Fuck Brickwall Mastering

[Guest skept]

This anti-loudness war thing is quite the fashionable subject at the moment. Idea - if you dont like loud tracks then just dont listen to them?

 

"Engineers" ranting about the loudness wars are not working professionals, 99.9% of the time they are wannabe audiofools with no actual paid work.

 

Mastering is a service based industry where the engineers do what they are told to make money - it's that simple.

 

The ranting forum engineers are fakers - FACT.

 

If you are looking for mastering, hire a proper engineer and tell them what you want. A working professional should listen to you, and deliver what you asked for, not lecture you on the benefits of dynamics and rant about how all music nowadays is crushed to oblivion. Why the hell would potential clients want to hear that shit?

 

A good tune is a good tune. ;)

[oscillik]

:handegg:

[cruising for burgers]

loudness war

[Guest sirch]

:handegg:

 

:dick head: lol

 

seriously tho, that's quite a lot of "hand eggs", osc. you just look like a nob now mate. :)

[ascdi]

i like clipping and distortion (in my music) but it's all part of my punk rock aesthetic, i have had complaints....

punk rock has nothing to do with clipping. punk rock has everything to do with analogue distortion that comes from over-cranking amps and such, which creates additional harmonics, but most importantly doesn't clip waveforms.

 

clipping, on the other hand, occurs only in the digital realm, and just turns whatever hits the 0dB wall into a square wave that crunches, and not in a good way.

I don't think clipping and "analog distortion" are as different as you are saying. I used to think they were, but I am starting to change my mind a little bit.

you can change your mind all you like, doesn't change the fact that digital clipping and analogue distortion are both technically and sonically very different.

 

:handegg:

I'd be interested to know exactly what you mean by that aside from the usual pseudo-theoretical bullshit most people read on blogs and in plugin promotional copy.

 

I'm not trying to be argumentative (although I do disagree), but I am curious what specifically you are talking about.

[bossman]

listening to bibio right now and it feels like every tone and instrument is being thrown ontop the other, a mountain of strained sound that plateaus into a gross imposing car advert demanding yr ears. it sounds terrible when played loud on decent speakers, worse on proper ones. it's like a quick fix to make a track buff, demand attn with fully flexed beats+tones, punchy and roided out, often way too saccharine, like a sparks w/ a cupful of whey protein. it's just gnarly.

[oscillik]

i like clipping and distortion (in my music) but it's all part of my punk rock aesthetic, i have had complaints....

punk rock has nothing to do with clipping. punk rock has everything to do with analogue distortion that comes from over-cranking amps and such, which creates additional harmonics, but most importantly doesn't clip waveforms.

 

clipping, on the other hand, occurs only in the digital realm, and just turns whatever hits the 0dB wall into a square wave that crunches, and not in a good way.

I don't think clipping and "analog distortion" are as different as you are saying. I used to think they were, but I am starting to change my mind a little bit.

you can change your mind all you like, doesn't change the fact that digital clipping and analogue distortion are both technically and sonically very different.

 

:handegg:

I'd be interested to know exactly what you mean by that aside from the usual pseudo-theoretical bullshit most people read on blogs and in plugin promotional copy.

 

I'm not trying to be argumentative (although I do disagree), but I am curious what specifically you are talking about.

well firstly, i'm not going to go into what qualifications i may or may not have on the subject, but i shall go as far as to acknowledge that the information i have gleaned is not from internet blogs or plugin promotional copy.

 

secondly, if you can't actually hear the difference, then there's no real point in discussing this because whatever is said, you can't hear the difference so what does it matter?

[ascdi]

i like clipping and distortion (in my music) but it's all part of my punk rock aesthetic, i have had complaints....

punk rock has nothing to do with clipping. punk rock has everything to do with analogue distortion that comes from over-cranking amps and such, which creates additional harmonics, but most importantly doesn't clip waveforms.

 

clipping, on the other hand, occurs only in the digital realm, and just turns whatever hits the 0dB wall into a square wave that crunches, and not in a good way.

I don't think clipping and "analog distortion" are as different as you are saying. I used to think they were, but I am starting to change my mind a little bit.

you can change your mind all you like, doesn't change the fact that digital clipping and analogue distortion are both technically and sonically very different.

