Prepare for Take-off
Setting the Stage for Success
Here in part 2 of a 3 part series on home audio production we examine the process of recording and a general overview on how to use the basic tools discussed in part 1 along with some newcomers to the party. Enjoy.
If audio production is both an art and a science it might be said that recording is the science and mixing is the art. That said, our ultimate goals in the recording stage are relatively few, four to be exact.
1. Isolated Signals
2. Quality Signals
3. Clear Signals
4. Strong Signals
I will expound on these in detail but generally stated...
We isolate tracks so that we have the most discrete and surgical control over them during mixing without interference from any other sounds in the mix. In a professional studio this is done with isolation booths and baffles (portable walls that block sound). At home you can improvise using mattresses, carpet, home made isolation boxes for amps, closets, bedrooms, bathrooms, etc. Also (generally) you want to use rooms that are fairly "dead" meaning when you clap you hear little to no echo or natural reverberation. This way you can add the desired ambiance electronically without competing with the original room ambiance. You can help this by hanging carpet, foam or mattresses on the walls. However this is not a hard and fast rule since many rooms sound great and may have just the right ambiance for the sound you want. Use your judgement.
We must make sure the signals we get are of a good quality meaning the overall sound and timbre of the instrument is good. This is accomplished by playing a well maintained instrument, playing it well, using the proper amplifier settings, the proper microphone(s), and using just enough (if any) E.Q. to give us more than enough frequencies to work with during mix-down.
We take signals that are clear and undistorted. If you're encountering unwanted noise from your equipment it is probably being caused by a grounding problem and there are a few things you can do to try and solve it. You can try and plug the offending gear into a different outlet. You can try using a power conditioner. You can also try using a direct box. If all else fails you may want to call an electrician to check the wiring in your home. That may seem extreme but anything you can do to get rid of unwanted noise is usually worth it in the long run. If you don't believe me then just try using a dirty signal all the way to mix down and if you end up with any hair left try and save yourself some aggravation next time.
We take signals that are strong to maximize signal-to-noise ratio. This means recording the signal as loud and strong as possible without being distorted in any way. We can always turn things down, but if you try and boost a weak signal after it's recorded you boost any noise that is present along with it and that's not good.Credit: http://commons.wikimedia.org/wiki/Main_Page
To best illustrate the recording process I'm going to take you through the signal chain step by step and discuss each step along the way. The signal chain is simply the path the signal follows on it's way to the recorder and eventually the final mix. It's important to note here that it's best to keep the signal chain as short as possible while still getting the sound you want to tape (or disk). More links in the signal chain multiplies the possibility for not only unwanted noise but distortion due to unnecessary "gain stages". A link in the signal chain can be called a gain stage if it has a potential effect on the signal's gain (volume). If you're patching through a piece of equipment that has a gain control (volume knob) it's possible that the gain can be set too high causing distortion and ruining the signal. We need to remain careful and conscious of this along the way.
Input Stage (Microphones vs. Line Level Instruments)
For the most part, sounds we record come in three flavors:
The strongest are line-level input sources. Powered sound modules like keyboard synths and drum machines are capable of coming into the board either at or very close to line level. These instruments have small amplifiers in them already which can boost the sound to an acceptable level for recording all by themselves. Most mixers (real and virtual) have either dedicated line level channels or a switch that allows you to choose between line level and microphone at the input stage. If you're recording a line level instrument such as a keyboard synth the gain or volume control on the instrument is your first gain stage for that signal and you should start by turning it all the way up. Chances are very slim that the unit will distort itself. Just make sure the mixer fader for that channel is set to zero (about two thirds of the way up on most boards). This is called "unity gain" and it means that nothing is being added to or subtracted from the signal coming into the mixer. If the signal is coming in too hot, back off on the output control of the device until you get a good level. You should always try and bring every signal you record to line level as early as possible in the signal chain. Also, the default setting for some keyboard sounds is in stereo. Be sure to investigate this and if that's the case either set the keyboard to "SUM" the output to just one channel or you'll have to use two channels/tracks for that sound.
