February 7, 2019

Recording is a Lie

At some point, every artist with whom I've worked has made some statement about "being true" in the studio.

What is "true"?

This is a very philosophical discussion, meaning there is probably not really a definitive answer, just an endless list of things to consider.

Also, it is a very serpentine discussion, meaning I'm not sure where to start, so I'll just pick a spot on the canvas and begin painting with broad strokes.

What is "true"?

By "true", most bands are referring to being true to 'their sound', how accurately they are portraying it, usually based on their live performances.  They want to sound like they do live or, more accurately, what they want their live performances to sound like:  the quality of sound of each instrument (such as a certain distortion on the guitar or an un-gated snare), the instrumentation (eg. guitar + bass + drums + vocals and nothing else!), and other qualifiers (feel, energy, tightness, etc.).  There is a respectable amount of desire to make the recording as representative of the band's sound as possible, but even aside from agreeing on what is meant by "true" and "sound", there are a host of other things to consider.

Let us suppose we record some songs as you play them with the instruments you normally use to play them and the musicians with whom you normally play.  Where do we put the microphones?  Microphones are the ears of the studio recorder, and no one ever listens to a concert with their ear in the kick drum.  And if they did, they could not very well have it in front of the guitar amplifier speaker at the same time.  And up to the vocalist's mouth as well?  (Or, rather than the vocalist's mouth, should that be the PA speaker, where the vocalist's voice is amplified to be heard over the band?)  So, merely placing microphones on each instrument is a lie because we humans have (at best) two ears permanently affixed to an orb on our shoulders, making it impossible to place our ears in front of all these instruments in the manner microphones are placed.

And there's a time-space situation to be considered since a microphone next to a speaker hears the sound almost instantly, but a microphone across a large room will hear that same sound 18 milliseconds later, and that can greatly alter tone quality when blended with other sounds.  Lies!

And different microphones have different frequency responses, missing out on some frequencies and accentuating others.  Technically, this is not a true representation of the sound created, that is to say:  Lies!

Let's overlook that for now.  Do we put reverb on the voice?  Compress the bass?  EQ on the guitar?  LIES!  More lies!

Ok. So maybe "lies" is starting to sound a little heavy-handed.  Let's call them "compromises" because what is most true is often not the most flattering.  There's a reason snares get gated, voices get reverb, basses get compression, and so on.  Our ears hear a wider dynamic and frequency range than most microphones and definitely more than most speakers, so the microphones are losing some of the information and the speakers are turning around and losing more.

And we haven't even discussed what happens when someone digitally compresses the audio.

Too, what will your audience be listening on?  Bluetooth earbuds?  Laptop speakers?  Home stereo?  Car stereo with road noise?  All of those?  Something else entirely?  Whatever the answer, it affects the sound.  ...and it is out of your control.  So, whatever the truth -er- compromise was up til now, it's been filtered with the end listener's choice of playback device.

Too, I feel it worth mentioning that most albums I've recorded have been well-served by a nice tambourine track somewhere on the album even if the band doesn't have a tambourine player.  We add it because it sounds nice, but it's technically a lie (in regard to representing the band's sound) since they don't have a tambourine player.  But it's usually an acceptable lie, serving the greater good and all that.  So, what about an extra acoustic guitar rhythm track to "beef up" the mix, even though the band doesn't use an acoustic guitar live.  Should we double track the lead vocals?  (Freddie Mercury, Eminem, Lynyrd Skynrd, Elton John)  Perhaps bring in a vocal ringer come in to sing harmonies?  (Shannon Hoon was not a member of GNR.) What about a keyboard track?  (Billy Preston on the rooftop?)

My humble opinion.

As Alan Moore says (through his character V), "Artists use lies to tell the truth."

The studio is for recording the platonic ideal of the song, whether it be something you can approximate in a live setting or not.  My goal as an engineer is to go unnoticed in the final product.  By extension, I think the goal of recording should be to sculpt a sonic product that conveys the song to the listener without drawing attention to the production.  When you watch a movie, you should never think "that pan shot was shaky" or "the background action is distracting", rather you should think "How will they make it out of this situation?" or "Just kiss her, you fool!"  To the extent your budget will allow, re-do the parts until you're happy.  Use the sounds that please your ears.  Modify as necessary to capture the essence of the song.

