Hi my friend...very beautiful jobs !!!! Congratulations for you !!! The sound is fantastic...what did you use? What is the DAW do you use? Could you talk about the process for us? I really like to know... Once more congratulations for you and all musicians and the fabulous singer . the song is fabulous
hey guys! thanks so much for the kind words. after putting so much love into a track it's always nice to hear if people are enjoying the music!
I use cubase 6 for recording, editing and mixing. Because I use UAD plugs (32 bit only) I have to keep virtual instruments outside of the daw using Vienne Ensemble pro. The 32 bit version of cubase gets buggy afte about a 2.5gb ram size (3 is max for a 32 bit program). Vienna Ensemble almost never crashes (like native instruments always will inside cubase) so that allows me to use as many virtual instruments at 64 bit (like z3ta 2, omnisphere, trillian)
Another unique thing about my setup is the console which I sum 16-24 channels output through the radar 24 at 48k to the MCI-400 series console with upgraded opamps and back into cubase. This smooths everything over!. Nebula actually is amazing at this by using the console emulations and some of them actually sound much better than my mci console however many times I find myself wanting to use nebula instances on other things (like tape and reverbs) and I will save the processing power by using the external console instead of using the nebula consoles. however this is ONLY because of convenience and not at all because I think nebula doesn't sound as good... at 96k the alexb A*I console, vintage N**e AND the ssl9000 all sound amazing to me.
Because production and mixing involves so many things, I'll just focus on the key points that I feel are most important to getting this kind of sound if it interests anyone.
1.) SONGWRITING - The song is great to begin with (written by skylar grey who writes hooks for eminem, rihanna etc.) and is an excellent framework to build upon (8 bar verses, 8 bar chorus, middle 8). just textbook songwriting and excellent harmonic content with simple chord arrangements that makes it much easier to have a good production/mix already from the start.
2.) PLACEMENT - It is my personal belief that 90% of mixing at least is PLACING the elements in the production properly. That means that hardly anything hits the grid at the same time. I'll explain more for each element.
3.) ACOUSTIC SPACE - My engineer has taught me a lot about using space properly. He treated my studio tracking room with wood ceilings, and diffusion and baffles all over the place that are moveable. That way we can get the sound we want as it is being played and recorded. To me that's one of the biggest differences I've discovered in the last few years working in professional audio. Every element of this mix (except the pads) were live tracked / dubbed in my medium sized oak-wood live room. Without the live room I would have to work thousands of times harder to get the piano to sound right and the drums would always sound a tiny bit off to my ears even with excellent artificial reverbs.
4.) CONVERTERS - I've invested in two sets of 24 channel radar converters at 48k. i use them as my front end converters for i / o. I stress to everyone around me HOW MUCH of the sound depends on the converters. I've converted half the studios on my street to radar because the sound is SO good. I also like prism, mytek, the new digi-hdx (i used them the other day for the first time and they rock) and apogee symphony, black lion and burl b2. There are other excellent converters and they all sound different. For drums / strings I think radar is unbeatable.
I personally have a distaste for apogee rosetta/duet 1( the second one is fantastic though), protools hd192, digi, RME, Maudio, tascam.
5.) EDUCATION / LISTENING - I'm a religious follower of Pensado's Place so I always am listening to what very successful producers and engineers are doing. I also try listen to every new influential track that is being released so I can find sounds and grooves that I feel will impact modern music in the near future. That means I'm always updating my influences so I don't remain stale in my sound choices.
It helps so much to be listening to the latest sounds all the time because it helps when tweaking an EQ to actually be going for the right sound for "RIGHT NOW" instead of what a good sound from last year might be.