The next strongest are the signals that come from electric guitars and basses, if that is, you are plugging in direct (without an amplifier). These signals can vary in strength so a direct box is needed. Every home studio should have at least one. These devices serve a couple of very necessary purposes. They boost signals so they are stronger coming into the mixer. Also, if there is excessive hum or noise coming in with the signal due to grounding problems there is a ground switch on the direct box that many times will eliminate the problem.
Sounds picked up by microphones are the weakest signals we encounter. In order for the recording device to properly "hear" signals coming from a microphone, those signals must go through a pre-amp stage where they are boosted to an acceptable level. Mixers, 4-tracks and other portable studios usually come with built-in pre-amps for every input channel. Sometimes these knobs are labeled "trim" and are the dial nearest to the top of the mixer channel strip. As we've said, computer systems rely on audio interfaces to not only convert the incoming analog audio data to digital, but also to serve as a mic pre-amp, boosting the signal to an acceptable level. Separate outboard mic pre-amps are also available. If you plan on recording more than one or two sources at the same time as we've said you will need to make sure your interface or tracking device has a satisfactory number of inputs AND be sure your computer system is strong enough to record/playback all that data at once.
These days electric guitars really don't need to be miked. You may disagree and that's fine but do yourself a favor. If you have an amp modeler (real or virtual) give it a try. You just may be surprised at the quality and flexibility this gives you when laying down guitar tracks. This means that you will only record the dry guitar sound and then route it through the modeler during playback. This gives you the freedom to choose many different sounds on the fly during the recording/mixing process. Also, if your guitar amp has a line out jack you can run a cable directly from the amp to the mixer. You won't need a direct box in this case unless you get unwanted noise caused by a "ground loop". Then you can try the direct box and use the ground lift switch to see if that solves the problem.
The guidelines for recording electric guitar and bass are very similar. There are basically two routes to go with each: recording direct, or miking an amplifier.
When being recorded direct (without an amplifier) both electric guitar and bass should be recorded with compression (see Dynamic Processing below). As I've said, a direct box should be used between the instrument and the mixer (or interface) then a compressor should be inserted in the signal chain at the insert point on the mixer to get the dynamics under control. Since the bass is more percussive and has more of a pronounced attack a higher compression ratio should be applied to the bass. Start with a ratio of 3:1 and adjust from there so that compression kicks in when the bass is being played at around 60% full strength by the musician.Credit: http://commons.wikimedia.org/wiki/Main_Page
When you mic a guitar or bass amp, using a short boom stand try placing the mic a few inches away from the speaker. This is where you put on the lab coat and goggles (metaphorically speaking) because depending on the sound you want you're going to experiment with the mic placement. If you aim the mic more towards the center of the speaker cone you're going to get more treble, high-end, edge. Aiming the mic more towards the outside edge of the cone will give you a warmer sound. It's true that SM57's are a good choice for this purpose but don't be afraid to try different things. These methods are only guidelines and the possibilities are endless depending upon how creative you want to be. If the guitar amp is not going to be set that loud you can also try using a condenser mic to capture some ambiance too, or you can try both mics, the 57 close in and the condenser farther away recorded on a separate track to blend in later as you wish. Speaking of volume, we know that some amps, especially tube amps sound a lot better when they are turned up. You might get a better overall sound by doing this. Most guitar players agree with this idea because we just love playing loud anyway. If you do crank it up during recording that's fine, just make sure you maintain proper isolation if anything else is being recorded at the same time and think twice about using that condenser mic on a source that is strong enough to blow dry your hair. Also try and get the take as quickly as you can before the neighbors call the cops.
If you're not going to use an amp modeler, now comes the discussion of whether or not it's best to record guitar with all it's effects (wet) or not (dry) and adding effects later. Depending upon your set-up and if you have tracks to spare you can either try capturing both wet and dry signals on separate tracks using either the line out or effects loop send on your amp (provided this doesn't shut the speaker off) OR try to do it with two different takes. Another option is to record the signal wet with all effects in place except reverb. Save that for the final mix stage since it plays a special role and can easily be added later.
