MistaMayne
10-28-2008, 11:41 AM
I've been quiet in music for a while, because I really want to get further into it. So I've decided to analyze random instrumentals from producers Eminem, Timbaland, KanYe West, Dr. Dre, DJ Premier, Marly Marl, The Neptunes, The 45 King, Mr. Porter and Mobb Deep. And I saw that they all have a few differences that none of the underground producers have.
The volumes. The way the drums, snares, melodies and vocals are mixed down, varies greatly from producer to producer, but thanks to Cool Edit.. (and a shitload of time and looping) I've managed to get the basic rules applied. I'm going to go back to basics in all of my mixing to get more towards the industry standard. But I'm sharing this with you guys to give you more of a start as well.
This is a snippet from Eminem's "Soldier" instrumental. See the image below. I've selected the bassdrum kick, to get a basic leveling done. With every beat I've looked into, I've noticed that the bass-drum is always set to -3 DB in the Cool Edit window.
Picture of Eminem's "Soldier" instrumental, with the bass drum looped
The snare varies from producer. Eminem varies between -3 DB to -9 DB while Timbaland and Dr. Dre keep the snare between -2 DB and -6 DB on average.
Eminem's "Soldier" instrumental, with the snare drum looped
Dr. Dre's "Forgot About Dre" instrumental (ripped from YouTube, with the snare drum looped
The melody is where the major difference is between the industry producers and the underground producers. Let me show you what I mean.
I've used the "Soldier" beat yet again, and inserted Shadowville's "Mosters" beat and VTZ's "Sins" beat. VTZ and Shadowville's beats volumes have been equaled with the Eminem instrumental. Meaning that I set it that all the drums in the three beats are at the same volume.
Eminem's "Soldier" beat, melody is around -10,5 DB
VTZ's "Sins" beat, melody is around -6 DB
Shadowville's "Monsters" beat, melody is around -3 DB
I know different producers have different styles. But I'm merely stating that the mainstream producers all have the drumkicks at -3 DB, the snares vary from time to time. But the melodies are much softer.
I've tooken about 10 main stream producers, took 3 beats from each of them and averaged the volumes. To get a bit of a more accurate idea of the volumes.
The drums are always at -3 DB. I've not seen 1 exception to that. The snaredrum is around the -4,6 DB on average. The melody is around -10,5 DB on average.
The conclusion we can draw from this, is that the drums are relatively more dominant in the mainstream instrumentals than with the instrumentals found on let's say Soundclick.com.
This has an effect on the vocals as well. I've gotten my hands on Nelly Furtado's "Maneater" a cappella and on Eminem's "Cleaning Out My Closet" a cappella, as well as the "One Mic" (Nas) and the "Lapdance" (N.E.R.D) a cappella. I've chosen those four, because they're on 4 different styles. And just for the heck of it we'll use Gwen Steffani's "What Are You Waiting For" as well.
I'll insert them all into Cool Edit Pro. Please keep in mind, that these a cappella's were taken raw from the albums. So they are at the volumes that they were mixed at with.
We'll start Nelly Furtado's vocal takes. Upon listening to it, just with the plain ear, I'm noticing a few things. There's no reverb on the main vocals (Nelly) but there's an echo on Timbaland's addlibs. On the pre-chorus, Nelly's voice in triple layered. Plus underneath it, there's another layer running, with just a harmony. These extra layers all continue in the chorus, but stop at the verse. The vocals also get more upfront at the chorus, bridge and after the 8th bar of every verse. (Pre-chorus).
Then we'll listen to Eminem's "Cleaning Out My Closet" takes. Upon listening to this, I'm hearing some more things. Eminem always uses a triple layer on his verses. There's one running with about a note lower than his dominant voice. This one is calm and goes solidly with the beat. The dominant one has a lot more emotion. Then there's the adlibs, which are a tad bit lower in tone. The chorus has 5 layers running. This is more obviously heard on "Loose Yourself". The same 3 from the verses, then he goes with the melody. Humming, so to say. And he adds another add-lib track. The vocals remain at the same place. They don't become more upfront or fading away. Again, there's no reverb, no echo, no polishing what-so-ever. And the overdubs are very slightly off-beat. Think of a delay of a few miliseconds. There's very little bass levels in there, but more frequencies in the middle and higher levels.
Going on to NaS's "One Mic" a cappella. Of course these get louder at the verses, due to the whole concept of the song. So we'll disregard that. Apart from that, we hear his vocals doubled. Like Eminem, he has one vocal take going perfectly with the take, and completely overdubbing it. Unlike Eminem, both takes are at the same pitch. There are no overdubs, though. There's a slight reverb and a small echo on the chorus. It is noticable that NaS's vocals are pretty much having the bass levels filtered away.