It always changes so the only way to stay up on a good sound is to be aware of whats out there. I tend to lean toward trends from the underground music scenes even though i'm always aware of the major label projects being released every tuesday. To me if a record sounds like a major artist record that was released yesterday, it is already too old to use for influence. That is only my opinion and plenty of producers I know that are way more successful than me disagree with me
6.) DRUMS - What makes this song work to me is the drums. In my opinion beat (an evolving 8-bar phrase) is the most "interesting" thing to listen to in the production and definitely holds the entire song together since I opted for minimalist production. There aren't any crazy production tricks like counter-leads, instrumental hooks, rises, badass transitions or anything so I put all of the production into the drum part. I opted for just a kick, snare and a hat. There's no cymbals or toms. That makes it easier to manage when I'm slicing and dicing a part.
Also instead of using a metronome inside the daw, I used a program on my ipad called DM1 to program a loop that we all played to while the video was shot. That was played on a speaker and we all performed the song while it was on beat. THe coolest thing about DM1 though is that the tempo isn't actually consistent no matter what you set it to. So I put it on 70bpm but it ends up floating around each measure between 70.010 and 71.199 or something like that. It's cool because the feeling actualy sounds more human that way because it isn't locked. Because we sync the overdubs to the video, I had to have cubase create a tempo map around a recording of my drum loop on the ipad so the daw can keep up with the floating tempo. That's a small part of the way the song feels right there.
Before we overdubbed the drums I programmed an 8-bar loop on dm1 of the general idea of what I wanted the drummer to do. He came in and learned my loop and played it exactly like I wanted. I tracked everything through the MCI console preamps and eq.
I used an sm57 on the kick beater however I liked the sound of the pedal hiting the drum so i put it on the face side of the kick drum instead of putting it inside like you are supposed to. I didn't know it at the time but that gave me an incredible BOTTOM mic for the snare drum as well so I got lucky!. I also used an akg kick mic for the deeper sound out in front of the drum. My engineer positioned it so the air from the drum hit would smack into the microphone because we thought it sounded cool. Again I guess you aren't supposed to do that either.
I had an overhead pair of tube condsener mics (mojaves) for the roomy sound. My engineer didn't think I needed a room mic but I wish he had it because I ended up using a L*****n for a room reverb to blend into the track ever so slightly. Apparently I just realized that you aren't supposed to use reverb on drums either. Mike shipley is one of my fav mixers and he says it's a bad idea. Oops. well i thought it sounded good so I did it in this track.
After the drums were tracked I did a general beat quantize to the grid favoring the hi-hats so they were almost always on my grid. That meant the snare and kick had to be moved around manually for every bar to get the feeling just right. Small price to pay for the right feel. That also means the kick and snare feel more organic than if I had quantized the KICK to the grid and let the hi-hats float around. To me that's a perfect balance of tightness and human feel.
Most mixers these days use drum replacement to get the right sound. In fact I hired a VERY well known mixer to do this mix originally because I thought I wasn't good enough to do it myself. However when i heard the mix he did I was unhappy because he had replaced all my drums and they sounded very bad to me.
So instead of using samples I always use the ACTUAL drum sound. Because we have such great control over transients now (with spl and modern editing) I think it's somewhat silly to replace drums without at least trying to get it right...
--------------------------------- KICK -
I took each mic from the kick recording and treated them as two different "samples". That means I sliced each transient and had every hit on both mics separate and ready to edit. I moved the entire KICK-OUT track back and forth until it blended perfectly with the KICK-BEATER track. Yeah it puts it out of phase but because I mix the kicks like a jigsaw (they don't fit eachother for room in the mix) so you can't hear anything bad.
After each slice was independent, I would mute certain KICK-OUT hits so the double kick hits feel more dynamic. Some are missing, some are there. On the big choruses I leave some of them in where I would usually take them out. This means the 8-bar loop is always changing. i'm pretty sure every 8-bars has something in it that's unique so it never feels like a "loop"
Also because software gates SUCK in my opinion. I now usually trim each transient manually. Some I leave long, some I leavereally short so it sounds tight. This is what makes the kick sound really cool I think. Like the first hit is booooooom then second one is like Bap. I just trim the lengths around until i think it sounds cool.
