Making hip hop beats fatter...

In my studio, I'm not a beat maker. I can, but they usually come out sounding less Dr. Dre and more Trent Reznor (thanks to my rock background). So most of my hip hop clients bring their beats in to me. They usually bring in a stereo mixdown of their beat, and often it's an mp3. If I can get a .wav of their beat, it's much better, because as an mp3 this trick can bring out more noisy artifacts if it's not encoded properly.

The trick is parallel compression. I usually do two things when I get a new beat in to the studio. One is to notch some bad frequencies out of the beat to clean it up, and the other is to parallel compress it. Let me explain.

Not everyone who makes beats is doing so in a great listening environment. This doesn't mean that their beats are bad, it just means that there may be frequencies that they didn't hear and didn't know to take out of their mix. Poorly treated rooms can have null's at certain frequencies, and this means that you will not hear that frequency because it cancels itself out of your listening environment due to bad acoustics. This is why acoustic treatment is so important, and if you can't afford treatment, I recommend mixing through flat headphones (NOT
Skull Candies or Beats By Dre's…although those can be great for referencing the mix, but I wouldn't make any critical listening decisions using them).

Knowing that no two rooms sound the same, the first thing I do is bring the stereo mix of the beat into a single track. I pull out an EQ, and start notching out bad frequencies. This is done by narrowing the Q on an eq band, and spiking it up toward the ceiling. It looks like this.

I then slide that eq band left and right while listening to the source audio. I'm listening for any sound that doesn't sound right, that resonates, or sounds just plain bad. Once I find it, I pull that frequency down.

Repeat this until there are no more bad sounding frequencies in the beat. Feel free to bypass the eq once you're done to hear the difference. You'll hear a lot more clarity in the mix.

The next thing I do is duplicate the track (I call this "cloning"). You should have two channels of the same stereo file instead of just one, and both channels now have the same eq settings.

I solo the cloned track and then bring in a compressor. The goal here is to over compress the cloned track so that it's pumping like crazy. Squash the hell out of it! Then un-solo it and turn it all the way down.

The only step that remains is to blend the two together. Level out the first track so it's not peaking, and is barely poking into the orange on the meters. Then slowly turn up the compressed track until it fattens the beat up just enough. Only your ears can tell you when is the right amount, but your eyes will trick you into hearing a difference when none is actually made. So I typically close my eyes when making ANY fader adjustments, including in this situation. This way, my ears have to tell me when the compressed track is loud enough, and I won't look at the number values on the faders and subconsciously decide it's right based on what my brain wants to see. Taking my eyes out of it leaves only my ears to make the judgement call, and isn't that what you're supposed to use after all? Trust your ears, not your eyes.

Once I've got the two tracks blended just right, I'll group the two tracks together so that if I adjust one fader, the other one moves with it to keep them balanced. I use pro tools, but I'm sure that there are other DAW's with a grouping function. If not, you may consider routing the output of the two tracks to a bus, and adjusting the bus instead.

Now it's time to record your flows on top of a much fatter beat. This really brings more out of the beat, just be careful not to over do it. And remember, use your ears, not your eyes, to make your decisions about levels.

(Brandon S. Hire is the head engineer at Skyline Sound Studios in Grove City, Ohio. Visit skylinesoundstudios.com for more information!)

Polishing A Turd...A Mixing Rescue Story!

Mythbusters proved that it is indeed possible to polish a turd.


While it's not ideal to have to polish a turd as an audio engineer, knowing how can help save a recording and make the clients LOVE you! Case in point...

A band brought in a recording and asked me to try to give it a better mix. They recorded it in a church, the soundman for the church did the tracking, and there were sonic issues all over the place. The lead guitar was a stereo D/I out of a distortion pedal, yet the rhythm guitar was single tracked in mono. The bass was full of mess-ups through the whole song, and they were so bad that they literally were no longer following the song at one point. It turns out that the band rehearsed the song all of two times before recording it. Even the drums sounded like complete garbage, especially the kick. The drummer has a soft hit with his left foot, so when the left foot struck the bass drum on the double bass pedal, that hit would get lost and there would often not even be a transient (making drum replacement pretty difficult). The vocals were off key something fierce, and the entire mix sounded like one of my very first home recordings from over a decade ago.

So let's address the issues I was presented with one at a time. First, the vocals. They were pretty pitchy, but that's nothing that some light use of Melodyne can't fix. Of course, if automatic pitch correction isn't an option, then the audiosuite pitch correction plugin that comes with Pro Tools can do the job one note at a time. While not ideal, I've done some great things using that plug in.

So now that the vocals are in tune, let's look at the drums. They were poorly mic'd, so I opted to just replace everything other than the cymbals. The cymbals were mono mic'd with a single overhead near the center. Not a bad approach, I've done it before. But it would appear that they mic wasn't placed properly, as some of the cymbals just sounded bad. Nothing I can do about that though, so let's address the individual drums.

The kick drum was the first to be addressed. I used the Tab-To-Transient feature in Pro Tools to find the transients and trigger Steven Slate Drums at the transient. The problem is the way the drummer plays, considering that his left foot hit's very lightly. Since we're talking about metal music here, that is a problem. It's an even bigger problem that he hit the drum so lightly with his left that the transients would disappear in the huge transient created by the right foot hitting the same drum during the fast double bass parts. So this made tab-to-transient drum replacement a problem, and damn near impossible. I had to listen to the drum track in slow motion to determine where to trigger the samples.

The entire drum kit was sample replaced. There was just no easy way to save those tracks. But now that the drums had been sample replaced and the vocals had been tuned, it was time to move on to bass.

This was tough. I had to anticipate what the bassist was trying to do and recreate it using those same notes as they had been played earlier in the song. But the song wasn't recorded to a click track, so this meant I had to employ heavy use of elastic audio to get the bass to fit correctly. Once I was done with the bass, I had essentially reconstructed the entire track.

I did the same to the rhythm guitar, as I wanted a big full metal guitar sound in the mix, but only had a mono single guitar track! So I did some voodoo to reconstruct a new guitar track from the ground up that would act as a double. This included more elastic audio. I basically took the guitar track and cloned it in the DAW. Then I copied the second chorus over the first and the first chorus over the second. This gave me two distinct performances of the chorus on guitar and enabled me to double the chorus even though it was single tracked! I followed suit with the verses and the other parts of the song until I had reconstructed the entire song on guitar from the ground up!

Last was the lead solo, which was recorded through dual outputs of a distortion pedal onto two channels. I tossed one of the channels, as they were both identical and both sounded like garbage as is. Then I re-amped the remaining take to give the lead solo some real amp tone, and cranked the bejesus out of the amp in the live room. IT WORKED! The guitar solo sounded really good as opposed to nasty and digital!

Once it was mixed, and I delivered the mix, the band was ecstatic at how big of a difference I had made! It was also a fun project to help me test my boundaries and problem solving skills. But in all honesty, it's better to get it right to begin with. Surgery on this song took HOURS, and retracking the song correctly would've honestly taken less time than reconstructing it from the ground up. But the band wanted me to save what they brought me, so that's what I did. But the moral of the story is, and should be, to find a way to make it work. You may have to put some time in, but solutions are out there to get it done!

(Brandon S. Hire is the head engineer at Skyline Sound Studios in Grove City, Ohio. Visit skylinesoundstudios.com for more information!)