I have far from exhausted this topic, and, unless you really pin down the question, there really is no answer, but I hope this has given you something worthwhile to think about.

July 19, 2018

Preparing to record

Preparing to record 

(for the musician who has never recorded in a studio)

The more you plan beforehand, the more you can enjoy your session later.  These are just a few topics to consider to keep your productivity high, your stress low, and your focus centered.  Added bonus: a brief glossary of terms at the end.

Practice

Before you record, practice the song.

Oh, you'd be surprised.  A lot of people show up to the studio with absolutely no idea what they're going to do when the light turns red.  Or they have a loose idea and want to cobble a song together with digital editing.  I'm a pretty fast editor (if I do say so myself), but I'll never be nearly as fast as a good take.  A good take means I hit "record" and the performer plays/sings their part as intended.  Done.  But comping, on the other hand, requires recording a series of takes, listening back to each take, comparing them together (often a few times each), then splicing together the good bits to make a final "take".  I have had people record over 20 takes singing a single chorus and then want me to splice together the best parts word by word until we have a "good take", which I'm happy to do, but it's costing the customer.  Most artists don't have that budget.  I'm just trying to save you money by saving you time.  The better prepared you are, the better takes you get, and the less comping we'll have to do.

But if you're ok with it, I am happy to comp all day.

Pre-record

Before you come to the studio, it's a great idea to record at home, even it's with just a cell phone.  If you're lucky, you have a friend with a recorder.  It doesn't have to be elaborate or high-quality, just something to reference.  If you haven't recorded before, this is invaluable.  You can finally listen back on yourself without focusing on actually making the sounds, whether is singing or playing an instrument.  Plus, it will acquaint you with the process a little.

If you simply cannot pre-record, try to mentally think through the process.  Shoot the mental free throws, so to speak.  Imagine showing up to the studio.  Getting out your guitar.  Putting on headphones.  Listening to the engineer talking to you through the headphones from another room while trying to get microphone levels set.  The microphones are now your ears (when you have on headphones).  Weird, huh?

Schedule realistically

You probably don't want to record vocals at 7 a.m. or jump into a 5-hour session after a double shift at work.  Vocalists tend to suffer fatigue after about an hour on the mic, if not sooner.  And although you can play guitar for 6 hours straight, your ears might not be such good judges after that fourth hour of critical listening.  There's all kinds of peculiar factors to consider when scheduling a recording session.

Plan on at least half an hour to set up drums and drum mics.  If you're not picky, that might be all you need.  If you are picky, plan for longer.  I usually spend numerous hours on drums setup when it's for an album, but once they're set up, we can track song after song until the drummer is fatigued.  If there's still more drumming to be done, leave the drums set up and work on something else until he/she is ready to track some more.  

Never move past the drums.  If you're multitracking one instrument at a time, get drum takes for that song as early as possible, preferably first.  Trust me.  Most other instruments can overdub pretty easily, but drums can get tricky.  Plus, a good drum take sets the foundation for everything else.

Plan to lay down all the songs for given instrument in one setting.  (See Figures A & B, below)  This doesn't mean you shouldn't take a break, but it helps to avoid setting up microphones for that instrument (again) or trying to match how it sounded earlier.  Go back and forth between, say, acoustic guitar and bass tracks.  Those two instruments will probably not be using the same microphone (or performance space), so everything can be left set up for each one while the other is taking a turn tracking.  Vocals versus lead guitar.  Background vocals versus tambourine.  You get the idea.
Fig A: More time is spent setting up and tearing down than is spent actually recording the instrument.  In this example, only one seventh of the time is spent actually recording.
Fig B: More time spent recording than setting up and tearing down, even with numerous breaks for resting or to listening to playback.