That brings us to the fourth acapella.. N.E.R.D's "Lapdance". At the intro it's already hearable that there's an echo, which only has the higher frequencies bouncing. No middle or bass levels. Pharrell also doubles his vocals, on the rapping parts, though. There's very little pitch difference between the two layers. And they do go together perfectly. There's not a milisecond difference between them. The reverb is gone on the rapping. But it's present on the singing parts. When the lady sings the chorus, you hear another harmony track going on. It's vocoded, but it's still hearable that it's somebody singing a basic melody. The chorus is also overdubbed at some parts.
The fifth acapella, Gwen's "What Are You Waiting For" shows more of the way mixers handle singles. The verses are single in this. There's no overdubs. But there are addlibs. When the chorus starts, you hear added harmony tracks with it. The bridge has Gwen's vocals quadruppled. There's a very small reverb on the vocals.
What are the guidelines/rules in the Industry Standard when mixing down?
- Bass drums are always at -3 DB
- Snare drums are on average around -4,6 DB
- Melody is on average around -10,5 DB
- Vocals on the verses have the mimum levels are on average around -10 to -8 DB
- Vocals on the chorusses are averaged around -9 to -7 DB
- Melodies on the hooks are 0,5 DB louder.
With rap vocals
- Do the first take without emotion to the instrumental. Have your voice at any preference at pitch.
- Overdub the entire verse with emotion
- Overdub the verses with only the words you want to empathize.
- Hum with the melody on the hooks
- Overdub again on the hooks, over all the other tracks.
When mixing down, make sure that all the takes blend together nicely. Keep the overdubs around 25% of the volume on your main vocal takes.
With sung vocals
- Do the first take with emotion and very strong harmony control.
- Possibly overdub, depending on personal preference.
- Add addlibs.
- Use another take for extra harmonies on the chorusses, pre-chorusses and bridged.
When mixing down, keep the added harmonies around half the volume of the main track. That way it dives under the melody of the instrumental, but stays effective.
Reverbs, echo's and all other effects are personal preference, they are NOT a must.
I don't know the settings for the compression or anything, though.
Here's an added bonus that I salvaged from somewhere. It's the basic guideline for applying EQ. And how to use it.
Don't just stupidly apply these EQ settings. These are the guidelines. From these settings, tweak around untill you get the perfect sound you want. But notice how the bass and treble levels roll off. And where the boosts are. The boosts are to add warmth. The "lead" parts are used for the main takes. The "back-up" parts for the addlibs and overdubs.
EQ Settings
If I find out more from the industry way of mixing stuff down, I'll be sure to post it.
The volumes. The way the drums, snares, melodies and vocals are mixed down, varies greatly from producer to producer, but thanks to Cool Edit.. (and a shitload of time and looping) I've managed to get the basic rules applied. I'm going to go back to basics in all of my mixing to get more towards the industry standard. But I'm sharing this with you guys to give you more of a start as well.
This is a snippet from Eminem's "Soldier" instrumental. See the image below. I've selected the bassdrum kick, to get a basic leveling done. With every beat I've looked into, I've noticed that the bass-drum is always set to -3 DB in the Cool Edit window.
Picture of Eminem's "Soldier" instrumental, with the bass drum looped
The snare varies from producer. Eminem varies between -3 DB to -9 DB while Timbaland and Dr. Dre keep the snare between -2 DB and -6 DB on average.
Eminem's "Soldier" instrumental, with the snare drum looped
Dr. Dre's "Forgot About Dre" instrumental (ripped from YouTube, with the snare drum looped
The melody is where the major difference is between the industry producers and the underground producers. Let me show you what I mean.
I've used the "Soldier" beat yet again, and inserted Shadowville's "Mosters" beat and VTZ's "Sins" beat. VTZ and Shadowville's beats volumes have been equaled with the Eminem instrumental. Meaning that I set it that all the drums in the three beats are at the same volume.
Eminem's "Soldier" beat, melody is around -10,5 DB
VTZ's "Sins" beat, melody is around -6 DB
Shadowville's "Monsters" beat, melody is around -3 DB
I know different producers have different styles. But I'm merely stating that the mainstream producers all have the drumkicks at -3 DB, the snares vary from time to time. But the melodies are much softer.
I've tooken about 10 main stream producers, took 3 beats from each of them and averaged the volumes. To get a bit of a more accurate idea of the volumes.
The drums are always at -3 DB. I've not seen 1 exception to that. The snaredrum is around the -4,6 DB on average. The melody is around -10,5 DB on average.
The conclusion we can draw from this, is that the drums are relatively more dominant in the mainstream instrumentals than with the instrumentals found on let's say Soundclick.com.
