Any mastering engineers - headroom?

[samkablaam]

I know a lot of people say that it's best to leave 3-6 dbs of headroom when sending stuff of to be cut, and the reason sort of makes sense to me.

But I've started to convince myself it doesn't make sense. Correct me whenever I go wrong, but surely if you're working digitally, the mastering engineers will be rearranging all these bits anyway and can deal with a 24 bit file that doesn't have many spare bits left.
And if you're working analogue-ly(?) you'll be doing you're own dac-ing anyway.

A lot of producers I work with send masters off with almost no headroom left (like 0.1 dB) and the mastering guys never complain, and these are seasoned mastering companies, with really serious client lists.

I've sent electronic music off with loads of headroom and haven't noticed an improvement in mastering compared to sending stuff off with no headroom.

Is it something to do with the way daws sum stuff if you touch a fader? Or is it just trying to make sure the stuff sent to mastering hasn't got like intersample clips? Or?

safe

[jaimelee]

The gain structuring comes to mind, if done incorrectly the engineers need the head room to fix that. (As to bring out the full potential of track)
I'm not 100% on your question but that's what comes to mind mate.


Also 0.01% headroom!?!?!? I'd be too afraid of distorting when mastering that :O

[the dub lemon]

-6db leaves the mastering engineer some thing to work with and if you're working at 24bit there's no need to run it as hot as possible. However if you send a mastering engineer a track peaking at 0db and they need some headroom they'll just turn it down, simples.

[samkablaam]

The gain structuring comes to mind, if done incorrectly the engineers need the head room to fix that. (As to bring out the full potential of track)
I'm not 100% on your question but that's what comes to mind mate.


Also 0.01% headroom!?!?!? I'd be too afraid of distorting when mastering that :O

mastering doesn't fix gain structuring. thats mix shit.

but what im asking is, some mastering engineers dont ask for any headroom, and it never comes back distorted. so why would you be afraid of distorting? im not saying anyones wrong, im just asking for you're reason for being afraid.

-6db leaves the mastering engineer some thing to work with and if you're working at 24bit there's no need to run it as hot as possible. However if you send a mastering engineer a track peaking at 0db and they need some headroom they'll just turn it down, simples.

thats what im asking about, why do some mastering places ask for 6dBs of headroom if its so easy to turn it down. :S

[Electric_Head]

This is all assumption but I would assume it is related to your system and creating a mix that will fundamentally sound good on most systems because of the gain structuring.
Rather than sounding good on one system but bad on another.

[samkablaam]

what? how is headroom going to make your mix sound good on most systems?

dont know if ive explained my question properly.

im just wondering what the point of leaving headroom is when tracks are sent to mastering. some guys ask for it and some dont, and tbh, the most experienced mastering engineer i heard of, doesnt.

[grooki]

Check out the Gain Structure and Mixing Sticky:

http://www.dubstepforum.com/viewtopic.php?f=8&t=74832

It has a whole pile of questions, including ones likes yours, answered by Macc, who is a well known and respected mastering engineer.

[jobbanaught]

as far as i understand headroom refers to the rms level of the track. Leaving -6db means that your track on average is -6db, peaks can be up to -0.1db still. in other words leaving headroom means not compressing your track to fuckery before having it mastered. correct me if im wrong though....

[AxeD]

I they can turn it down that easily then I see no reason to send it off with headroom.

[Sharmaji]

it's less about having "space" to work with, and more about removing the potential for intersample peaks and the resulting audible distortion.

which sounds like cat piss.

quite often, when things are pushing up at -1, -2, there will be digital overs that happen so fast, the meters don't recognize them (try it-- take a transient track that peaks at around -3 or so, and put a compressor on it w/ super-fast attack, but the threshold all the way up at zero. why does it respond?).

