Loudness War - Reports from the Frontlines (1 Viewer)

[aoboa]

You sure about this?
There are physical limits to how much your eardrums and the little sound-sensitive hairs inside you ears can actually move, I would have thought, so if what you're listening to is loud enough you will get clipping in your ears, regardless of whether the source is clipped or not. I think you might be onto something there Conor

Absolutely!
Your brain adjusts the threshold of your hearing for loud sounds much the same way it does for quiet sounds. When you enter a loud environment your brain adjusts the threshold of what's going on so you can understand it. With quiet sounds, the opposite. This is a seperate mechanism to your ear. Above 85dB your ear starts running into physical problems particularly with high frequencies.
I guess this is clipping of a sort but it's more accurate to compare it to the frequency response of a mic which is normally measured with a signal at a particular amplitude. Outside this spec the quoted frequency response doesn't hold. I guess it's the same with your ears.

 

[taubstumm]

don't want to get too stephen pinker on you, but...

it's actually an internal mind/brain process. your ears/eyes/nose, as physical devices resulting from evolution by natural selection, recieve stimuli from the environment at whatever force they're at. the stimuli are passed to the brain, where various (unconcious) processes are carried out on them in various orders in order to render the stimulus interpretable to the concious brain. so loud sound is squished, quiet is amplified, etc... this is why you can adjust to a loud gig and be fine while you're there, even though you're doing yourself damage at the time.

the human equivalent of digital clipping is when you get 'ouch'-level uninterpretably loud sounds followed by plenty of tinnitus.

 

[Super Dexta]

there is no such thing as "too steven pinker".

 

[guy of the tiger]

nah i think your eardrum just bursts when it's pushed to it's limit then you start bleeding out your ears apparently.

not nice stuff.

your ears adjust to the loud volumes to an extent. It's sudden bursts of loud noise that's the real killer because your ear isn't ready for it.

You sure about this?
There are physical limits to how much your eardrums and the little sound-sensitive hairs inside you ears can actually move, I would have thought, so if what you're listening to is loud enough you will get clipping in your ears, regardless of whether the source is clipped or not. I think you might be onto something there Conor

 

[Pantone247]

nah i think your eardrum just bursts when it's pushed to it's limit then you start bleeding out your ears apparently.

I'll tell you what the weirdest thing is, when exposed to prolonged loud sounds is how folk go from very restless and agiatated, to a weird state of collective zoned-out-ness, the you get that strange vertigo/claustrphobia disorientation thing going on. It's really weird to see peoples reactions to extreme volume, like a Bullets gig... people either get really angry and can't hack it, or go into some sort of weird trance thing...

It's a very tanigable trippy experiecne... and a 'deadly' 'buzz'

 

[taubstumm]

when exposed to prolonged loud sounds is how folk go from very restless and agiatated, to a weird state of collective zoned-out-ness

from http://www.buddyhead.com/music/kevinshields/page2.html -

Buddyhead: Awesome. What were people’s reaction to the noise? Did they get bummed out?

Kevin: Usually people would experience a type of sensory deprivation, and they would lose the sense of time. It would force them to be in the moment, and since people don’t usually get to experience that, there’d be a sense of elation. There would be a feeling of, “Wow, that was really weird, I don’t know what happened, but I suddenly heard this symphony…” Because it was such a huge noise with so much texture to it, it allowed people to imagine anything. Like when you hypnotize somebody, and nothing becomes something. That was what the whole purpose became. 1/3 of the audience would always think it was really shit, and try to leave, or get as far away as they could, and the other 2/3 really liked it. One time half the audience tried to leave, and it caused mild panic because they all tried to leave at once and got stuck at the door and got crushed. Then a whole gang of them came back towards the stage and tried to get us because they were so angry that they couldn’t get out. Like a village mob or something.

Buddyhead: Haha, where was that?

Kevin: Canada.

Buddyhead: Figures.

 

[guy of the tiger]

yeah bullets is pretty intense allright.