As far as any mixing to the kicks. For the BEATER track I used an S*L EQ with a boost in 800-1k for the nasty room sound that I loved. You aren't supposed to do this normally but I liked it here. I also boosted the tight sound of the beater around 4-5k can't remember exactly. Also I used SPL transient designer to tighten the attack and also lengthen the apparent sustain (both knobs cranked 90 degrees). It pushed the sound around nicely as the different trim lengths of each kick reacted differently every time it hit the spl. That made it always sound evolving.
I also used a UAD 1176. I don't like the sound of this compressor at all, I only like the TIMING. The timing is absolutely like a really fast 1176. I set it at 20 ratio (don't do this because you aren't supposed to) and about at the middle of the attack (so you could hear the attack before it compressed the sound) then a release that was timed with the quarter note breath of the song.
The outside kick had an S*L EQ boosted around 50 and I took some ugly out of 200-400 and left the high end alone. I also used an 1176 on this for timing. About the same settings as the main kick but I attacked faster and set the release faster so it would stay out of the way until the SUSTAIN of the drum released and that's what you get to hear.
One last thing is that I took a reversed kick that had some snare bleed and had that come in on the last part of the 8 bar drum phrase and it creates this smashy snare/echo/delay sound at the end of the loop. Again that is unique each time so it always sounds different. Just a by-product of reversing an entire kick drum track and having a cool snare bleed to work with. That was actually on the KICK-INSIDE channel as well so the eq, and compression was of the main kick was being used for that backward snare effect sound. Just happened to be perfect and I didn't need it's own channel. Cool. less work. --------
I then bounced the beater kick to a new track and then created a new slice at each hit. I then REVERSED every single hit so I had a REVERSED version of every drum hit. I muted most of these but at key moments in the groove I let a reverse kick come through (with a low end rolloff and some tricky compression) to enhance the bounce of the groove. I also filtered it using a UAD MOOG filter with a little bit of drive and most of the high end rolled off using the filter frequency knob. I automated that track so sometimes you hear more high end and sometimes you don't. Just moved it up and down when i felt it. You can hear it as almost a "delay" type sound behind some of the kick hits (or sometimes right before it).
Very old-school hiphop sounding. They used to do that a lot in the 90s.
SNARE - I made a copy of the KICK beater track and sliced out all of the kick sounds just to leave the SNARE hits. That track was intended to be just the kick but I ended up have such a GREAT bottom sound to the snare I used that as the main body of the snare sound.
The other mic that was directly on the snare was used just for a bit of the crack up top.
The bottom snare sound was manually gated using my slicing techniques (lots of cross fades as well ha) so it gets that really tight SPLATTY sound. I also sliced the top snare track a little different so it would sound good with the bottom.
The snares were placed a little behind the 2 and 4 beats to get the "pocket sound" just right.
I use 1176s on the snares as well to get the timing of the transient just right. UAD 1176 is amazing for this although I still wish it didn't sound so BAD GRRR.
First I cut out every 2 and 4 of the hi-hat track because it was just too many things trying to hit the back beat and it sounded cluttered. After that I used S*L eq to get 7k OUT and since the drummer used SABIAN hats (they sound so crappy and clunky) I rolled a lot of bad low end off.. However I left a bit of that cardboard sound in because i thought it sounded cool at a certain point. I used a D*x 160 on this hat because the timing was just right for the 8th note (generally is on the D*x). It's compressing about 4-6db at a 3 ratio.
I panned the main hat track about 35 to the RIGHT and then sent it to a SOUND TOYS 8th note delay with some tape saturation. The DELAY is panned 35 to the left and I automate it IN and OUT on every measure so it doesn't clutter up the sound. I only need it like between the kick and the snare in certain ways and as a FILLER sound on Verse 2 so I just rode the fader up and down the entire song.
I bounced the entire hat track and split each hat out and reversed them so I had those to play with as well. Most are muted but I use them to rev up into the snare on two. That's the WHOOSHing sound leading into the 2nd beat (clap/snare) of certain measures. It's only on the first measure of the 4 bar phrase (half of the 8). Sometimes not where you remember it so it doesn't get old. It's also ALWAYS a different reversed hat and not just the same one over and over which would be predictable.