 

:handegg:

I'd be interested to know exactly what you mean by that aside from the usual pseudo-theoretical bullshit most people read on blogs and in plugin promotional copy.

 

I'm not trying to be argumentative (although I do disagree), but I am curious what specifically you are talking about.

well firstly, i'm not going to go into what qualifications i may or may not have on the subject, but i shall go as far as to acknowledge that the information i have gleaned is not from internet blogs or plugin promotional copy.

 

secondly, if you can't actually hear the difference, then there's no real point in discussing this because whatever is said, you can't hear the difference so what does it matter?

I dunno, let's pretend we're in a hypothetical universe where you do have something truly worthwhile to bring to the discussion. In that universe, for you to not enlighten other people with your wisdom would make you kind of a bad person.

 

I have other reasons, that's just the first one that came to mind.

[oscillik]

well firstly, i'm not going to go into what qualifications i may or may not have on the subject, but i shall go as far as to acknowledge that the information i have gleaned is not from internet blogs or plugin promotional copy.

 

secondly, if you can't actually hear the difference, then there's no real point in discussing this because whatever is said, you can't hear the difference so what does it matter?

I dunno, let's pretend we're in a hypothetical universe where you do have something truly worthwhile to bring to the discussion. In that universe, for you to not enlighten other people with your wisdom would make you kind of a bad person.

 

I have other reasons, that's just the first one that came to mind.

let's pretend that we're on the world wide web, browsing a forum for music where there is a large percentage of people that, just like anywhere else on the internet, have made up their own minds about stuff and won't budge from that line of thought.

 

now let's try to have a discussion.

 

everyone is right and everyone is wrong.

[ascdi]

Well now I'm really confused. I stated my opinion already which you "refuted" with facts. I was just hoping to get an idea of what some of these facts were, where they came from, what exactly they are talking about, etc. So I could decide if it was worth re-thinking about all this stuff or not. So you see, it's not useful TO ME unless I know more about what you are exactly referring to. Otherwise this whole thing just turns into an Audio Penis Contest that benefits nobody.

 

Anyway in the interest of stimulating discussion here are two "provocative" "opinions":

 

1. Most distortion discards information, therefore most distortion reduces the uniqueness of audio. Therefore, there are not many different kinds of "distortion", in a digital sense.

 

2. There is a continuum between "clipping" and "compression"—they're not two distinct things. Consider a very simple lookahead limiter with a hard knee being fed a signal that is entirely above the threshold. In what way is that different from clipping?

[Braintree]

It's not something like Clark's crazy sound auteur idea of using compression & limiting as a type of dynamic instruments in the song, it's straight up ignorance on the part of sound engineers and the producers who just say "yeah that sounds heavy, bro, awesome"). Why doesn't anybody realize that a DJ or a home listener CAN EASILY JUST TURN THE VOLUME KNOB UP IF THEY WANT IT LOUDER without the track having to be totally butchered before it even gets to the consumer?

 

This is why all the DJs and clubgoers are deaf anymore.

 

It's so it's louder on the radio. The loudest song usually gets more attention and sells more records.

[Braintree]

Anyway in the interest of stimulating discussion here are two "provocative" "opinions":

 

1. Most distortion discards information, therefore most distortion reduces the uniqueness of audio. Therefore, there are not many different kinds of "distortion", in a digital sense.

 

2. There is a continuum between "clipping" and "compression"—they're not two distinct things. Consider a very simple lookahead limiter with a hard knee being fed a signal that is entirely above the threshold. In what way is that different from clipping?

 

1. There's soft distortion and hard distortion in varying degrees. Why does the guitar distortion in a Slayer track sound different than a Pantera track?

 

2. If it's above the threshold, it's not compressing. Clipping/peaking is when you cross over 0dB...

[ascdi]

Anyway in the interest of stimulating discussion here are two "provocative" "opinions":

 

1. Most distortion discards information, therefore most distortion reduces the uniqueness of audio. Therefore, there are not many different kinds of "distortion", in a digital sense.

 

2. There is a continuum between "clipping" and "compression"—they're not two distinct things. Consider a very simple lookahead limiter with a hard knee being fed a signal that is entirely above the threshold. In what way is that different from clipping?

 

1. There's soft distortion and hard distortion in varying degrees. Why does the guitar distortion in a Slayer track sound different than a Pantera track?