Miking vocals as we've said can generally be done with a decent condenser mic. Also, just like with bass, be sure and insert that compressor early in the signal chain and begin with a compression ratio of 2:1 for softer vocals and 4:1 for louder songs. Sometimes when you're working with a loud singer who is a screamer or a yeller a dynamic omnidirectional like an SM58 is a better choice. Always use a pop filter which is a round physical screen that clips to the mic stand and prevents the singer's P's B's S's and T's from ruining the take by clipping the track. If this becomes a problem during recording instruct the singer to direct his or her voice off to the side of the mic so their breath is not hitting the mic directly. This way the sound will still get picked up but plosives and sibilants will be minimized. You should also be aware of the "proximity effect". This means that the closer the singer is to the mic, the more bass frequencies the mic picks up. Be sure and experiment to get the best sound. Monitoring is also an important part of recording vocals. A singer has to be comfortable with the monitor mix in his/her headphones in order to deliver the best performance possible. Most singers prefer at least some reverb in the monitor mix. If you have the tracks available it's also usually a good practice to record two or three good takes and apply them in thin layers over the final vocal with just enough volume and judicious panning to thicken things up a little.
The drums can be a song's most important element. The best song can be ruined by bad drum sounds and an average song can sound a lot better with good drum sounds. For that reason, and especially for novice engineers/producers I advise using sampled drum sounds using midi triggers or an electronic kit. If you know how to program beats properly you don't have to sacrifice any human feel and the sounds themselves will be consistently good because they have a good signal to noise ratio and have already been put through a dynamic processor. If you use a sampled drum kit I reccommend having a few kits already programmed and ready to go. You don't want to get bogged down by this process when it comes time to record. You can always swap sounds out later during mixdown. Make sure the samples themselves are dry so you can start from scratch with effects that are appropriate for the specific song you're recording.
If using samples is not an option there are many different approaches to miking drums. Most home studios don't have the luxury of being able to mic every single piece of a drum kit, and that's a good thing. You rarely will ever need more than four mics. One dynamic mic on the kick drum (just inside the shell aimed at the point of contact between the beater and the head), and one dynamic mic on the snare drum (underneath the hi-hat, aimed at the center of the head, away from the hi-hat to minimize bleed). Two condenser mics or omnidirectional dynamic mics over head. If you only have three mics opt for two condensers overhead and one dynamic on the kick. If you only have two inputs available choose the two overheads. If you're going bare bones and only have one track use one overhead condenser pointed down at the kit. With all of these methods you'll have to make some test recordings to find the sweet spot for each mic. Yes, it's time consuming and a pain in the ass, but still the best way to go.
These set-ups gives you some degree of isolation on the kick and snare which are the most important and frequently used pieces. It also gives you a decent stereo picture of the rest of the kit to work with. In theory, a good drummer should know how to balance his sound on the kit anyway. Even with this abbreviated method of drum miking you will probably spend the most time getting a good sound from the drum kit. Again, take time, experiment. Have some extra muffle material (blankets, pillows) on hand for inside the kick drum if necessary. Have a roll of duct tape handy for any rattling hardware. Spring for new heads on the kit and tune them up correctly. If you don't know how then Google up some tutorials. In the long run, as with anything, you will only get out what you put in.
If you're dealing with an acoustic (real) piano, God help you (just kidding). The best approach to this is to open the lid (whether it's an upright or a grand piano) and hook up two condenser mics to do the job. One for the high strings and one for the low. You also may want to try panning them hard left and right to get some good stereo separation and a more realistic sound. If you can't manage that then just go with one omnidirectional dynamic.
Any other acoustic instrument you may be dealing with such as guitar, winds (flute), brass (trumpet), reeds (sax), strings, etc. can usually be adequately miked with a condenser but again you will have to play around with it for a bit to find out what positioning works best. With an acoustic guitar placing the mic a few inches from the strings at the body fret (the point where the neck meets the body) is a good starting point. The closer you get to the sound hole the more low end will be picked up. If you have the resources you may want to use two condenser mics and get a fuller stereo picture for pan separation later during mix-down.
