

Sorry but this sounds like nonsense. Vowel distinctions absolutely matter in English. Think of how many words would be indistinguishable otherwise: bout, bought, bet, bat, bit, beet, boot, boat, bite, but, and bait are all distinguished from each other only by a vowel.

(And, yes, these all sound quite different to me, an American, though non-native speakers often have trouble making or recognizing some of the distinctions. Some native speakers further distinguish “bot” from “bought”, but I don’t.)

It may sound like nonsense, but in comparison to many other languages, English is super flexible with how vowels are pronounced and toned (entoned?). If you've ever tried to learn a tonal language, or one with more specific vowels like Khmer (which has 33 consonants and 22 vowels) you'd realize how relaxed English can be.

Generally this is in English's favor, I pronounce button differently than my NZ friends, but they still understand me.

American English has something like 15 distinct vowels. That’s not that much less than your citation of Khmer, and way more than many languages. Any claim that American English has a uniquely poor vowel inventory is just wrong.

Perhaps I didn't specify that perfectly. I wasn't saying English is poor in vowels, I'm saying that it is flexible in how you use them. Khmer has 22 specific vowel letters, each letter that indicates exactly how the sound is made, even the differences that of how the previous letter will effect the sound.

English has 5, and no real consensus on how they are used, let alone the tone of voice that you need to use to indicate a specific word.

In tonal languages a rising vowel note is different than a falling, rising and then falling, falling then rising, flat or tumbling tone. The closest English has is the rising tone one makes when asking a question, but that doesn't change the meaning of the words entirely, just the context.

“Letters” are irrelevant. We are talking about the spoken language, not the writing system. American English has 15 or 16 vowels, and (not counting tone) the distinctions between them are just as meaningful as in Khmer or any other language. “Bit” and “beet”, or “but” and “boot” are different words, after all. Where are you seeing “flexibility” in the use of vowels?

Again, we are not talking about “letters” or the writing system here. Yes, there are many distinct vowels that are written the same, and conversely many different ways of writing the same vowel, but that’s irrelevant to the discussion of spoken language.

And, yes, I know what a tonal language is, and you’re right that English isn’t one. It’s hardly unique or special in that. Most languages aren’t tonal.

There are tons of examples, take "no" for example, you can pronounce it 'nooh', 'nah', 'nuh', 'neh', and yet it will be transcribed as "no". That one could be dismissed as negating is a pretty primitive concept.

My previous example of how to pronounce 'button' varies depending upon where you're from...but is still recognized by the most English speakers. Some folks pronounce it but-ton, others but'n.

Water can be pronounce in a variety of ways. Actually the list is endless...I'm surprised you're arguing the opposite. Compared to Spanish, or any number of languages where the vowels are very specific (disregarding the Castilian ascent), English is very forgiving. Bringing up Khmer's letters was to show that the vowel sounds have been formalized in that language, where English allows for much more variance.

Listen to any non-native English speaker, they can be all over the place and still understood.

> Some folks pronounce it but-ton, others but'n.

This is a distinction in consonants! Different realizations of the /t/ phoneme. The whole claim that started this thread was that English expresses information almost exclusively in consonants, not vowels. Your example directly contradicts that very claim.

Anyway, English is _far_ from the only language where different dialects have a different set of vowels. Quebec French for example has completely different ones from French French.

I get that, but it matters very little in a typical sentence. I bet you could understand almost everything with all vowels replaced with schwas.

> I bet you could understand almost everything with all vowels replaced with schwas.

No. Have you tried understanding someone who can only pronounce consonants but also whose dialect is foreign? It's unintelligible. Vowels absolutely serve a purpose. You're just used to hearing your own words spoken back to you in the same way you've always expected them.

Different accents use different vowels, but they remain comprehensible. It's specifically those accents that also change consonants that are taken as hard to understand, such as Scouse.

You can replace unstressed vowels with schwas and usually end up with something understandable. Stressed vowels cannot be schwaed. This is also specific to american English and doesn't work for other dialects, or even all american accents.

> Many of today's actors /actresses almost mumble their lines and make it absolute hell for the audio people to pick them up, whereas theatre trained individuals were loud and enunciate the crap out of everything.

There's zero correlation between how loud someone says their lines and how loud they are in the resulting mix when ADR is used.

There is a limit to how far you can push this for lines recorded live because of environmental noise, but in that case there's more control than ever before because of software like iZotope RX (which is close to magic).

This. These are largely mixing decisions, although it would be kinda weird to have a whispering actor boosted to sound as loud as a train in a scene.

The technical term for this is “Dialogue LRA” measured in LU: https://s3.amazonaws.com/izotopedownloads/docs/insight200/en... - this link also mentions some of the standards around this and now to use their software to adhere to those standards.

