|
|
|
|
The Frank G Guide™ to Creating, Recording, and Producing a Hit Record Using Only a PC Workstation |
|
BY FRANK G |
|
|||||
|
|
||||
Part Two: Cutting Vocals, or Keeping the Band from Killing One Another
When we left off (in part 1), I had just finished arranging a new song - which I created entirely in my PC - and gave it to the artists to write to. Two days later...
Now I know why I produce these guys! They loved the track, took it home, and wrote 4 verses and a chorus -- overnight! In fact, one over-productive member of the crew wrote himself a 64 (!) bar verse....we'll deal with that in a moment. First, let's get something recorded.
I load the project back into Nuendo. Everything is saved exactly where I left it. There are still NO audio tracks. Everything we're hearing is MIDI, playing back through virtual synthesizers. The mix sounds great, and is more than suitable for a cue mix. Good thing, because the only cue system I have is (and I kid you not) a wireless speaker setup, which sends the output of my stereo bus wirelessly downstairs to a speaker, I then take the line out of that speaker (which is intended to go to the OTHER speaker) and plug it into a cassette deck in the vocal booth. Plug headphones into the cassette deck (oh, and don't forget to put a blank cassette in there, in record-ready/pause so that audio can pass through..) and we're groovin'! Ingeniously inexpensive... I always like to ghetto-engineer existing gear into doing what I need, rather than running to the store and buying things (as some of my previous employers were known to do.)
Since I'm running my cue system wirelessly, how do you suppose I run my vocal mic? Yep....that's right! With a standard XLR cord, just like everyone else. The actual vocal booth is in the hallway downstairs. The studio is upstairs. I simply dropped a mic line over the stairwell, tucked it into the corner, and gave myself an XLR input to my board from the bottom of the stairs/vocal booth. When the studio door is closed, leakage is extremely minimal. My talkback system is equally impressive - I took a semi-functional EV RE-16 and plugged it (yes, the mic itself, directly into the jack - minus cord) into an input on my external mixer. When I need to communicate through the cues, I simply turn that channel on and talk. If the place wasn't such a mess right now I'd take some pics (my house is in a temporary state of disarray while we rearrange several rooms. I decided to move the lizards upstairs into the studio, because they love noise, and clients love them.)
Now that the setup is taken care of, we're ready to cut some vocals. I take the song back to the top and hit play. I usually let the vocalist run through the song while I set levels, do a bit of monitor EQ, throw a bit of reverb on it -- essentially, I mix the vocal in with the song WHILE cutting is taking place, so that the vocalist has a decent idea of what the 'final sound' will be like. In doing so, I discover that there is a spot where some yelling takes place midway through the track. If I was unaware of that, chances are it would wind up clipped on the resulting recording. This is one of the best reasons why you should
NEVER WALK AWAY WHILE CUTTING TRACKS!!!
I compensate by applying a compressor to the input at a very slight setting of 2:1 -- just enough to reduce the level on that yell to something that can be dealt with. Levels are set, vocalist is warmed up, let's do it.
I hit play and record on my 428 and we're rolling. Having actual transport controls feels much more intuitive than clicking the play and record buttons on screen, even after close to 10 years of digital audio work. We get about halfway through the first verse, and a mistake occurs. Oops, wrong lyric. Rather than start back at the top (the first part was very good) I decide we should punch-in. Doing a punch-in with the 428 is a breeze. Back the transport up so the performer has time to find their spot, hit play, then hit record right when you want to punch-in. Sound familiar? That's because it's the same procedure you'd use to punch-in whether you were in the computer, using da-88s, or even good ol' 2 inch analog. That's the beauty of the interface -- It lets you work the way you are used to working, rather than forcing you to learn 'new skills' such as rapid mouse handling. In fact, one studio I know of refused to get their control surface repaired when it broke down, stating "The engineers like to use it too much. No one is bothering to learn the keyboard shortcuts, and it makes the Pro Tools mixes go too fast!" Now, while I like a good keyboard shortcut as much as anyone else (perhaps more so, being a video guy....check this pic of my keyboard for proof) anything that increases productivity AND a desire to work can't be a bad thing....
When a punch-in is performed in a digital audio workstation, it -- like most everything else -- is a non-destructive edit. Simply put, you're not actually 'recording over' the spot as you would be on tape. You're 'recording on top of' the existing audio data.
 |
You have to think in three dimensions here, think of the timeline as a gameboard laying flat on the table that you can stack pieces on. This means that you can adjust the in/out point of the punch-in. Did you forget to punch out and 'erase' the next line? No problem, simply drag the end point of the punch-in region to the left revealing the original data underneath. Is the new performance great, with the exception of that one word that was nailed on the prior track? No prob, find that word on the new track, select it, and delete it. Voila! The new performance of that word is gone, and the original shines through. Each time you press record, the DAW creates a new audio file with its own individual name. For this reason, you must be careful if you intend to take your tracks elsewhere. If I do a bunch of vocal overdubs in Pro Tools, I have to be sure and consolidate them into one cohesive vocal track before burning the project onto cd, otherwise I'll have to figure out how to reconstruct the track from 20+ files called 01_Vox0000xx.wav. Once you forget and have to do this once, it'll become habit not to do it again!