 

2. If it's above the threshold, it's not compressing. Clipping/peaking is when you cross over 0dB...

Hi.

 

1. Okay, maybe I can try and get at this a slightly different way. Say you take a pure sine wave, clip it super hard, and then compare it to a square wave at full scale. How different are those two things?

 

Suppose you do the same thing, but then normalize both waves back down to relatively quiet levels (say, -30 db), and then lowpass them at like 2k. How different are they now?

 

I'm not saying I know the answers to these things, I'm just asking.

 

2. By above I mean louder than the threshold. Suppose you feed a primitive peak limiter audio that is entirely above its set threshold. How is what it's doing different from what clipping the signal at that threshold would do?

[Braintree]

Anyway in the interest of stimulating discussion here are two "provocative" "opinions":

 

1. Most distortion discards information, therefore most distortion reduces the uniqueness of audio. Therefore, there are not many different kinds of "distortion", in a digital sense.

 

2. There is a continuum between "clipping" and "compression"—they're not two distinct things. Consider a very simple lookahead limiter with a hard knee being fed a signal that is entirely above the threshold. In what way is that different from clipping?

 

1. There's soft distortion and hard distortion in varying degrees. Why does the guitar distortion in a Slayer track sound different than a Pantera track?

 

2. If it's above the threshold, it's not compressing. Clipping/peaking is when you cross over 0dB...

Hi.

 

1. Okay, maybe I can try and get at this a slightly different way. Say you take a pure sine wave, clip it super hard, and then compare it to a square wave at full scale. How different are those two things?

 

Suppose you do the same thing, but then normalize both waves back down to relatively quiet levels (say, -30 db), and then lowpass them at like 2k. How different are they now?

 

I'm not saying I know the answers to these things, I'm just asking.

 

2. By above I mean louder than the threshold. Suppose you feed a primitive peak limiter audio that is entirely above its set threshold. How is what it's doing different from what clipping the signal at that threshold would do?

 

Jeesh, I didn't get enough sleep last night. *If it passes over the threshold it's compressing.

 

1. a. Sine & square waves have different partials. A simple square wave is 50% harmonic/non-harmonic. There's a matter of phasing that makes them sound different.

1. b. The sine wave is still distorted, and the square wave still has a more buzzy sound considering its non harmonic partials.

 

2. A limiter limits based on a ratio. Without the ratio, I can't tell you how it will behave since that's the thing that dictates how many decibels it cuts, and at what point it cuts [0dB, -1dB, -2dB, etc]. It's designed to reduce audio that's set over the threshold, and snags the audio before it goes through the output to clip.

 

A 2:1 ratio means for every 2dB in, there's on 1dB sent out. 10:1 means for every 10dB in, there's 1db sent out, ∞:1 means anything that goes in, only produces 1dB out, and so on.

[Guest sirch]

i like clipping and distortion (in my music) but it's all part of my punk rock aesthetic, i have had complaints....

punk rock has nothing to do with clipping. punk rock has everything to do with analogue distortion that comes from over-cranking amps and such, which creates additional harmonics, but most importantly doesn't clip waveforms.

 

clipping, on the other hand, occurs only in the digital realm, and just turns whatever hits the 0dB wall into a square wave that crunches, and not in a good way.

I don't think clipping and "analog distortion" are as different as you are saying. I used to think they were, but I am starting to change my mind a little bit.

you can change your mind all you like, doesn't change the fact that digital clipping and analogue distortion are both technically and sonically very different.

 

:handegg:

I'd be interested to know exactly what you mean by that aside from the usual pseudo-theoretical bullshit most people read on blogs and in plugin promotional copy.

 

I'm not trying to be argumentative (although I do disagree), but I am curious what specifically you are talking about.

well firstly, i'm not going to go into what qualifications i may or may not have on the subject, but i shall go as far as to acknowledge that the information i have gleaned is not from internet blogs or plugin promotional copy.

 

secondly, if you can't actually hear the difference, then there's no real point in discussing this because whatever is said, you can't hear the difference so what does it matter?

 

:dick head:

 

"qualifications" - gotta love oscillik! lol

[ascdi]

Anyway in the interest of stimulating discussion here are two "provocative" "opinions":

 

1. Most distortion discards information, therefore most distortion reduces the uniqueness of audio. Therefore, there are not many different kinds of "distortion", in a digital sense.