That's kinda 'the way' your supposed to listen to noise music apparently. i.e. it just sounds like drone at the begining where you would normally be listening for obvious changes in the sound. After a few minutes, when youre enveloped by the sound, you begin to notice the subtle fluctuations in the drone and the sound varying slowley over time.

when i stop listening to noise after listening at high volumes, i feel dizzy and disoriented.

I'll tell you what the weirdest thing is, when exposed to prolonged loud sounds is how folk go from very restless and agiatated, to a weird state of collective zoned-out-ness, the you get that strange vertigo/claustrphobia disorientation thing going on. It's really weird to see peoples reactions to extreme volume, like a Bullets gig... people either get really angry and can't hack it, or go into some sort of weird trance thing...

It's a very tanigable trippy experiecne... and a 'deadly' 'buzz'

 

[nlgbbbblth]

I'll tell you what the weirdest thing is, when exposed to prolonged loud sounds is how folk go from very restless and agiatated, to a weird state of collective zoned-out-ness, the you get that strange vertigo/claustrphobia disorientation thing going on. It's really weird to see peoples reactions to extreme volume, like a Bullets gig... people either get really angry and can't hack it, or go into some sort of weird trance thing...

It's a very tanigable trippy experiecne... and a 'deadly' 'buzz'

I remember My Bloody Valentine's SFX gig in 1992 - the extended mid-section of You Made Me Realise which lasted for what felt like 40 mins. I was ok with earplugs but my friends weren't.

 

[kraster]

When audio signals reach or exceed the threshold of pain the cochlea Becomes rigid from over stimulation. It's a kind of Mechanical Compression within the Ear. Repeated RMS signals above the threshold of pain will damage your hearing over time but repeated transient peaks above the threshold of pain will kill your hearing over a much shorter times. The problem with mastering methods nowadays is that the whole concept of loudness is mis-perceived. Someone mentioned Highway to Hell in a thread as being a loud album. It is a loud album but if you play it against any modern recording it seems quieter. It doesn't matter how technically loud something is. The important thing is to keep the integrity,fidelity and dynamics of the original recording. Things don't seem loud when there is no sense of silence. A great example of this is a good quality recording of Ravels Bolero. It gets uncomfortably loud but only because it's almost imperceivably quiet at the beginning.

 

[Jimmy Magee]

When audio signals reach or exceed the threshold of pain the cochlea Becomes rigid from over stimulation. It's a kind of Mechanical Compression within the Ear. Repeated RMS signals above the threshold of pain will damage your hearing over time but repeated transient peaks above the threshold of pain will kill your hearing over a much shorter times. The problem with mastering methods nowadays is that the whole concept of loudness is mis-perceived. Someone mentioned Highway to Hell in a thread as being a loud album. It is a loud album but if you play it against any modern recording it seems quieter. It doesn't matter how technically loud something is. The important thing is to keep the integrity,fidelity and dynamics of the original recording. Things don't seem loud when there is no sense of silence. A great example of this is a good quality recording of Ravels Bolero. It gets uncomfortably loud but only because it's almost imperceivably quiet at the beginning.

Trouble is, they don't seem loud - but if you switch with something that's at a lower overall RMS it'll seem immediately quieter, even if over time it seems louder because it has more dynamics

 

[egg_]

Been listening to System of a Down's first album, and it's mastered by the infamous Vlado Meller (the loudest of the loud) - when you import a wav and look at the waveform it's at -1 dB almost all the time
But here's the thing - it sounds amazing. Obviously it's not just the mastering that makes it sound good but ... well I was a little surprised to find that after all the criticism of "modern" mastering a record that I think sounds great was mastered in this way.
Maybe SOAD get away with crushing it cos they have sudden silences that make up for the lack of dynamic range elsewhere?