 |
Verse one is done! We listen back and decide it's a keeper. Rather than go on and record the first chorus, I usually go on and complete all the verses, THEN work on the chorus. Who's next? Great! This guy usually requires a million punch ins, but he's been steadily getting better. I'm ready for this track to take forever, and then, without warning, he's dead on this evening! Somewhere around bar 16, we all comment on how great that performance was.....around bar 32 we start to wonder when it might stop....bar 48 hits, and we're looking at our watches and laughing at the insanity coming from the booth...finally, bar 64, he ends the verse! One take, 64 bars, no errors. That's enough to turn into a track for a solo EP.... We inform him that that was great, but that 32-48 bars would be lost when we mixed. Uh-oh! He went so long, that there isn't enough room for the other guys! A brief argument ensues, featuring language that Bob would only edit out if I described here. They ARE rappers, after all - which brings up an important point: If you have a low tolerance for obscenities, you should NOT become a recording engineer -- not in Detroit, anyway... What to do....I don't want to delete anything until I have a chance to hear it and absorb the good parts, but I don't want to just stop his verse at 16, because the next guy won't know where to come in as they're playing against each other. I select the 32 bars of music he shouldn't have been rapping over, copy them, and paste them right after themselves. In doing so, I've just extended the runtime of the song by 32 bars, but I've managed to restore the timing/structure they wrote the song to.
Next in the booth is the guy who never messes up, and after the last take, I'm thinking we may get done early. Then it happens -- these two guys must've reversed places for the evening, as the 'perfect' person proceeded to need many, many punch-ins. At one point, I hit record when I shouldn't and 'erase' a previous sentence that we wanted to keep. The take I recorded over it with was no good, so the easiest thing to do was hit "Undo" and make that last recording go away. When it's all over, there are 19 segments making up this one vocal track. I decide not to consolidate them right now, because I want to listen to a couple of the 'bad takes' when no one's around. At this point, I check my audio directory to see how big the project is. The entire 6 minute song is taking up 40 MB right now. Waitaminute! 40 MB? How can this be? Digital audio takes up 5 MB/track minute, so how can I have 18-19 tracks/6 minutes in 40MB? Very simply - remember that ALL the music is virtually tracked at the moment (and in fact will remain that way unless I need to take this project to another computer) The only thing taking up space are the vocals themselves. There are approximately 3 minutes of vocals taking up 30MB, and the remaining 10MB is made up of the various punch-in regions.
Only one verse left, but he decided to rewrite some things after hearing the other 3 verses. While we're waiting for him, we'll work on the chorus. I tell the other guys to get in the booth, and they shout their group chorus. It sounds cool, so I cut 7 more just like it. Now I have 8 tracks of vocals that I mix together into the chorus part. I group the 8 faders together so that I can control all the chorus vocal tracks with one fader, thus keeping them mixed together. Each chorus ends with a line from the person who performed the verse, so I have each of the three in the booth read their individual lines. Since these are at the end of the recording timeline, and need to be inserted in the choruses, this is going to require more cut and paste to get the sections in place. When using cut and paste, it is frequently quite helpful to use the snap to grid function. If, for example, I set the grid to 1 bar, I can only cut or paste things starting on the 1. This is essential to proper looping/adjustment of parts. One warning, if you don't have the project tempo set to match that of the song, the bars on the grid won't correspond to the bars of the song. Make sure you set the proper tempo and switch the display mode from Hours:Minutes:Seconds:Frames to Bars:Beats before attempting cut and paste.
As I finish pasting the chorus sections into the song, that last verse is finally ready. We settle in and record it in two takes, with no punches. That means we did it twice (and kept it twice. If you don't keep it, it is NOT a 'take'.) I pick the best parts out of the two and merge them into one final vocal track. We should be done cutting...Oops! We need an overdub track for adlibs, doubles, etc. I send everyone back into the booth and arm another track. 6 minutes later, they all come back out, happy with what they've just laid down. The song is starting to sound more finished.
I hit the 'go to zero' button on the 428 and hit play:
Click here to hear my monitor mix
*WARNING* File is 3MB
Song contains Explicit Lyrics
At this point, you can tell that a couple of things still need to be put back in place, and at least one verse needs to be considerably shortened. Of course, the entire mix still needs to be completed/automated, as this is just a monitor mix. Granted, it's closer to the finished product than you may be used to hearing from a monitor mix, but I'd estimate I'll need to spend an additional 1-2 hours mixing this song. I tend to 'pre-master' as I go by using many of my favorite mastering tools on the output of my stereo bus. This has a lot to do with the loudness and punch of the song. There are very few individual compressors working (and none of them on the music), so the mastering compressor (also known as a limiter with a ceiling control for the plug-in challenged) is doing most of the work. Having gone through several stages of mastering, this rough mix may sound better than many "finished" yet unmastered mixes. Mastering can truly turn an "OK" mix into a "Killer" mix in a relatively short time....but we'd better talk about mixing first!
Next issue: Part Three: Mixing/Automation, and keeping the band out of the room while doing so
|
|||||||||
|
|
||||||||