Live vs Overdubbing

Most sessions are a combination of "live" recording and overdubbing.  

Typical session:  The drummer records his/her part while the guitarist and vocalist (in a separate room from the drummer to minimize bleed) record a "scratch".  Once the drummer gets a good take, other instruments can starting laying down their tracks:  bass, guitar, keys, vocals, background vocals, auxiliary percussion, horns, etc.

A lot of times, the band will lay down all the basic tracks "live", meaning they do them at the same time.  This usually means the drums, bass guitar, rhythm guitar, and maybe another instrument or two record their parts at the same time.  It keeps the energy high but it usually means a little bleed and everyone playing their part correctly.  If one person messes up, the take is ruined.  This is why you only track "basic" tracks together, ones that are pretty easy to get right with minimal chance of error.  
If the band really needs the vocals and/or the guitar solo to capture the real feel of the song, those two parts can be performed in an isolated area to prevent bleed, then they can be re-recorded later.

Terms

These are some common terms you'll probably encounter in the studio, along with a brief description.  If you want more detail, Google.
  • Video:  One Verse One Song https://youtu.be/M5tm5dmI6l4?t=62 This is a video I made some time ago to show how one person can use multitracking to record an entire song alone with one microphone and different instruments.
Autotune - correcting pitch with software/hardware.  Ideally, this is never used, but occasionally a singer hits a sour note in an otherwise perfect take, and autotuning can correct that.  It can also make you sound like a robot.  (Samples:  https://www.youtube.com/watch?v=WzUHKagIhoA )

bgvox - background vocals

Bleed - sound from one instrument is picked up by a microphone set on another instrument

Cans (aka Ear Goggles) - headphones

Clipping (aka Overmodulation, Distortion, Fuzz) - when the signal is too strong.

Comping - splicing together mulitple takes.

Master - a post-mixing process applied to final audio mixes before publication

Mixdown - to blend the recorded tracks together and export to a single file that can be played back on consumer audio equipment

Mixer (aka Console, Desk, Blender) - the device that is used to set the volume level of multiple sound sources at the same time.  It's the heart of every multitrack studio.

Overdubbing - recording an instrument against previously-recorded tracks.

Pop filter - keeps plosives and spit off the mic diaphragm.  When you sing close to a microphone, certain words result in a burst of air from your mouth, and when it comes through the speakers, it sounds somewhat akin to thunder, and not in good way.

Punch In - To record over a segment of a prior take by going into RECORD mode as the track is being played.  The machine is taken out of RECORD mode at the end of the segment (be left playing) to "punch out".  Punching in is typically done to get the performer "in the groove" by letting them perform along with playback as opposed to recording from STOP.

Red Light Syndrome - psyching yourself out.  Some people tense up when they know they're being recorded.  Don't.

Reverb (short for Reverberation) - the decaying residual sound after a sound occurs.  If you record in a small room, you can make it sound like a cathedral by adding reverb.

Scratch track - a guide track (to be discarded later).  When multitracking from the start, a scratch track is recorded as a rough sketch of the song, often just a guitar and vocals, but whatever is suitable to cue future players that they know where they are in the song.

Slapback - a fast echo.  Think rockabilly guitar or Elvis' vocals.

Take - a specific recording.  "You got a little raspy on that last take.  Let's take it once again and see which take we like best."

Track - 1) v. to record.  2) a single lane of audio.  3) a recorded song.  4) an audio file

Tracking - recording

Vox - vocals

XLR [extra long run] (aka. a microphone cable)  This is the cable used with most microphones.  It has a barrel and three pins.

July 11, 2018

Mastering

What is mastering?

Mastering is a final process for preparing your audio before submitting for distribution.  It is essentially the final step to ensures that each track on your album sounds about the same level without having to constantly adjust your volume or EQ when listening back to it.  Think of a song you hear on the radio that is predominantly acoustic guitar, then it is followed by a song that has drums, electric guitars, synths, and tons of other sounds, and yet the radio doesn't need to be adjusted for volume.  Or better yet, a single song that contains massive contrasts in dynamics and instrumentation, such as Queen's "Bohemian Rhapsody".  (Freddie sings softly over a piano at the beginning, but later on, he's wailing over a guitar guitar solo with bass and drums and chorus of voices backing him, and it all sounds perfectly fine without adjusting the volume or EQ.  That's what mastering does.)