Sometimes I cut out part of the hat toward the end of the 8 bar phrase and replaced it with the reverse version so it would sound cool and skippy. This was just here and there so it's always helping to change the feel of the 8 bar loop.
I used a moog filter on that so the high end wouldn't get annoying. the drive on that UAD moog is so good as well. Nebula's moog sounds great but the UAD drives way harder so that's why I use it sometimes over neb.
There are two claps. These are the only extra sound other than the actualy drums that were tracked. They are claps I made in rob papen punch which is a really cool drum synthesizer I just bought a few months ago because I can't afford the tempest from roger linn
I set them ahead of the snare drum a little bit so they don't hit at the same time. I try not to have ANYTHING hit at the same time unless the frequencies don't mask. That makes the song sound bigger in my opinion.
Overheads are just barely in the track at all but they function as a room glue sound to the drum track. I rolled the low end off on the S*L eq and then compressed using an 1176 with fast attack and fast release. That pumps a bit and helps the drums to breathe. Waves S1 imager pushed them a little wider as well.
The entire drum mix was compressed together using D*x 160 with a 2 ratio and about 4db of compression. Then I sent into a moog filter and put some drive on it and filtered just a bit of the high frequency out. However I only mixed in about 10% of that wet sound and used 90% dry.
All of those steps help to create that drum sound which I think is the most interesting part of the track. I put tape on the drums after editing but it didn't help at all so i ended up NOT using tape. It was just a little too much saturation after using the moog drive as well. Sometimes its good sometimes not. I was surprised that I didn't need it
7.) VOCALS - Tracked using Nueman u87 through a N**e 1064 preamp. Lead vocal has just a tiny 3k boost from a N**e eQ on a N**e filter at 70hz. I used a UAD precision de-esser on her as well. The only other thing is the UAD la2a compressor. I had to work too fast to use a neb compressor on her and the la2a sounded find to me. I didn't cut out her breaths at all because I think it adds to the live video element.
Vocals have two plates on them, a 140 and a 250. The 140 is hard panned left and the 250 is hard panned right. The pre-delay is timed to the track tempo. Hard panning two separate plates and having predelay keeps the vocals untouched in the center which is VERY important I feel. BGV mains also have the plates as well.
8.) MAIN BACKGROUND VOCAL - I sang my main bgvs into a mojave large diaphragm condenser through a N**e 1064 as well. I did a N**e 3k boost on my voice but I compressed using a fairchild with the timer on 3 for an ultra smooth sound. Our voices blend perfectly anyway so I wanted to treat both vocals as parts of ONE lead for the chorus. I ran a small amount of de-essing but probably didn't need it. Safe than sorry I say.
Most important part is vocaligning the bgv to her vocal (which is already hand edited for feel). That way it's ultra tight and the timing just snaps.
9.) BREATHY BACKGROUND VOCAL - My engineer put two small diaphragm tubes about 4 feet apart and had me stand in the dead center a few feet away. I sang the same chorus REALLY breathy into those mics. I used one take from one mic and panned to the left. I used a different take from the other mic and panned it to the right. That created a really cool stereo image on an fx type vocal track to mix in. I compressed with waves rcomp and didn't eq it at all cause the texture was perfect with the tubes. Just a high pass using a N**e filter around 80 . Of course I HARD VOCALIGNED this vocal to my main BGV part. That made it perfectly tight to the vocal arrangement. There was extra room noise in there too because of the HIGH GAIN needed to grab the vocal so I used izotope RX which is my favorite denoiser. I also used that on drum tracks and piano tracks that I felt were a little too much hiss to be in a pop track.