For some mixes, this isn't an issue-- there's a very dense high end already, etc. for other stuff, those fast peaks that occur because of the calculus of sampling can build up and turn into a white noise-ish haze that runs thru the track. in those cases, you're stuck with it. EQ and dither can help, but in that case, you're talking about fixing and masking a problem, rather than letting the great sounding parts of the track shine.

mixing down quietly generally gets you out of that danger zone.

the super-loud mixes i get in to master are often from younger producers who aren't mixing in the best of environments, which is totally fine by me. Only problem is that quite often, there's unintended bits of distortion and modulation-- especially in the low end-- that could have been avoided had everything just been quieter. At that point, all i can do is turn the material down and try to do some corrective work to hopefully find elements that i can accentuate. now we're talking about sidechaining various frequencies, doing parallel compression, maybe using mid-side to mask the peaks in the center while making the sides bright, etc...

contrast that to something like Moldy's "Embrace the Filth," which was one of the best raw mixdown's i've ever worked on (and probably peaked at around -6 or 5). super-clear highs, chest rattling bass, and an intense, full vibe through the whole thing. Essentially all I had to do was turn it up, and do a tiny tiny bit of EQ on the midrange.

[samkablaam]

seen, thats what was starting to think. so, the guys that dont ask for it are just trusting the mix engineer to be careful of intersample clips and cat piss.

thanks for clearing that up.

[Sparxy]

it's less about having "space" to work with, and more about removing the potential for intersample peaks and the resulting audible distortion.

which sounds like cat piss.

quite often, when things are pushing up at -1, -2, there will be digital overs that happen so fast, the meters don't recognize them (try it-- take a transient track that peaks at around -3 or so, and put a compressor on it w/ super-fast attack, but the threshold all the way up at zero. why does it respond?).

For some mixes, this isn't an issue-- there's a very dense high end already, etc. for other stuff, those fast peaks that occur because of the calculus of sampling can build up and turn into a white noise-ish haze that runs thru the track. in those cases, you're stuck with it. EQ and dither can help, but in that case, you're talking about fixing and masking a problem, rather than letting the great sounding parts of the track shine.

mixing down quietly generally gets you out of that danger zone.

the super-loud mixes i get in to master are often from younger producers who aren't mixing in the best of environments, which is totally fine by me. Only problem is that quite often, there's unintended bits of distortion and modulation-- especially in the low end-- that could have been avoided had everything just been quieter. At that point, all i can do is turn the material down and try to do some corrective work to hopefully find elements that i can accentuate. now we're talking about sidechaining various frequencies, doing parallel compression, maybe using mid-side to mask the peaks in the center while making the sides bright, etc...

contrast that to something like Moldy's "Embrace the Filth," which was one of the best raw mixdown's i've ever worked on (and probably peaked at around -6 or 5). super-clear highs, chest rattling bass, and an intense, full vibe through the whole thing. Essentially all I had to do was turn it up, and do a tiny tiny bit of EQ on the midrange.

Great post. I was reading this thread thinking the same as the OP. I definitely learned something there!

[grooki]

Oh yeah and Sharmaji Sorry Dave!

[OlzaMK]

Great post Sharmaji, very informative.

[Electric_Head]

I think Sharmaji explained it a hell of a lot better than I did

I understand it better now as well.

score

[samkablaam]

ha. well he explained it for a start.

[Depone]

Yeah dave has it.

The other thing i can think of, is that its a pratical help. If the track is peaking at -0.1db, and your first port of call is some positive db eq. You are already going over the digital db threshold. And this doesnt effect the playback, but at the last stage with limiting, you have lesser control over the threshold as its already driving hard into it.

I know you can turn the raw wave down, but its better to have started with the more headroom from the initial bounce.

[samkablaam]

Yeah dave has it.

The other thing i can think of, is that its a pratical help. If the track is peaking at -0.1db, and your first port of call is some positive db eq. You are already going over the digital db threshold. And this doesnt effect the playback, but at the last stage with limiting, you have lesser control over the threshold as its already driving hard into it.

I know you can turn the raw wave down, but its better to have started with the more headroom from the initial bounce.

yeah, thats the thing i was confused about. the masters we normally send off from the studio have little to no headroom and mastering engineers just turn it down. and then most mastering engineers on here say leave the headroom to begin with. so i was just wondering what the difference was.

macc + sharm + depone = answer.

[paravrais]

macc + sharm + depone

I'd love to hear that collab XD

[zerbaman]

I do have a question on this.
In reference to the first post:
Can't they just turn the volume down on the 24 bit .WAV if necessary?