Should you do it?

There is a lot of debate over the necessity of mastering and its benefits, or even detriments.  Shouldn't a properly-mixed song be complete upon mixdown? Personally, I recommend mastering for a few reasons.  One easily overlooked aspect is that mastering puts a fresh set of ears on the project, so an engineer in one studio might hear offensive tones that were not prominent enough to notice in the other studio.  Even after mixing, all the songs on an album might have different volumes, and mastering "fixes" that, makes the album more cohesive. It also ensures that you get the maximum volume.  Granted there is some loss of dynamics, but one thing to keep in mind is that most of your listeners will probably be listening on poor-quality speakers (such as a laptop or dashboard speakers) or poor-quality headphones (like Beats) in a noisy environment (car, subway) from a low-quality source (MP3), so you probably don't want too large a dynamic range in the first place.  It would frustrate the listener.  Or if your song is played between two other songs by other artists, and their songs have been mastered, you probably want your music to "keep up" with theirs, at least in overall volume.

There are a number of other reasons to have your songs mastered after mixing, a lot of them overtly obvious and many of them more nuanced, but these are the usually the most dominant.

Where do I do it?

Unless you've recorded an album before, there's a good chance you've never even heard of mastering, much less know someone who does it, but your engineer probably has someone in mind.  Personally, I refer all my clients to Safe&Sound in Memphis because the engineer does good work, he's fast, and inexpensive.  Plus, we've worked together for over 20 years, so we have our own workflow efficiency of sorts:  I know how he wants his files processed and labeled, and he knows how I usually need them returned.  Once a song or album is mixed, I usually upload uncompressed audio files at high resolution for Safe&Sound to download, master, and re-upload for either my client or for me to give to my client.  Often, my clients never have to talk with the mastering studio; I simply pay the mastering studio (ie. Safe&Sound) then add that cost to as a line-item in my fees at the end of a session.

A word about vinyl.

If you plan to press vinyl, talk to the vinyl guys about it, ie. the people who are doing the actual pressing.  Usually, bands that press vinyl need to get two masters:  one for digital distribution (CD, USB drives, cloud services, etc.) and another for vinyl, usually through the actual vinyl pressing company via their in-house mastering.

June 27, 2018

Mixing a song

Mixing a song

You're about to mix the song you've been recording.  Unless you have a record label footing the bill, and unless you have serious engineering experience, and here are a few things I would suggest.

Let's hear it from the engineer

First off, let the engineer have a crack at making a mix.  At the very least, wait until he* has provided a "rough mix" before you start lobbing suggestions.  Why?

  • Chances are he has ideas, knows the material fairly well (since he's repeatedly heard every part through a microscope for the past how many hours?), and will be relatively quick since he knows what knobs to twist rather than conveying his ideas to someone and hoping they understand and know how to make it happen.  In most cases, the engineer is the only person who has sat at the monitors to hear every track as it has been laid down giving him or her a unique knowledge of what there is to work with when it is time for mixdown, and he's probably has developed some thoughts about the project along the way.  
  • Mixing your own material can be dangerous for the same reason you shouldn't proofread your own writing:  you read what you intended to write rather than what you actually wrote.  
  • You might be pleasantly surprised by what a third-party hand makes of your material.  I'm not saying you shouldn't voice opinions or suggest edits, but try to refrain from getting too picky about how the guitars are panned until you hear the result of what the engineer's mix sounds like.  

Wait.