LUFS is also relevant here as a measure of loudness: https://blog.landr.com/lufs-loudness-metering/

https://auphonic.com/blog/2020/10/08/dialog-loudness-normali...

and

https://www.pro-tools-expert.com/home-page/2018/11/10/is-dol...

Give some insight into how this works.

Are lines ever recorded live? I imagined that they almost always re-record them in a sound booth later, it just seems like it would be so hard to get good sound on-set in most circumstances.

Yes, on-set audio is often used if possible. Re-recording is a bunch of extra work (you need to pay your potentially quite expensive actors for), needs matching, might not be as "in-the-moment" as the original, ... But re-recording is of course done where needed.

Many of today's actors /actresses almost mumble their lines and make it absolute hell for the audio people to pick them up, whereas theatre trained individuals were loud and enunciate the crap out of everything.

Sorry, but I work with a lot of Hollywood production companies and have been on plenty of sets, and this is a load of crap.

Today's actors and actresses speak their lines just fine. The problem is the trend toward louder audio f/x and music, which can drown out voices if not mixed correctly. See, e.g., Tenet for a horrific example of audio mixing, and the characters can barely be heard even when yelling; compare to The Batman where Pattinson rarely speaks above a whisper but is audible and understandable in every scene.

I always wondered what people have against Tenet's audio. I originally saw it with headphones and using vlc and it was fine. Then I saw it on a TV and it was awful. I wonder if it was tested in normal environments.

I remember that back in the day, actors re-voiced what they said on the set, but from a studio with proper recording gear, and sound engineers mixed these voices into the general soundscape.

Are modern movies shot in reporting style instead, with all the visuals and sounds recorded once and simultaneously? That would surprise me.

In some recent musical-to-film adaptations they recorded the voices ON THE SET during filming, often with no backing track or metronome, and then forced the music director, post-filming, to conduct the orchestra along with whatever the hell the actor was filmed doing: https://www.youtube.com/watch?v=1ikqU6G6Xgs

It seems pretty short-sighted to blame it on actors. They are a low-level cog in the machine that are trained to exactly what they are told. It's the job of the director to get the performance out of them that will look good on screen. Can't the sound teams talk to them?

Seems much more likely that, as other posters have said, movies are edited to play on the big screen at high volume.

> I'm forced to watch every movie with subtitles

The silver lining here may be that people stop treating subtitles as torture and see some non-English titles more often. There are some fun movie cultures out there beyond Hollywood, but I've found the "ugh, I've got to read‽" response common.

I've always found that subtitles for a non english movie actually results in me watching the move an walking away with a more memorable experience than if I'd just been listening. I'm not sure if that's a common experience but it's mine at least.

In the UK people prefer subtitles to dubbing, although that is different across parts of Europe. Foreign language programmes are quite popular.

I do love watching foreign shows on their original audio track. Korean especially has an awesome sound! For example right in the first ep of Squid Games, protagonist talking with his elderly mom, lines where he is disagreeing or making excuses end with this long up-and-down "huuuuuuuuuh!" sound that is the coolest thing ever :-)

I appreciate your inclusion of an interrobang here, which, like non-English movies, is too often overlooked.

I definitely watch all movies, English and foreign, with subtitles, and can't stand dubbed movies. I've found an amusing effect of well subtitled movies is that sometimes I forget that they're on, and, if I walk away and can only hear the dialogue but not see the subtitles, then I have to take a minute to process why I can no longer understand it ….

I have the opposite problem. Subtitles draw my attention, even when I understand what's said just fine. And then it's extra distracting when the subtitles change the meaning of what is said (which happens both when watching Japanese movies with English or French subtitles, or English or French movies with Japanese subtitles)

> which happens both when watching Japanese movies with English or French subtitles, or English or French movies with Japanese subtitles.

Also, though with less frequency, sometimes when the subtitles are in the same language as the speech, which I agree is distracting. (I am monolingual, so do not experience it otherwise.) But I had rather experience that dislocation than dubbing, which to me reliably ruins even a beautiful movie. Of course à chacun son goût!

As it happens, I was watching a foreign film (Italian) last night with english subtitles but the dialogue was also muffled and I found it just as difficult to watch.

I’d wager that many engineers are still mixing for the cinema experience, but the vast majority of people are consuming media via their TV’s, laptops, tablets and phones, with an extremely diverse range of audio quality and placement. Findings a happy medium is incredibly difficult, if not impossible.

There’s certainly a need for a ‘context aware’ format that can dynamically change the way sound - particularly speech - is delivered based on the device, volume, or even physical setting. (I suspect that this is about as useful as the ‘loudness’ button on many older amps though.)

I can’t wait for the day Netflix gives us the ability to adjust volume by category like video games. Different sliders for dialogue, background music, explosions, etc.

That would be a fair bet, but a lot of modern movies are difficult to understand in a cinema as well.