 

[kraster]

Again It's dependant on program material. Ravel's Bolero uses the whole dynamic range. I think the snare at the start is about -60 db all the way up to 0 on the version I have. Using a program limiter or brickwall compression on this would destroy the dynamics so important to this piece. On louder material with an increased RMS signal you don't notice the compression as much as their are less transient peaks particularly if it has been recorded hot onto tape or a lot of compresion has been used. But things like snare drums and Bass drums can be pushed back into the mix if they are over-compressed. Personally I think the best mastering keeps the integrity and dynamics of the Mix intact while making the overall material louder. I know that's probably what most people prefer but 9 out of 10 recordings seem to be smashed into a collection of square waves. Add to this the signal processing on Radio and you'll be able to hear what the singer had for breakfast. Every music delivery sytem has a volume knob,slider or dial if I want something louder.

 

[RED(tape)MENACE]

Yeah, and with this last post I totally agree, but to play devil's advocate herre for a minute,,,
I've heard bands worrying about their masters being too quiet. It really does seem like no-one trusts the listener to know how to use their volume knob. How can you argue with someone who's worried that the music they spent a lot of time, and probably plenty of their own money on, is ignored because the DJ/promoter/A&R(!)/whatever is too used to hearing everyone else's smashed pancake abominations?

 

[C24]

Is it not true to say though that most people who make records of 'loud music' want their music to be perceived as louder than everyone else's (or at least as loud)? Surely the job of the engineer is now and always has been to deliver the music as the artist wants it?

Another effect of the perception of heavily limited music is the effect of power... Heavily limited music has a feeling of weight and density that cannot be achieved by applying dynamics to any single element of a mix. It is an effect applied to the programme material, the point being that is an effect in the same way a reverb or dealy is.

Another thing to throw into the discussion is this. What are the excuses for square wave clipping these days? Almost all level increases in mastered music these days are attained by using dsp algorithms as opposed to classic digital or anolag limiters. It seems that in my opinion it's generally laziness in mastering that causes this.

 

[eoin5]

Speaking of throwing something into the discussion, when you record a piece of music you try to make it as loud as possible so that the signal to noise ratio is good and the louder it is the more quality is captured, to convey this quality on mixdown and mastering it should also be as loud as possible, especially because its going to be put from 24bit onto 16bit media persumably.

So its quite hard to pin down, you can have a very dynamic piece of music but not convey the tonal textures as best as possible no matter how the listener messes about with their volume controls, or you can compress the fuck out of it and convey the true quality of the recording.

If you pick examples you can see its up to the taste of the musicians, it works for some and not for others. Think of Ash's Meltdown, a highly compressed piece of music that you can hear all the instruments clearly and appriciate the tone of each fully (the bass for clones sounds class), or think of Carmina Burana[size=-1] by Carl Orff, which would be ruined if you had the start as loud as the rest. Personally I like a good bit of compression, but I'm not Carl Orff .[/size]

 

[RED(tape)MENACE]

Speaking of throwing something into the discussion, when you record a piece of music you try to make it as loud as possible so that the signal to noise ratio is good and the louder it is the more quality is captured, to convey this quality on mixdown and mastering it should also be as loud as possible, especially because its going to be put from 24bit onto 16bit media persumably.

This is much less of a problem now with the spread of 24-bit recording systems. It was much more important to be pushing the digital levels with 16-bit, since you were much closer to the low level where you'll run into quantizing (sp) noise. Once you're avoiding quantizing the 'quality' remains the same, no matter how loud it is.

So its quite hard to pin down, you can have a very dynamic piece of music but not convey the tonal textures as best as possible no matter how the listener messes about with their volume controls, or you can compress the fuck out of it and convey the true quality of the recording.

Hmm. This needs a bit of work. Again, there's a difference between tone and volume. Compressing the fuck out of it will alter the tone (and not usually for the better) and NOT convey it properly. As for how a listener percieves tone at different volumes, do a search for Fletcher-Munson Curves to learn more.

Think of Ash's Meltdown, a highly compressed piece of music that you can hear all the instruments clearly and appriciate the tone of each fully.

I can hear the instruments just fine on all the albums I own that aren't compressed to shit. And if you can listen to meltdown twice in a row without getting burnt-out ears, good luck to you.