If time allows, just wait.  Let it simmer.  It's so tempting to lay down that last take, all eager and ready to jump into mixing the project into a final product, but it's usually a good idea to wait a while between tracking and mixing.  (In fact, if you let the engineer have a go at it, you can kill two of these birds with one stone:  you get to wait while he mixes.)  During the tracking process, your mind is prone to focus on all the mistakes you just made and obsess over them, often missing out on listening genuinely to everything else, fixating on insignificant errors and missing egregious ones.  What I have found is a few days after tracking, I go back to listen and usually don't even notice things that were driving me crazy during the tracking sessions, however I do start to notice new things I didn't notice while tracking.  Fresh ears.

Set a deadline.

Songs are never done; you just stop working on them.  If you work on them too long, you'll likely produce the energy right out of it, so you have to strike a balance somehow.  There will never be an end to things you can find to fix; at some point you simply say "done".  If you're spending more time in post-editing than you are in actual recording, it might be wise to consider re-recording parts.

Kill the puppy.  

Be open to the idea of changing things if it benefits the whole project, even it it means cutting a really cool part, or adding one you didn't intend.  It's easy to latch onto an idea because of how it worked in your head or before the interplay of other instruments, but reality usually alters a few supporting facts and you might have to reassess the value of a part in the new context.  Maybe the guitar feedback sounds awesome, but does it muddy up the vocals?  Perhaps muting a couple of instruments in the verse (rather than adding a couple more in the chorus) will give the chorus more oomph.

Step back and try to be as objective as you can, hard as it may be.  Is the crux of the song the vocal hook, or is the song a showcase for the kick drum?  As a bass player, I know well that the role of my instrument (usually) is far from the focus.  I am trying support and not get in the way of the focal points of the song which are often the lead vocals and/or lead guitar.  As cool as it would be for me to slappin'-poppin' wanking, it might work better to simply play the root or [gasp!] do nothing.

KISS

Keep It Simple, Stupid.  There's a reason adages are repeated through the ages.  More and more I aim for less and less.  What is the fewest amount of instruments and least amount of processing I can use to convey the song?  Is this particular part necessary?  Does it really add to the wall of sound or merely add to the noise?

These are far from the only factors to consider, and they don't apply to all sessions, but they are things I encounter time and again.  Hopefully, this serves to benefit your next recording session and give you pause to reflect on how to get the most out of your studio time.

So, now that your song is mixed, it's time to get it mastered...

______________________________________________________
* "He" in the gender neutral sense.  A lot of audio engineers are women, and hopefully more each day.  Regardless, "he" reads better than "he or she", drawing less attention to the text and (hopefully) more to the content.

June 20, 2018

How much does it cost to record?

How much does it cost to record?

A question I am often asked is "How much to record a song?"
This is much akin to asking "How much is a car?"
I often reply with my hourly rate and then mention bulk rates for long sessions.
But the question still stands:  How much to record a song?
To answer, I really need the following.
  1. How long is the song?
  2. How many tracks?
  3. How prepared you?
  4. How picky are you?

Granted, there are plenty more variables, but these are some of the biggies, each one able to affect the duration of your session exponentially.

1.  How long is the song?

Obviously, a longer song takes longer to record because, quite simply, it's longer.  But what I've noticed over the years is that people's ability to play a perfect take diminishes at an accelerated pace over time.  Or, to put it another way, the longer a song is, the higher the chances the performer will make more mistakes.  For whatever the reason, if a performer makes a single mistake for, say, a 30-second take, they'll likely make two mistakes for 60-second take.  But if it's a 2-minute take, they're likely to make six, eight, ten mistakes.  It's as if "red light" syndrome makes it harder to recover from each mistake, and each one compounds the previous.

Even if rate of mistakes stays the same (for this example, 1 for every 30 seconds), a longer song requires more searching time to re-locate errors, find a spot to punch in (and back out!), time to run the rehearsal with the performer to make sure they know where the punch is, then the time to get a good take.  Plus, during the course of the original take, the performer's tone (voice, amp, settings, etc.) may have changed slightly, so they might have to make adjustments so their punch in matches the sound quality of the original pass, otherwise it might be noticeable.  For a long song, these punch-ins might be scattered around, taking a while to re-locate each one to fix it.  As with any creative process, there is a certain momentum needed to keep the process alive.  When you're constantly stopping to make notes, listen back, punch in, that momentum can easily be lost.