I've noticed an even newer effect: Making scenes almost all black. It might be a software problem with the HBO Max app's gamma and brightness settings, but it's absolutely the case that, during the day with thin curtains drawn, I had to put a thick blanket over my head and an iPad in order to see the highlights in some of the darker scenes in one of their recent original productions (probably Raised by Wolves Season One; can't remember for sure).

These were action scenes with grunting and no dialogue, so the blanket was the only way to follow the plot.

This sounds a lot like "dynamic contrast", i.e. varying the display's light source brightness with the scene brightness. I can't speak to whether or how the iPad implements it, but it's a common behavior of LCD screens. It's configurable on some devices.

You can just cover the light sensor (right next to the camera) if you don’t want to mess with menus instead of using a blanket.

But that would dim the display. Selectively illuminating the sensor would boost display brightness.

Yep I subbed to HBO to get GoT a couple of hours before "other sources" and immediately cancelled mid-way through last season. Night-time fights were just pitch black and a few smudges moving around the screen.

They probably just throw the thing into ffmpeg with the fewest amount of arguments that satisfies a codec and bandwidth cost metric and call it a day.

However, I've never had any problems under the same situations on Netflix using a "blessed" 4K HDR10 set-up (Fire TV Cube and a cheap Hisense TV).

In the past I have adjusted the black level on my video player to make movies like this watchable. It won't look good, but you can make the few shades they used visibly different.

I'm guessing you're using a stereo setup then, dialogue is usually mixed into the centre channel, downmixing a 5.1 or 7.1 stream to stereo will produce those results.

even worse is if you don't downmix to stereo, and all you are hearing is front left and right, which is often just the reverb effect channels for dialogue

Don't many people have just a stereo setup still (myself included)? Why would a newer standard be backwards-hostile like that?

As a counterpoint, I have both 5.1 via ARC and Dolby Atmos system hooked up via eARC and dialogue is still quiet in some modern releases. Older movies sound great though.

this will be because more recent movies will be utilising high dynamic ranges for audio, worked around by compressing/limiting but in doing so, you lose quality.

trade-offs for everything.

there was a time when stereo was seen as unnecessary faff, too.

edit: movies are mixed for theatres and no longer are the home releases adjusted to suit.

In many cases, people ostensibly using stereo sound may as well be using mono sound. Stereo speakers built into a TV that's 20 feet away, or watching a video on their smartphone without headphones; technically these people are using stereo sound but with the speakers so close to each other relative to the listener, it might as well be mono.

I think if you want to make widely accessible audio content, you should be taking this into account. Make it intelligible in mono first, then worry about the rest.

I watch using headphones and dialogues are still difficult to hear. Now I use vlc method to make them audible.

https://lifehacker.com/how-to-fix-movies-that-are-really-qui...

How to do this in mpv? Then we'll automate it with a short shell script!

There are a few ffmpeg filters you can use with mpv to get this effect, though you may have to fiddle with their options to get what you want. acompressor or dynaudnorm might be what you're looking for.

https://ffmpeg.org/ffmpeg-filters.html#acompressor

https://ffmpeg.org/ffmpeg-filters.html#dynaudnorm

To use them with mpv you'd do something like `mpv --af=lavfi="[dynaudnorm=param1=value1:param2=value2:param3=value3]" ...`

(make sure the first parameter is preceded by a '=', not a ':' because reasons)

that would be because headphones are stereo.

I've got the full klipsch THX ultra 2 system (7.2), and I also find dialogues in modern movies much harder to understand than in older releases. I think they should mix in more of the dialogue in the front left and right channels

considering the vast majority of people listen to movies in stereo, why don’t they include a decent stereo mix?

Christopher Nolan's Tenet was a prime example of horrible stereo translation. I could hardly understand the dialog. I've never changed volume so much during a movie. IIRC he refused to produce any audio mixes aside from those meant for the most premium 7.1 stream theatres. What a shame.

That mix was garbage even in 7.1. I had to add 10db to the center channel.

> will produce those results.

If that's commonly the case then whatever is doing the downmix isn't fit for purpose.

I found on Kodi you can boost the centre channel when downmixing, this helped quite a bit for me but I still watch with subtitles.

Gameplay and enunciation look quite standardized in modern movies. I remember a quote from a vietnamese director who said he cannot stand movies where you do not see people breath. Look at blockbusters: you never see people breath. When you do see them breath it is very much overplayed.

Female characters get to breath. Look at marvel movies. Black Widow and Scarlet Witch are seen pausing and breathing, mostly breathing in, but that is more for male gaze than anything else. In the pauses of fight scenes they are mouth agape, almost panting as the male characters stand like stones.

The “let’s turn luminosity to 0 to drive up the sale of screens with 5000000:1 contrast” got me as much as “let’s make the sound inaudible on stereo to drive up the sale of 7.1 sound systems.”