Into the Mixer
When the signal starts passing through gain (volume) controls either on an amplifier, mixer, or effect processor you always have to pay attention to how hot the signal is coming out. You do this by keeping a close eye on the channel meters and a close EAR on how it sounds. Headphones are great for this. On almost every mixer channel there's a signal meter that looks like the flag lights at a drag race. Green at the bottom and red at the top. Very old mixers may have a V.U. meter with a needle that bounces back and forth with the signal strength. This provides a visual clue about how hot the signal is coming in. You want to keep as strong a signal as possible coming in without "clipping" or distorting. Any sign of clipping on the meter (red lights) or distortion through the monitors (or headphones) can be fixed by backing off a bit at the latest "gain stage". Remember, we want to maintain a strong signal, but not a distorted one.
Now I'm going to take you through the various stages of the mixer. A lot of this may be review for some but it should be covered to bring everyone up to speed. If your working with a virtual mixer it functions in much the same way as a real (hardware) mixer only it is much more scalable depending upon the size of your project. As you look at a mixing board (either real or virtual) you are really looking at a series of channels, each with it's own set of controls arranged in a thin strip from the top of the board to the bottom (where the fader lives). Signals always flow from the top of the mixer to the bottom.
The first control the signal flows through is the mic preamp. This is your first chance to screw up the signal by setting it too hot. When setting up to record, tell the musician being recorded to play as hard and as loud as he/she will during the recording. Then set the level so it is just shy of hitting the red zone without actually getting there. Again, this is where you want to be for all signals you record. Sick of hearing this yet? That's how important it is.
Some hardware mixers come with a 1/4" jack on every channel that is labeled "insert". These jacks are capable of accepting a TRS (stereo) plug that effectively becomes an output/input loop for the signal immediately after the mic preamp. They're kind of like an effect send/return on a guitar amp. Inserts are useful for introducing things like a compressor or an outboard E.Q. early in the signal chain. A "Y" splitter (two 1/4" female mono plugs to one 1/4" male plug) allows you to do this. (See "Dynamic Processing" below)
Effect send or auxiliary controls send a copy of the signal to an outboard effects unit such as a reverb or a delay. There are "effect send" and "effect return" (aux send / aux return) jacks on the back of the mixer you will use to patch the effects in. Many mixers have more than one effect send channels and each one can be used for a different effect or processor. Once connected to an effect, turning up the effect send 1 knob on channel 1 (let's say it's a rhythm guitar) sends that amount of guitar to the effects unit for processing. The same effect is patched into effect send 1 across the board on every channel. When the processed signal returns to the board there are separate "effect return" faders which set the level of the effected signals. Some mixers also have a "pre/post" button next to the effect send control meaning "pre-fader and "post-fader". Pre-fader sends a copy of the signal to the effects unit that is not affected by the channel fader. Post-fader sends a signal that is affected by the fader. So for example, you can fade out a vocal and its delay at the same time. Most of the time you will keep these buttons on the "post" setting. You can also get creative and bring the effect returns into another vacant mixer channel if you want more control over the effect returns (such as E.Q.) or if you want to layer effects on top of one another such as a reverb on top of a delay. If you run short of channels on the mixer effect returns can also be used as additional inputs since they also usually have their own bus controls (covered below).
Initial recordings of sounds should be done with as little E.Q. as possible. How to use equalizers in general is will be covered shortly but the E.Q. section of a mixer is your first chance to tweak the signals frequency spectrum to your liking. Depending on the complexity of your mixer on each channel you may have between two and seven controls or more. The simplest set-ups will have little more than bass and treble shelving controls. Higher end units may have high, high-mid frequency, high-mid Q (bandwidth), low-mid, low-mid frequency, low. The more control you have over E.Q. the better but you can often get away with using fewer controls than you would think if you brought good, full, natural sounds to tape (disk) in the first place.