10.) BASS - the bass guitar was a fender jazzbass DI through a GK head. I also tracked a GK cab in the live room with an AKG mic. VERY IMPORTANT that the bass player had to play much lighter than he usually plays (barely pluck the strings and let the amp do the work) and also play BEHIND THE KICK DRUM. That means every bass note is "late" on every beat that has a kick. This is somewhat of a lost art these days where bass players try to play ON the kick drum and it just turns into a mess. That's the key to making sure the kick and bass SOUND perfect together; especially on a ballad. I also sidechained about 4db off the bass (using a kick trigger and Vengeance Sound multiband sidechain plugin) JUST a bit but most of the feel comes from the right pocket with the kick drums.
I phase aligned the DI to the cab (very common when tracking both) and I didn't use the DI at all during the verses and bridge. I just added into the choruses for a slightly more exciting sound. Most of the DI low end is rolled off so it highlights the strings a bit more for the choruses. The cab sound is mostly just the low end. D*x compressor on the bass buss and the rest is automation to get each tone to sit right. (I kill the bass toward the ends of the verses and bridge so it sounds bigger when it comes back in at the chorus)
11.) PIANO - Live tracked with mics on the hammer, two on the lower part of the sound board and a blumline for the room. Piano rooms have cupwise radio stuff mixed in them and there is an 8th note delay on the hammer track that has alexb vinyl on it. I automate that in and out to accent the feel of the track.
Piano buss is compressed with a dbx160 just barely so it pumps with the song.
12.) FAKE STRINGS - I used omnisphere warm strings as a base for the sound. they make up the low mids of the track and are featured on the verse mostly. eqd a little bit of 500 out using the cubase eq (for speed not because they sound good... they suck). I sidechain compressed using waves C6. The triggers are the piano and the kick/snare because they all borrow the same space. So whenever the drums and piano aren't as active, the string pad comes into to warm up the space.
13.) PADS - ONLY on the choruses. There are 4 stereo pads from omnisphere. They are all different, some are warbly and some are glassy. But they are all panned so the left and right are staggered. I hard limited them all with UAD L3 so the sound was VERY consistent the whole time. I then used a moog filter on the entire buss and used an LFO synced to the track that moved the filter band around so they sound like they sweep around. I also used a waves s1 imager to spread them a little wider.
The key to the pads though is that they are sidechained to the vocals and the high end of the piano (using waves c6). That means they can be LOUD but they'll never cover up the vocal or the click of the piano. The release was set to the quarter so the high end filled in as sooon as the vocals were done with a phrase and the piano quarter note pumps create a glue-like sound between the high end elements. That is definitely a huge part to the sound of the chorus. The side chain is taking off about 10 db of pad sounds so you can really feel the pumping with the motion of the piano. Have each element dynamically effect another is certainly where most of the sound comes from in this mix. Otherwise the sound would be very flat and I would have to compress things harder and lose the dynamics. Or I would have to automate a lot more like I usually do. This track has much less automation than normal.
14.) MAIN MIX - there is a L*****n room that I sent a little of each element in the mix to (except bass and kicks). I mix the room reverb solo and then send stuff into it and pann them around until I like the image of the room and it sound realistic of the instruemtns and sounds I have in the mix. I use a little eq and filtering so it sounds clean before I add it back into the track. Also sometimes it needs some heavy pumping compression to help with the ambiance. Not this time though.
Then I take the fader all the way down and slowly bring it up until it sounds just perfect with the REAL track.
After I stemmed out 16 channels to the mci (this is where I would use a nebula console if I had time) I came back in and used a N**e 33609 UAD plugin on the bus giving it some glue and I used OTARI ATR 3db as the tape. I tried every tape stock and that sounded the best with this mix. I sent that version for mastering. In mastering nebula GEQ was used to boost the high end a bit (VERY NICE SOUND) and the alexb pultec was used to enhance the low end sound. WHAT AN AMAZING NEBULA LIBRARY ALEX THANK YOU SO MUCH. Mastering guy really liked the mix so he just did some low end harmonics and a little bit of multiband limiting _3db here and there_.
That's about it. If you have any questions just post and I'll try to answer the best I can. Hope that helps in understanding whats going on in this song. Thanks for listening to the song and again thanks to the amazing developers behind nebula who have brought all this expensive hardware to the entire world of producers and mixers!