|
How to create more headroom in your mix (with EQ)? (pg. 5)
|
View this Thread in Original format
| Raphie |
| quote: | Originally posted by Subtle
I meant more like flat as little peaks, which means you gain more volume without loosing dynamics.
here is a picture of what i consider a bad and a good mix.
BAD MIX
GOOD MIX
As you can see one of them is spiking like , and is a tough track to master at a decent volume, while the other is evened out and is easy mastered to a decent sound level. |
I think most engineers will not agree. your "good" mix has no dynamics left. those aren't spikes, they are transients. your "bad" mix is actually the better of the 2... start reading about the loudness war... |
|
|
| Subtle |
| quote: | Originally posted by Raphie
I think most engineers will not agree. your "good" mix has no dynamics left. those aren't spikes, they are transients. your "bad" mix is actually the better of the 2... start reading about the loudness war... |
Yeah i know about the loudness war, and if i see spikes in my tracks i keep them, because i know limiting them will make the sound worse.
I realise posting pictures was the worst idea ever.
What I mean about bad mix is that the volume cannot be raised alot without loosing dynamics, while the good mix can have its volume raised without it sounding different than in the sequencer. Which means lesser processing to get an even volume overall, thus a more correct mixdown.
The bad mix sounds good, it just cannot sound as high in volume as other tracks going with it. |
|
|
| Raphie |
| quote: | Originally posted by Subtle
Yeah i know about the loudness war, and if i see spikes in my tracks i keep them, because i know limiting them will make the sound worse.
I realise posting pictures was the worst idea ever.
What I mean about bad mix is that the volume cannot be raised alot without loosing dynamics, while the good mix can have its volume raised without it sounding different than in the sequencer. Which means lesser processing to get an even volume overall, thus a more correct mixdown.
The bad mix sounds good, it just cannot sound as high in volume as other tracks going with it. |
hence you need to cut and control your transients on track level in order to optimize headroom and perceived volume |
|
|
| Subtle |
| quote: | Originally posted by Raphie
hence you need to cut and control your transients on track level in order to optimize headroom and perceived volume |
Yes absolutely. That is why it is a bad mix, it is an old mix of mine.
I have noticed that my mixes are just getting flatter because i get better at mixing. But only by following the ears and nothing else, and i need the mixes to not spike, as i do not master my tracks afterwards, i simply compress once at 3 db threshold, slap a limiter on them and there they go.
And before anyone says "no no dude flat is bad" i have to also say that my kicks usually peaks between -3b and -6db which means room for transients and dynamics. |
|
|
| cronodevir |
The whole notion that a mix needs to be loud [peaking at 0db] is the loudness war in action. The concept is basically "What is the best way to get maximal loudness and not distort?" And that is the approach DJ RANN is coming from I believe. This is accomplished by compressing audio all to , flattening it out, and pushing it to the 0db wall. [Which, Subtle is what your "good" mix looks like, of course ive not heard it.]
As for the noise floor, he is referring to my maxing at around -5 to -8db ...I try not to let anything go over those. |
|
|
| Omega_Blue |
| quote: | Originally posted by Subtle
Now, this is what i dont understand fully what has to do with the master channel.
How can there be a difference between having having the master channel at higher volume and the individual channels set lower, as opposed to having the individual channels set higher and the master channel lower.
The output should be exactly the same, no ? |
i might've misunderstood you and dj rann but this is what i get from it (and how i've been doing it myself). you don't necessarily turn every individual channel down, only the problematic ones
you have a track but one or two instruments are clipping. you would turn down only the clipping instruments instead of turning down the whole track via the master. relative to the rest of the mix, the clipping instruments are too loud. turn down everything and the clipping instruments are no longer clipping, but still too loud when compared to the rest of the mix, which now has lost dynamic range since you cut out its maximum possibility of loudness by decreasing the master volume.
it's like cutting down the whole tree because two branches are in the way of your house. why wouldn't you just trim the branches instead? |
|
|
| Subtle |
| quote: | Originally posted by cronodevir
As for the noise floor, he is referring to my maxing at around -5 to -8db ...I try not to let anything go over those. |
Okey, i see.. well that makes perfect sense, you should of course render with the MAX as close to 0 db as possible.
The thing with the loudness war is that you compress/limit so you LOOSE the transients, if your mix is maxing at 0db you arent loosing any transients at all (hence it got nothing to do with loudness war having the peak close to 0db), its when you limit/compress that you are taking transients away.