Some boards may have a monitor section which allows you to build a custom mix that is different than the faders. This is useful for creating a monitor mix for musicians to listen to as they record their tracks. If you have more than one monitor section to work with you can create more than one mix in case you're tracking more than one instrument at once and your performers have different needs in their headphones. If you don't have a monitor section on your board you can use effect sends for this purpose. This will sacrifice some of your ability to apply effects to the mix early on but at mix down you will have these channels back.
Busses are pairs of channels that can accept any combination of signals to create a "su-bmix". These channels have assign switches that can be found at each channel fader and effect return fader. The master stereo output is also a type of bus and is usually labeled L/R. When you're recording using a physical (real) mixer you will use the buss outputs to bring post fader signals to each track on the recording device so they can be recorded. During mixdown if you have four or more tracks of drums and you get the balance between them just right you can bus the lot of them to a single pair of busses and control the volume of the whole drum kit with just one or two faders. This makes life a lot easier than trying to proportionately adjust all the drum faders at once when trying to get a mix.
Pan controls place the signal in the stereo picture much like the balance control on a home stereo. When the channel is assigned to one or more busses pan controls also proportion the signal between the selected pairs of busses. Bus 1 is left, 2 is right, 3 is left, 4 is right, etc. So if you want a signal to go JUST to bus 3 you would press the assign switch for 3/4 on that channel and set the pan control hard left. Then bus 3 would get all of the signal and bus 4 would get none of it.
These powerful tools are actually effects that control the volume or dynamics (loudness / softness) of sound. The "dynamic range" of a piece of music is the difference between it's loudest and softest parts. Effects in this category include compressors, limiters, and expanders. While these are technically categorized as effects they serve a very special and necessary purpose at the tracking (recording) stage instead of the mixing stage as most other effects do. Almost everything you record can benefit from some degree of dynamic processing while it is being recorded with the exception of sampled sounds (such as drums) that have most likely already been processed this way and don't need it. All dynamic processing is generally best applied at the recording stage through the mixer's insert mentioned above, just after the mic preamp.Credit: http://commons.wikimedia.org/wiki/Main_Page
Compression is one of the most important tools you have when recording. Compressors stop the volume of a track from rising too high and clipping which would otherwise render the take unusable. They also help smooth out the volume overall giving the track a sense of uniformity and continuity. This helps provide a consistent level for mixing AND perhaps most importantly controls transient spikes in volume which allows tracks to be turned up to increase signal to noise ratio for a stronger yet quieter (less noise) mix. A compressor can also be used on an entire song during the mastering process (discussed later).
Common stages of compression include...
Gain: Sets the overall volume of the output.
Threshold: Sets how strong the incoming signal needs to be before compression kicks in. The higher this parameter is set, the louder the signal has to be before it starts getting compressed.
Attack: Controls the length of time in milliseconds the compressor takes to actually start to change the sound level after the volume reaches the threshold.
Release: Controls the length of time in milliseconds the sound should be held by the compressor after the volume level falls below the threshold.
Ratio: The difference between the incoming and outgoing signal. A 3:1 ratio says that when 3dB of sound comes past the threshold, an increase of only 1dB comes out the other side.
A limiter is actually an extreme compressor with a ratio of about 10:1 or more. Limiters are good for turning up an entire track via the gain parameter with absolutely no fear of clipping. Usually however, compressors will suffice unless you're dealing with a signal that has a very wide dynamic range. Limiters are often used during the mastering process to get a full, loud sound from a final mix.
Expansion is the flip side of compression. Compressors limit the high end of volume and expanders "quiet down" the low end. They eliminate noise during otherwise silent sections by effectively muting the track temporarily until the signal returns past a certain preset level. A noise gate is a very basic form of expander.
Common stages of an expander include...
Threshold: Sets the volume level at which the gate opens and closes. When the signal falls below the threshold, the gate closes and volume is completely attenuated. When input signal returns and reaches the threshold the gate opens and the signal is heard again.
Ratio: Controls how much the expander attenuates the signal.
Attack: Sets how fast the gate reacts when the volume reaches either the open or close threshold. This is usually set to be fairly quick.
