Compressed music might be harmful to the ears
49 comments
·May 16, 2025lewispollard
kreetx
A less misleading title would be: "Constantly loud music might be harmful to the ears".
clan
A.k.a "The Loudness War"
Flex247A
By compression they mean dynamic range compression, which loudens softer sounds so they are more audible.
Pop songs for example, are heavily compressed because "compressed" music sounds "better" on cheap external speakers with bad dynamic range. Kinda like increasing the saturation on a cheap TFT panel.
perching_aix
Put briefly, researchers blasted guinea pigs with 102 decibels of music for 4 hours. They found that while both the compressed and non-compressed test tracks caused temporary damage, the compressed variant made the damage longer lasting, suggesting that there's more to ear damage than just the loudness.
From this, what I would take away is not necessarily that compression on its own is harmful but that... there's almost certainly more to ear damage than just the loudness, and compression is one way this can exhibit through. So I'd say the title and angle of the article is a bit misleading.
Naturally, when I say compression I mean it in the audio effect sense, as they do in the article.
tumult
It’s possible their compression settings were actually exaggerating the peaks instead of compressing them, and then they did nothing to control those peaks afterwards. This is a really common thing that can happen with a superficial use of compressors. Especially if you did averaged loudness-matching of the compressed signal with the uncompressed signal. It ends up being spikier than before compression. I would entirely believe a waveform with those added spikes would be more damaging than a controlled waveform that had been saturated or limited after compression. I don’t have access to the original publication, so I can’t check and find out.
ri0t
Or in short (as every modern producer already knows): Put a limiter on that!
Actually, put a limiter everywhere - if you work in software, they're real cheap..
jmkr
My Bitwig template: limiter on master track. My Renoise template: limiter on master track.
Now I have no fear of loading random presets.
lupusreal
Unless I understand compression all wrong, isn't the point to make the quiet parts loud? So in one case you have a song peaking at 102dB and in the other case virtually the entire song is that loud.
It's already known that hearing damage is cumulative, longer exposure to loudness is bad, so it seems like a common sense result in line with all existing research.
tumult
The pre-compressed and compressed versions were loudness matched to be the same overall loudness, according to the description. My guess is that they set the compressor to actually make the waveform spikier, without fully understanding what’s going on. Just a guess, though. I can’t check to find out.
titzer
If the damage is caused by overpressure, then obviously compression will deliver more overpressure for more time, i.e. a higher total transmitted acoustic energy. I didn't read the article because of paywall, but did they normalize by acoustic energy?
sixthDot
Compression reduces the range between the lower and the higher level so your ears are faced to a more or less "constant" pressure. Personally when I was into home production I only dared eating the peaks with a limiter and reasonable settings, e.g to gain one or two db, never more.
On top of that another problem with compression is that it is not neutral, bad compressors, especially in the digital domain can introduce aliasing.
Also search for the "loudness war". This is how we called the problem back in the mid 2000's.
jampekka
To nitpick: the ears are faced with more or less constant pressure variation. Constant pressure is silence.
sholladay
It’s worth noting that, on its own, dynamic range compression (DRC) merely reduces the volume of loud sounds above a certain threshold. But it’s not usually used on its own. The added headroom provided by DRC enables the use of additional makeup gain (MG), which turns up the volume of the whole signal after the compressor, including the noise floor. The end result is that the loud sounds are just as loud as they would’ve been before DRC+MG, but the quiet sounds are louder than they would’ve been, making the overall average volume over time also higher, which is what we tend to perceive as “loud”.
Beyond any potential health effects, DRC also just sounds bad after listening for more than a few seconds. It kills the liveliness of the audio and the signal-to-noise ratio.
So why is it used? Loudness is like candy to your ears, we are drawn to it and people tend to perceive louder music as sounding better, even if it’s exactly the same except for a change to the volume knob. DRC+MG is a bit like turning up the volume knob on your behalf, except it’s worse in almost every way.
One legitimate use case for DRC+MG, and in fact one of the original justifications for its widespread use, was the rise of portable music, where you want to be able to hear all of the instruments even though you are in a loud environment like the subway, for example. But this devolved into the “Loudness Wars”.
Tasteful use of DRC, perhaps without much MG, can also help “glue” an instrument within an audio mix. I might use it on a bass guitar, for example, to keep its volume under control so it stays in the background, providing rhythm, and doesn’t accidentally become the lead.