This has an effect on the vocals as well. I've gotten my hands on Nelly Furtado's "Maneater" a cappella and on Eminem's "Cleaning Out My Closet" a cappella, as well as the "One Mic" (Nas) and the "Lapdance" (N.E.R.D) a cappella. I've chosen those four, because they're on 4 different styles. And just for the heck of it we'll use Gwen Steffani's "What Are You Waiting For" as well.
I'll insert them all into Cool Edit Pro. Please keep in mind, that these a cappella's were taken raw from the albums. So they are at the volumes that they were mixed at with.
We'll start Nelly Furtado's vocal takes. Upon listening to it, just with the plain ear, I'm noticing a few things. There's no reverb on the main vocals (Nelly) but there's an echo on Timbaland's addlibs. On the pre-chorus, Nelly's voice in triple layered. Plus underneath it, there's another layer running, with just a harmony. These extra layers all continue in the chorus, but stop at the verse. The vocals also get more upfront at the chorus, bridge and after the 8th bar of every verse. (Pre-chorus).
Then we'll listen to Eminem's "Cleaning Out My Closet" takes. Upon listening to this, I'm hearing some more things. Eminem always uses a triple layer on his verses. There's one running with about a note lower than his dominant voice. This one is calm and goes solidly with the beat. The dominant one has a lot more emotion. Then there's the adlibs, which are a tad bit lower in tone. The chorus has 5 layers running. This is more obviously heard on "Loose Yourself". The same 3 from the verses, then he goes with the melody. Humming, so to say. And he adds another add-lib track. The vocals remain at the same place. They don't become more upfront or fading away. Again, there's no reverb, no echo, no polishing what-so-ever. And the overdubs are very slightly off-beat. Think of a delay of a few miliseconds. There's very little bass levels in there, but more frequencies in the middle and higher levels.
Going on to NaS's "One Mic" a cappella. Of course these get louder at the verses, due to the whole concept of the song. So we'll disregard that. Apart from that, we hear his vocals doubled. Like Eminem, he has one vocal take going perfectly with the take, and completely overdubbing it. Unlike Eminem, both takes are at the same pitch. There are no overdubs, though. There's a slight reverb and a small echo on the chorus. It is noticable that NaS's vocals are pretty much having the bass levels filtered away.
That brings us to the fourth acapella.. N.E.R.D's "Lapdance". At the intro it's already hearable that there's an echo, which only has the higher frequencies bouncing. No middle or bass levels. Pharrell also doubles his vocals, on the rapping parts, though. There's very little pitch difference between the two layers. And they do go together perfectly. There's not a milisecond difference between them. The reverb is gone on the rapping. But it's present on the singing parts. When the lady sings the chorus, you hear another harmony track going on. It's vocoded, but it's still hearable that it's somebody singing a basic melody. The chorus is also overdubbed at some parts.
The fifth acapella, Gwen's "What Are You Waiting For" shows more of the way mixers handle singles. The verses are single in this. There's no overdubs. But there are addlibs. When the chorus starts, you hear added harmony tracks with it. The bridge has Gwen's vocals quadruppled. There's a very small reverb on the vocals.
What are the guidelines/rules in the Industry Standard when mixing down?
- Bass drums are always at -3 DB
- Snare drums are on average around -4,6 DB
- Melody is on average around -10,5 DB
- Vocals on the verses have the mimum levels are on average around -10 to -8 DB
- Vocals on the chorusses are averaged around -9 to -7 DB
- Melodies on the hooks are 0,5 DB louder.
With rap vocals
- Do the first take without emotion to the instrumental. Have your voice at any preference at pitch.
- Overdub the entire verse with emotion
- Overdub the verses with only the words you want to empathize.
- Hum with the melody on the hooks
- Overdub again on the hooks, over all the other tracks.
When mixing down, make sure that all the takes blend together nicely. Keep the overdubs around 25% of the volume on your main vocal takes.
With sung vocals
- Do the first take with emotion and very strong harmony control.
- Possibly overdub, depending on personal preference.
- Add addlibs.
- Use another take for extra harmonies on the chorusses, pre-chorusses and bridged.
When mixing down, keep the added harmonies around half the volume of the main track. That way it dives under the melody of the instrumental, but stays effective.
Reverbs, echo's and all other effects are personal preference, they are NOT a must.
I don't know the settings for the compression or anything, though.
Here's an added bonus that I salvaged from somewhere. It's the basic guideline for applying EQ. And how to use it.
Don't just stupidly apply these EQ settings. These are the guidelines. From these settings, tweak around untill you get the perfect sound you want. But notice how the bass and treble levels roll off. And where the boosts are. The boosts are to add warmth. The "lead" parts are used for the main takes. The "back-up" parts for the addlibs and overdubs.
EQ Settings
If I find out more from the industry way of mixing stuff down, I'll be sure to post it.