The loudness war is when you compress everything so hard that the waveform is just huge brick. |
|
|
| Subtle |
| quote: | Originally posted by Omega_Blue
i might've misunderstood you and dj rann but this is what i get from it (and how i've been doing it myself). you don't necessarily turn every individual channel down, only the problematic ones
|
Yeah i understand that, but in my case i does not make a difference. I use my ears to spot the problematic channels, and when i render a Test mixdown of the track i import it in a wave editor, and if I see spikes i can easily recall where the problem areas are with looking at the waveform, and then fixing them when i open the project again. I just seem to get a lesser and lesser amount of problem channels and that only by hearing (which is what mixing is ALL about) |
|
|
| DJ RANN |
| quote: | Originally posted by cronodevir
Except every piece of music or audio from a major studio sounds like . Highly compressed garbage. The things they teach are about making your audio stand out, not about the best possible quality.
So sure, you may have went too school and you are a professional engineer, but the fact is 99.9% of "professionally" processed music sounds horrible. And this garbage is brought into the audio when ever a track is sent to be mastered.
So somewhere along the line, something tells me these processes fail at some point. Wether the processes are bad, or they are implemented incorrectly, who knows.
All of that said, as I mentioned before ive never had any problems with my mix from anyone, and I have had a few tracks released on compilations.
Yes ive used limiters on channels before, but sometimes stuff still gets by. |
I agree that a lot of pro processed music is horrible but that's not a fault of engineering quality, it's down to the producers (in the real sense), composers and artists.
Bear in mind that engineers don't get to make artistic decisions about the content of music so don't blame them or their methods of providing the best possible recording, even if someone else s it up. It's like blaming the company that makes brushes for the current state of impressionist art.
The failure is lies with the creators of the content, not those that engineer it. Also, a couple of peaks actually don't matter on individual tracks, and sometimes when you clear the red on the mixer they don't appear the second time.
If you lower the master say 3db, then bounce, the track will have a dynamic range of 93dbfs instead of 96dbfs (16bit). So the noisefloor is introduced at that level of 93dbfs. SO when you turn it up to reach it's full potential 96dbfs (by 3 dbfs, which is twice as loud) you are effectively increasing the proportional noisefloor by a factor of 2 as well. |
|
|
| cronodevir |
| quote: | Originally posted by DJ RANN
I agree that a lot of pro processed music is horrible but that's not a fault of engineering quality, it's down to the producers (in the real sense), composers and artists.
Bear in mind that engineers don't get to make artistic decisions about the content of music so don't blame them or their methods of providing the best possible recording, even if someone else s it up. It's like blaming the company that makes brushes for the current state of impressionist art.
The failure is lies with the creators of the content, not those that engineer it. Also, a couple of peaks actually don't matter on individual tracks, and sometimes when you clear the red on the mixer they don't appear the second time.
If you lower the master say 3db, then bounce, the track will have a dynamic range of 93dbfs instead of 96dbfs (16bit). So the noisefloor is introduced at that level of 93dbfs. SO when you turn it up to reach it's full potential 96dbfs (by 3 dbfs, which is twice as loud) you are effectively increasing the proportional noisefloor by a factor of 2 as well. |
Sure, but one is more likely to notice you lost a few hairs on your arm...than that extra noise. If in fact the noise exists in the first place. The effect of moving the master slider down is so insignificant it doesn't matter. |
|
|
| Eldritch |
| quote: | Originally posted by DJ RANN
If you lower the master say 3db, then bounce, the track will have a dynamic range of 93dbfs instead of 96dbfs (16bit). So the noisefloor is introduced at that level of 93dbfs. SO when you turn it up to reach it's full potential 96dbfs (by 3 dbfs, which is twice as loud) you are effectively increasing the proportional noisefloor by a factor of 2 as well. |
I don't know what DAW you're using but the master fader is usually within the 32-bit floating point part of the audio engine. So the bit depth of the export is irrelevant. (Unless you put dithering in pre-fader inserts of course). It does not matter if you attenuate the master fader or the individual channels. Of course 32-bit floating point has its limits but you'd have to do some stupidly large gain increases/attenuation to compromise the quality of the signal.
Personally I don't touch the master fader. But that's just because it will make the fader values of the channels meaningless. |
|
|
| Raphie |
Maybe, but if you cut your kick and bass above 500hz, there will be nothing decent of your kick or bass left. Kicks also contain vital higher frequencies, as bass (except for subbass ofcourse). [/QUOTE]
no one said ABOVE 500hz, I mentioned AROUND 500hz. (where you can play with the Q :) |
|
|
|
|