2.  How many tracks?

Multiply the answer to the first question by the number of instruments.  Drums + bass + rhythm guitar + guitar solo + vocals + backing vocals + whatever.  Sure, the tambourine track will probably lay down in one take, but it still takes the time to get out the tambourine, set a mic, and record for the length of the song.  (By the way, will you be doing a "scratch" track?)

For a lot of musicians, a recording session is the first chance to hear their song without focusing on what they're playing at the time, the first time they have the opportunity to actually listen as an audience member to what they're doing.  A lot of mistakes go unnoticed until you hear them in playback.  Is the busy guitar rhythm distracting from the vocals? Are the kick drum and bass guitar locking together?  Did an arrangement idea suddenly reveal itself when the keys were muted for second?  Sometime good ideas come up while recording, which is great, but can have a ripple out effect on the rest of the song.  As an artist, it's usually wise to maintain an open mind to explore new ideas, but it does come at the expense of time.

One consideration is that it takes less time to record each song if it is recorded in the same session with other songs.  I've had artists come in to record each time they have a song worked up, leaving days or weeks between sessions, which means if all these songs are meant to go on an album together, either the drums aren't going to sound the same song to song, or there's going to be a good deal of time and note-taking to ensure they match as well as possible.  I mention drums because they're usually the most time-consuming by virtue of what they are:  a lot of instruments being played together into a lot of microphones.  It's not uncommon for bands to spend a few hours getting sounds set up on just the drums.  But once they're ready, they can focus on laying down tracks.  Even if it only takes a half hour to set up drums, by the time you've recorded a dozen tunes one session at a time, you've spent 6 hours of studio time setting up drums.  (Notice how many successful musicians are actually successful business persons who simply chose music as their product.)

3.  How prepared are you?

This should go without saying, but one might be surprised how many people show up at the studio with no idea what they are about to do or who haven't rehearsed their parts.  A lot of people don't realize how sloppy or out-of-tune their performance is until they hear themselves on playback for the first time.  Many people never got a chance to rehearse their harmony vocal ideas because, well, they simply never had a way to sing along with their self.  There will always be curveballs in any situation, but try to imagine exactly what all has to be accomplished to achieve your recording goal.  Are you comfortable using headphones while you perform?  Do you require reverb in your monitor mix?  Do you have all the instruments to get the sounds you want?  Do you know how to convey what you want to accomplish to the sound engineer?

4.  How picky are you?

The Good Rule, to paraphrase David Miles Huber, is when you record a good musician with a good instrument during a good performance.   This should make for a good recording.

I've had a few performers show up to the studio with equipment far from "good" but expect to get "good" sounds, or they rely on recording numerous takes that I comp together to build a "good" take from best bits.  A clever engineer who knows his equipment can often improve upon the situation, but it rarely compares to a performer simply applying the Good Rule.  The Good Rule will usually save you time and money, and provide you with better results.

Keep in mind that "good" is subjective.  If you mean "good" as Hendrix' guitar tone, be prepared for disappointment if you're using a Danelectro Longhorn and a Fender Champ.  It's  beautiful combination in my humble opinion, but it's never going to sound like a Strat through a Marshall stack.

And although the era of Cut'n'Paste has made it possible to fix so many performance errors perfectly, how much time (ie. money) are you willing to invest to make that happen?

Conclusion.

These a far from all the factors in deciding how long it will take to record a song, and possibly not the most important, but they are a few that I have encountered many times over.

So, in short, if you have one 3-and-a-half minute song with standard instrumentation (drums, bass, guitar, vox, keys, a few harmony vocals, percussion), know how to play all your parts, and have realistic expectations for the sounds you will get with your voice and equipment, you can probably record one song in an hour if you're willing to trust the engineer's mic choices and placement.

And then it will be time to mix.


Recording is a Lie

At some point, every artist with whom I've worked has made some statement about "being true" in the studio. What is "tr...