VLC should have a setting such as “flatten sound to 70dB all the time and luminosity to 50% the ability of the display in average”. Even pirated packs are becoming unwatchable.

> The “let’s turn luminosity to 0 to drive up the sale of screens with 5000000:1 contrast.

Pitch Meeting recently gave up the game there.

(in the latest video, Ryan jokes that they're making the screen darker and darker until they don't even have to project a video and can just play an audiobook over the speakers)

I've also seen GoT.

That problem is due to wrong downmix from multichannel to stereo. both VLC and Kodi have options to compress audio so that louder parts can be lowered. Here's the relevant Kodi wiki page: https://kodi.wiki/view/Settings/System/Audio

The problem exists in the cinema too, so it's not that.

God...the headache I had coming out of the new Blade Runner...terrible.

I wear earplugs for live music when I go to the movies nowadays

I thought about something like that too but than there are dialogues which are already hard to understand because they're not loud enough just before you get blasted with sound effects again...well yeah..one more reason why I don't go to the movies so much anymore.

The types of ear plugs for music are high-fidelity and barely distort the sounds you actually want to hear, like dialogue.

Can you recommend a product?

Not in particular. At the time I was residing in the UK and just swung by a music store on the way to a show and asked the guy what he recommended and it's been on my keychain for several years.

It’s also because producers and directors make more use of the “dynamic range” that theatre sounds systems are capable of today. They paid a ton of money for that music and they’re going to make sure you hear it. In surround setups, most of the dialogue is mixed into the center channel, which is behind the screen in theaters. At home, if you don’t have a center channel, you’re either relying on the stereo mix and however they decided to balance the speech on that, or if you play the surround version on a nice processor, it’ll probably let you adjust the center channel.

An audio compressor would also help here on consumer devices, as that would allow a variable degree and ratio of dynamic range flattening. Basically, it can form a band of volume and ensure that quiet sounds and loud sounds both fit into that. Sadly, most consumer gear, although already equipped with all the DSP they need to implement this in software, thinks consumers are absolute morons and/or that this wouldn’t be a selling point, so we get watered-down poorly-implemented dialogue boost features or largely no option at all except for the consumer to “ride the fader” and constantly adjust the volume. https://www.nytimes.com/wirecutter/blog/soundbar-can-help-he... has some info on what’s out there today to address this, but yeah: it’s mostly fixable if consumers had control of the center channel mix + a configurable compressor + configurable EQ. There are ways to do this in consumer-friendly ways for folks who don’t know of care to learn about mastering audio for film, but the current situation is pretty bad and it really makes me wonder why so much mixing targets home theater setups vs built-in crap TV speakers. At very least, streaming companies could provide an alternate audio track [algorithmically] mixed specifically for built-in audio devices that has more clear dialogue.

I personally watch with subtitles because I don’t want to miss a word.

This was perfectly satirized a decade ago https://www.mrlovenstein.com/comic/87

...and still hasn't been fixed.

Can you give examples? I've heard this before but have never noticed it myself. Is it possible you have hearing loss that you're unaware of?

Almost every movie that features Tom Hardy: https://www.youtube.com/watch?v=2iSOWRsmibc

And from personal experience, from the last movie I watched: https://youtu.be/AJ0uqAr8Q4E?t=106 . I'm not a native speaker, but (1) I've been communicating exclusively in English for years, (2) my hearing is quite good, (3) I'm wearing headphones, (4) there's nothing else happening in the scene apart from dialogue and music. But I still cannot understand Anne Hathaway's two sentences following the time I linked.

This movie also had a lot of another pet peeve of mine, which is shoulder dialogue where the foreground character is clearly not speaking, or saying something different.

I personally don't find this happening as much as OP says, but nolan movies (e.g. interstellar) are an example where i have definitely had this.

I've watched Interstellar with multiple audio setups and didn't have an issue understanding the dialog. It may have a higher dynamic range than most movies but that's a good thing.

Definitely a known Nolan issue.

I will go to my deathbed attesting that Tenet was the absolutely worst sound mixed movie of all time!

I don't think not understanding the dialog matrially changes that movie.

I agree with you on the former, not the latter. I find the "mumbling" more realistic dialogue (often in scenes where the script is supposed to be cramped and confusing) and feels more like natural human interaction.

But the imbalance of dynamics makes it problematic.

If this were a conversation about mixing for recorded or live music, someone would invariably mention hearing damage. It's a huge problem. Could it be that the technicians are deaf?

> I'm forced to watch every movie with subtitles on

Me, too… but my kids say they can hear everything just fine and please turn off the subtitles. It’s possible we’re just getting old.

Could it be we are use to media/news reporting so we struggle with movies?

I certanly also struggle with modern dialog. Another common factor is twenty years ago my hearing was much better.

Reply on news.ycombinator.com