 

2. There is a continuum between "clipping" and "compression"—they're not two distinct things. Consider a very simple lookahead limiter with a hard knee being fed a signal that is entirely above the threshold. In what way is that different from clipping?

 

1. There's soft distortion and hard distortion in varying degrees. Why does the guitar distortion in a Slayer track sound different than a Pantera track?

 

2. If it's above the threshold, it's not compressing. Clipping/peaking is when you cross over 0dB...

Hi.

 

1. Okay, maybe I can try and get at this a slightly different way. Say you take a pure sine wave, clip it super hard, and then compare it to a square wave at full scale. How different are those two things?

 

Suppose you do the same thing, but then normalize both waves back down to relatively quiet levels (say, -30 db), and then lowpass them at like 2k. How different are they now?

 

I'm not saying I know the answers to these things, I'm just asking.

 

2. By above I mean louder than the threshold. Suppose you feed a primitive peak limiter audio that is entirely above its set threshold. How is what it's doing different from what clipping the signal at that threshold would do?

 

Jeesh, I didn't get enough sleep last night. *If it passes over the threshold it's compressing.

 

1. a. Sine & square waves have different partials. A simple square wave is 50% harmonic/non-harmonic. There's a matter of phasing that makes them sound different.

1. b. The sine wave is still distorted, and the square wave still has a more buzzy sound considering its non harmonic partials.

 

2. A limiter limits based on a ratio. Without the ratio, I can't tell you how it will behave since that's the thing that dictates how many decibels it cuts, and at what point it cuts [0dB, -1dB, -2dB, etc]. It's designed to reduce audio that's set over the threshold, and snags the audio before it goes through the output to clip.

 

A 2:1 ratio means for every 2dB in, there's on 1dB sent out. 10:1 means for every 10dB in, there's 1db sent out, ∞:1 means anything that goes in, only produces 1dB out, and so on.

Yo.

 

1. I'm a little foggy but I'm pretty sure a square wave has only odd, harmonic partials. It's a saw wave that has both odd and even partials. They are all harmonic (as in, whole number multiples of the fundamental) as far as I know.

 

b. I was kind of being a dick before by not really saying what I mean. Which is: if you clip a sine wave hard enough (imagine amplifying it by 999999 db), you get something close to (and eventually, you get exactly) a square wave. Not at all coincidentally, you have just added a ton of distortion, which happens to correspond with what you'd get by just starting with a square wave in the first place, only at twice the volume. It's one and the same process, in other words.

 

2. Let's stick with a simple hard limiter like L1. The whole purpose of that plugin is to prevent any signal at all from going over 0 dbfs (or whatever you set the ceiling at). So, it has an infinite ratio. ALL A LOOKAHEAD LIMITER DOES is look ahead to watch for samples that are going to pass above the threshold and "ramps in" the gain reduction to meet them. Then, it ramps out the gain reduction according to the release value. If you could somehow set a limiter plugin to have zero attack (that is, no lookahead) and zero release (that is, an instantaneous release) then all of a sudden you are clipping, not limiting.

 

So all "limiting" means (again, in the sense of a very basic algorithm) is clipping the signal, but with some ramp time in and out of the gain reduction to mitigate the WORST sounding parts of the clip. So there's a continuum between clipping and conservative limiting where, as you get less aggressive, you distort the music less (ie. you give the limiter more time to ramp in and out of its worst-sounding artifacts). The point is, these aren't two COMPLETELY DIFFERENT things. They are actually the same thing—that's why you hear things about L1 being distorted sounding and L2 being cleaner but way louder, etc, etc.

[ascdi]

What gets even interestinger is if you start to consider clipping as a production tool (which, as it's been mentioned in this thread, is happening all the time now), you start to think about, okay, what are the really bad types of clipping to avoid, and what are the "okay" types of clipping—in terms of the health of your audio.