 

[Pantone247]

So its quite hard to pin down, you can have a very dynamic piece of music but not convey the tonal textures as best as possible no matter how the listener messes about with their volume controls, or you can compress the fuck out of it and convey the true quality of the recording.

hhhhhhhhhmmmmm.... I'd disagree... I really don't think it's a toss up between the tone or dynamics... really really don't think so...

Maybe we're in a case of different tastes here, s'true the over compressed drums on most any flamming lips albums are fantastic and do bring out the charecter of the kit and room, it's a great rock sound. And if you're going for a slab of metal guitar sound compressing can really bring that crunch to the fore... but I don't think this is the strict rule...

I feel, that on acoustic instruments or keyboards, vocals too... compressing can really bring up the muddy mid freq, you can get terribly boxy sounding very quickly since, generally, that the freq range where the compressor, working at a heavy setting, is pushing or squashing the signal.

Compression is a bit off a WOW! effect, when you first get it right you can go mad and compress everything, cause it really can pull a lot of the life into individual recording "dude you can hear every finger noise of the guitar, it's awesome" or "you can actually hear the singers fillings resonating, it's THAT cool!!" but I garuntee when you put all these heavily compressed elements into a mix, most times it's just gonna be a big mess of everything fighting against each other.

Anyhow as I said I think we're getting into tastes here...

hey lets start a thread about reverb then I can be all "use anything less then 'Chatedral' and your a total pussy"

 

[aoboa]

Compression rules but 'space is the place'

It's best looked on as a way of controlling level as opposed to maximising it. I agree with acoustic instruments - over use destroys the sound. You can avoid this by eq'ing before compression but generally light compression is best.
Over compression on stuff like brass is terrible - it generally makes it sound like a cheesy keyboard trumpet preset.
Certain instruments just don't take compression well because the dynamics of are what makes the instrument sound the way it does - Acoustic guitar, brass, piano etc.
Like Pantone said... if you compress every element you end up with mush.
Vocals, bass and drums generally need it. After that it's down to preference but generally you're better to group instruments to different buses and compress them as required. Leads to a more focused and realistic sound and also means you don't need stupid hard compression on the stereo bus to keep your overall level in check.

 

[kraster]

I agree that it's totally down to the material listened to but even with very dense material smashing it to a pulp will only make it sound flat and boring. Right, the rms signal might be closer to 0 db but if it lacks transient peaks then the apparent loudness will be diminished. If you square off a signal due to over compression it will introduce harmonics that do not exist in the original signal. A square wave consists of odd partials which can sound awful as the partials bear no relation to the fundamental of the music. A common practise is to squeeze the bejaysus out of overheads on a kit. This can sound good until the crash cymbals come in and all your phase difference and harmonic content go out the window. It does have its uses e.g. The Flaming lips but can sound pretty crap too. Also over compression leads to compression artifacts. These are additional bits of noise not inherent in the original signal. Old Valve-based opto compressors like the the teletronix la2a sound good because the sound saturates to even partials the more you drive it. Radio has driven this ridiculous trend in squashing everything to a pulp. So even if you have your nicely dynamic master by the time it reaches the signal processors in the radio station it'll be mush by the time it hits your receiver.

 

[eoin5]

there's a difference between tone and volume. Compressing the fuck out of it will alter the tone (and not usually for the better) and NOT convey it properly. As for how a listener percieves tone at different volumes, do a search for Fletcher-Munson Curves to learn more.

I hear what your saying GrRrrrR you do change the tone when compressing but this brings out the subtle elements of a recorded sound that may otherwise not be heard (for good or bad), think of motorcycle emptiness and the awsome guitar tone. Imagine all the texture suddenly missing... nooooo! as aoboa said some instruments dont take kindly to compression at all, but again that can be down to taste.

I can hear the instruments just fine on all the albums I own that aren't compressed to shit. And if you can listen to meltdown twice in a row without getting burnt-out ears, good luck to you.

As a bassist i like to hear the bass, not just to acknowledge its existance but to hear how its sounds. There are so many albums that the bass sounds on par with your average ring tone , a little compression helps a lot.