Decay: After the hold releases the signal, how long until the gate fully closes. A medium setting is most commonly used.
The basic process of learning how to set expansion levels begins with miking a fairly strong signal source such as a guitar amp, inserting an expander into the signal chain with the ratio maxed out and the threshold all the way down. Put on some headphones and listen to the noise floor (the room ambiance). You may hear the air conditioner running or the amp humming slightly. Now turn up the threshold knob slowly until these sounds begin to disappear. Set the knob slightly above that. Now when the guitar is played it is easily heard but when it stops the background noise disappears with it. However, you usually don't want it to shut all sound out completely. That's where the ratio setting comes in. Try and set this so that the gate closes smoothly, doesn't cut off any signals prematurely and is never fully closed with some ambiance barely audible. This is a little nit picky but it does give the expansion a more natural and transparent effect.
Most things wired (guitar, bass, keys) can be recorded using speaker monitoring (in the control room/area) only if you are recording the signal either by plugging directly into your recorder/mixer (no open mics) or you are miking amplifiers that are acoustically isolated in another room. If you're working without a control room that is acoustically isolated from your signal sources (amps) meaning you have open mics in the same room as the monitor speakers this isn't going to work very well since the monitor mix (whatever the performer is listening to in order to stay in sync with the song) will "bleed" into the open mics destroying our isolation factor, which is bad. Remember, best case scenario, when you record guitar, guitar is the only thing you want to hear on that recorded track when it is isolated.
Okay, here's where the rubber meets the road. When you feel you're getting the best and strongest sound possible through your current mic/preamp/mixer signal chain (multiplied by however many tracks you are taking at once) then it's time to record. I recommend using a click track when recording to facilitate effective editing later on if necessary. Even if you're not planning on editing this way later, do it anyway. Believe me you will be glad you did. This is achieved by recording a click at the proper tempo on its own track and having the drummer wear headphones while he/she lays tracks with the click track playing in their monitor mix. Drum machines are usually a better choice when tempo changes are involved since they can be programmed as such. Most good drummers can handle this process and turn out a good performance. Virtually all pro drummers can as well. After a good drum track is captured and there are no drum-less parts where you still may need the click the click track can be erased and the track used for something else.
Now, if you're tracking a whole band at once and you have enough inputs and processing power to pull it off just press record and let 'er rip. If you're not tracking the whole band at once, drums are up first. If you can manage it I highly recommend recording a scratch (disposable) guitar track (monitored by the drummer via headphones) along with the drums. It helps the drummer keep his place in the song and lets him play more naturally than he would by himself. If possible it's a good idea to record a scratch vocal track as well just to give everyone more of the feeling of all playing at the same time. It helps preserve the normal feel of a tune. After that comes the rest of the rhythm section: bass, rhythm guitar(s), keys. Samples (if any). Acoustic instruments: guitar, strings, horns, etc. Then the icing: lead guitar, vocals, backing vocals.
So here we are recording (either separately or together). Intro, verse, hook, hook, hook, EPIC ROCK N ROLL ENDING - okay STOP. Listen back. Are the levels strong but not clipping? Good. Isolated? Good. Is it a good performance? Most players need a few takes to warm up and that's fine. If you're still sitting there an hour later trying to nail the same rhythm part, back to the woodshed with thee! Come prepared. Recording is an entirely different world than the practice space or the stage. Some players have to find this out the hard way. If you're working in digital you have no excuse to not keep every take. Save that stuff. You never know when you're gonna want to go back to mix and match takes. It happens. Also, AS you listen back...you're tweaking. Tweak, tweak, tweak, levels, E.Q., levels, E.Q., compression, maybe some light reverb through the auxiliary. Rinse, repeat... This process of "mixing as you go" in a sense saves you time in the long run when you're trying to get that elusive final mix that everyone loves. Plus it helps your monitoring situation, giving your performers a better mix to work with.
Now that you've captured your lightning in a bottle it's time to make sure it's ready to be released back into the world. Next we will examine the processes of editing, mixing, and mastering your project.