 

For example, there's this whole issue of intersample distortion, which is basically clipping distortion that results from the D/A converter trying to reconstruct your (non-clipping) digital waveform. To illustrate this, generate a naked square wave at a low frequency and high amplitude, and then lowpass it at a fairly high frequency. You'll see that the amplitude of the signal actually increases when you do this. This is exactly what the D/A converter is doing when it reconstructs your 44.1 kHz audio, and that's why you see all these limiters and plugins now on the market with oversampling meters—because it's not enough to make sure that no samples touch 0 dbfs if their analog representation goes on to clip some other part of the converter or amplifier.

 

Another way to say this is that different D/A converters behave differently when asked to play back signals at or close to clipping.

 

Then, if you want to talk about "distortion" in the musical sense, well there's all sorts of other things to talk about. For example, clipping a guitar amplifier is actually not all that different from clipping a digital signal, except usually there's a speaker attached to the amp that, because it can't keep up with all the high frequencies being thrown at it, acts a lot like a lowpass filter. I mean most guitar amp sim plugins are exactly that: something that models a circuit distorting, and then something that models the response of the speaker/cab/etc. Plug in direct to a distorting guitar amp sometime and you will see that it sounds like ass without a speaker to smooth it out. A lot of distortion plugins are just clippers with a lowpass filter on the back.

 

So the question is, what distortion is good? Personally, I think most distortion is pretty cool, and it's asinine to suggest that electronic music, of all things, should be free of it. Electronic music is almost exclusively MADE of distortion, from resonant filters all the way up to 12-bit samples and clunky compressed 909 hats. I mean fine if you are recording the new James Taylor album then okay, strive for a clean signal, but otherwise it's a fucking waste of time and will actually make your music SOUND WORSE.

 

That said there's always the possibility of overdoing it, which is what we were discussing before I derailed this thread.

Here's a Monolake interview with Rashad from D&M who in case you were wondering is basically the man

[Ragnar]

Ah, the 'wall of sound' mastering. I think it works well for music that is supposed to be loud and forceful, like the typical metal or hard rock fare, but music with more subtleties definitely won't sound good mastered like that.

 

Then again, I don't know what I'm talking about.

 

is this really what 'wall of sound' is if so I like how the loudness trend was started by a psycho killer

[Ragnar]

do you guys like my production/loudness level lol

[cruising for burgers]

omg there's no dynamics =0

[Guest Scrambled Ears]

Ah, the 'wall of sound' mastering. I think it works well for music that is supposed to be loud and forceful, like the typical metal or hard rock fare, but music with more subtleties definitely won't sound good mastered like that.

 

Then again, I don't know what I'm talking about.

 

is this really what 'wall of sound' is if so I like how the loudness trend was started by a psycho killer

It's not and I think it's a little misleading that they say its a related topic

http://en.wikipedia.org/wiki/Wall_of_sound

 

as Braintree said the loudness trend was mainly developed by a bunch of radio business experts who figure:

 

1. Loud music releases more endorphins

2. Endorphin rush gets people to buy music (lol)

3. Make music on the radio as loud as possible

 

proof: Loudest songs sell more CDs

 

Incidentally it's very similar to the way they make commercials like 6dB louder than the TV program they're on. Like pay attention you moron! Buy our shit! So if anything, think about this before you obliterate your dynamics, who/what are you catering to?

[Guest sirch]

holy shit, shut the fuck up you infant, sirch.

 

lol. bite me big boy! :P

[Guest Calx Sherbet]

It's not something like Clark's crazy sound auteur idea of using compression & limiting as a type of dynamic instruments in the song, it's straight up ignorance on the part of sound engineers and the producers who just say "yeah that sounds heavy, bro, awesome"). Why doesn't anybody realize that a DJ or a home listener CAN EASILY JUST TURN THE VOLUME KNOB UP IF THEY WANT IT LOUDER without the track having to be totally butchered before it even gets to the consumer?

 

This is why all the DJs and clubgoers are deaf anymore.

 

It's so it's louder on the radio. The loudest song usually gets more attention and sells more records.

 

the radio edits should just be that way for sales sake. just not the album

[Braintree]

Anyway in the interest of stimulating discussion here are two "provocative" "opinions":

 

1. Most distortion discards information, therefore most distortion reduces the uniqueness of audio. Therefore, there are not many different kinds of "distortion", in a digital sense.

 

2. There is a continuum between "clipping" and "compression"—they're not two distinct things. Consider a very simple lookahead limiter with a hard knee being fed a signal that is entirely above the threshold. In what way is that different from clipping?