Ideally, the base format for all audio would have little to no DRC, but there would be other versions or built-in metadata to help playback devices adjust the listening experience to the environment by enabling DRC when it makes sense to do so or if the user wants it. Just as movies can be watched with or without HDR and subtitles. Unfortunately, the audio world hasn’t progressed much in that direction.
ri0t
Compression with a side chain also has some very interesting use cases like ducking ( https://en.wikipedia.org/wiki/Ducking )
vanattab
Is there any evidence that anc headphones are bad for your ears? I also feel odd presure in my ears when using them. I know they are supposed to cancel out the wave form but I was curious if latency,timing etc could lead to high frequency and in audible but still damaging presure on ear drum? I have no evidence for this other then discomfort using anc and suffering from tinnitus after using anc
voidUpdate
I really dislike using ANC because of that feeling of pressure on my ears, and the increased battery usage. Annoyingly, my main headphones (Bose QC 35 II) default to it being on at power-on, and the app doesn't let you turn that off, only reassign the action button to toggle between high, low and off, so every time I turn them on I have to click it twice
Telemakhos
As I (not a doctor) understand it, the lack of low frequency noise, which the ANC headphones cut out, tricks your ear into thinking that it’s blocked and the pressure is wrong. There is no actual pressure issue, but that signal gets sent to the brain because of the sudden lack of low frequency noise.
OutOfHere
Sennheiser has a bit less pressure than the other brands like Bose, but of course it does still has some.
When listening to music, I keep the volume as low as feasible, and I disable ANC. When listening to podcasts, I enable ANC.
In general, in music apps, the volume is already too boosted, and so even the minimum volume is too high and damaging. This is not an issue in podcast apps like AntennaPod.
There actually is evidence that ANC music makes it hard for people to understand what others are saying in the real world. If this effect starts to happen, reverse course immediately before it becomes permanent.
mfld
I wonder if letting guinea pigs listen to Adele's 'I Miss You' at 102 dB for long periods required approval from an ethics council.
bitexploder
Maybe the poor things had depression after that and it impacted their behavior.
yobbo
The compression effect in music amounts to increasing the amount of energy/loudness in the sound within the limits of the playback device to avoid "clipping".
Put very simply, it increases the amplitude of the input sound. Its parameters are controlled by internal timers, external triggers, timers and signals ("side-chaining"). For example; lowering the bass ("ducking") during the critical milliseconds of the kick-drum so that the volume of both can stay maximised while their sum stays within maximum amplitude.
On voice-over, in radio and podcasts it makes voices sound even and "boomy/strong".
Originally probably necessary way to make recordings listenable in cars and noisy environments with lower spec speakers.
Nowadays producers use it to "sound louder" and thus make stronger impressions.
tumult
Lookhead limiting is more commonly used to sound louder, not compression. Compression is usually done for flavor. It’s not that great at making things louder, because traditional compression actually exaggerated the spikes in amplitude at the start of percussive sounds, which pop and dance music has a lot of, requiring additional work to tame. Modern drum processing usually uses a combination of compression (sometimes with upwards compression), lookahead limiting, and saturation.
ZoomZoomZoom
Ear fatigue is a real thing. No surprise deliberate listeners also are generally more interested in higher dynamic range material (speaking of loudness here, not encoding).
It's much easier to deliver reasonably dynamic masters nowadays than, say, 10 years ago, but the fight is still not over. Loudness normalization is still inconsistent between streaming platforms and one lagging/idiosyncratic actor (usually some smaller, locally-important one) can spoil the impression, breaking uniform listening experience, and force an author to reconsider what loudness and dynamic range they aim for.
posix86
Not sure if that transfers to compressed music in general. If even the non-compressed guinea pigs had hearing loss, that suggest the level of 102dB, though below the legal safety level, is harmful to their ears. Total harm done if you listen to music that's too loud is proportional to <loudness>*<time>. Since compression makes quiet parts louder, that amount would increase, explaining the increased hearing loss for guinea pigs that listen to the compressed amount. But if the overall level is below the harmful threshold, does that also happen?
eimrine
Is compressed music more harmful for the speakers? My understanding of compression is that a sinus input becomes more like a square wave (meander) on output which means more peak power in the pivot points.
This seems to be referring to the audio effect known as a compressor, rather than data compression techniques, for what it's worth.