 

1. There's soft distortion and hard distortion in varying degrees. Why does the guitar distortion in a Slayer track sound different than a Pantera track?

 

2. If it's above the threshold, it's not compressing. Clipping/peaking is when you cross over 0dB...

Hi.

 

1. Okay, maybe I can try and get at this a slightly different way. Say you take a pure sine wave, clip it super hard, and then compare it to a square wave at full scale. How different are those two things?

 

Suppose you do the same thing, but then normalize both waves back down to relatively quiet levels (say, -30 db), and then lowpass them at like 2k. How different are they now?

 

I'm not saying I know the answers to these things, I'm just asking.

 

2. By above I mean louder than the threshold. Suppose you feed a primitive peak limiter audio that is entirely above its set threshold. How is what it's doing different from what clipping the signal at that threshold would do?

 

Jeesh, I didn't get enough sleep last night. *If it passes over the threshold it's compressing.

 

1. a. Sine & square waves have different partials. A simple square wave is 50% harmonic/non-harmonic. There's a matter of phasing that makes them sound different.

1. b. The sine wave is still distorted, and the square wave still has a more buzzy sound considering its non harmonic partials.

 

2. A limiter limits based on a ratio. Without the ratio, I can't tell you how it will behave since that's the thing that dictates how many decibels it cuts, and at what point it cuts [0dB, -1dB, -2dB, etc]. It's designed to reduce audio that's set over the threshold, and snags the audio before it goes through the output to clip.

 

A 2:1 ratio means for every 2dB in, there's on 1dB sent out. 10:1 means for every 10dB in, there's 1db sent out, ∞:1 means anything that goes in, only produces 1dB out, and so on.

Yo.

 

1. I'm a little foggy but I'm pretty sure a square wave has only odd, harmonic partials. It's a saw wave that has both odd and even partials. They are all harmonic (as in, whole number multiples of the fundamental) as far as I know.

 

Yeah, that sounds like half the partials to me:

http://en.wikipedia.org/wiki/Square_wave

 

b. I was kind of being a dick before by not really saying what I mean. Which is: if you clip a sine wave hard enough (imagine amplifying it by 999999 db), you get something close to (and eventually, you get exactly) a square wave. Not at all coincidentally, you have just added a ton of distortion, which happens to correspond with what you'd get by just starting with a square wave in the first place, only at twice the volume. It's one and the same process, in other words.

 

You don't get a square wave. As aforementioned, the sine wave doesn't hold the same partials as a square wave. The wave form just starts to look like one since it's brickwalled.

 

2. Let's stick with a simple hard limiter like L1. The whole purpose of that plugin is to prevent any signal at all from going over 0 dbfs (or whatever you set the ceiling at). So, it has an infinite ratio. ALL A LOOKAHEAD LIMITER DOES is look ahead to watch for samples that are going to pass above the threshold and "ramps in" the gain reduction to meet them. Then, it ramps out the gain reduction according to the release value. If you could somehow set a limiter plugin to have zero attack (that is, no lookahead) and zero release (that is, an instantaneous release) then all of a sudden you are clipping, not limiting.

 

So all "limiting" means (again, in the sense of a very basic algorithm) is clipping the signal, but with some ramp time in and out of the gain reduction to mitigate the WORST sounding parts of the clip. So there's a continuum between clipping and conservative limiting where, as you get less aggressive, you distort the music less (ie. you give the limiter more time to ramp in and out of its worst-sounding artifacts). The point is, these aren't two COMPLETELY DIFFERENT things. They are actually the same thing—that's why you hear things about L1 being distorted sounding and L2 being cleaner but way louder, etc, etc.

 

The audio is supposed to go to the limiter before it gets to the output, hence compressing before clipping. That also depends on if it's clipping at the channel or the board.

 

Also, the more you compress, the more you raise the noise floor and all of that shit at the bottom becomes more apparent. It's not distortion. Phasing maybe [which could be considered a type of distortion], but the whole point to a limiter is that you don't clip anything.

[ascdi]

Samples are samples. If a sine wave is distorted to "look" like a square wave, then it "is" a square wave. The squarer you get, the more harmonics you get. That is what the squareness of a wave is.

 

Anyway, never mind