Soundfly

Home for the Curious Musician

A Guide to Best Practices for Seamless Audio Collaboration in Your DAW

recording-collab-header

Once upon a time, recording music worked like this: An arranger made a score, a copyist made parts, and a group of musicians got together to play while the tape rolled. The rhythm section’s pulse or the conductor’s baton were all that was needed to keep everything in sync until last note decayed and the master tape was labeled and stored.

Once upon a slightly more recent time, when multi-track “Sel-Sync” heads made overdubs feasible, musicians didn’t need to play together anymore. This made everything more complicated (and eliminated some musical give-and-take), but as long as there was a clear count off at the beginning, everything generally lined up fine (even when SMPTE timecode allowed multiple tape machines to be locked together). To keep track of such complex sessions, rigorous documentation and organization became the norm. If a collaboration with someone in a different city was required, copies of the of tape would have to be couriered to them at another studio. Because most processes were destructive (or at the very least, not without risk), fastidiousness and care were imperative — careers depended upon it!

bd-tapes

Today, things are at least theoretically more flexible and forgiving. The capacity for unlimited overdubs with sample-accurate time sync is taken for granted. Collaboration with people in other cities or even other countries can be rolling as quickly as an FTP upload and download. Overdubs can often be done at home. Few processes are destructive (and “undo” is usually an option), rendering the entire recording process less fraught with anxiety and stress. Decent recording equipment is less expensive than ever.

But these largely-positive developments are not without downsides. The feeling of lower stakes often results in less care and organization. Unprecedented access to low-cost equipment means that modern musicians are often expected to be recording engineers as well, despite lacking the requisite experience and training (not to mention proper facilities). This results in a lot of avoidable mistakes that lead to wasted time, frustration, and even some strained professional relationships.

So let’s talk about how to maximize what’s good (and minimize what’s bad) about recording and collaborating remotely in the 21st century!

+ Read more on Flypaper: “4 Tips for DIY Drum Recording”

Session and File Hygiene — Getting Started

A common problem area for the self-recording musician is file management — the art of naming, organizing and ensuring that files are in the correct format. Of course, many producers are not entirely innocent where this is concerned, either, so clear communication of needs and expectations is a good first step.

Some producers may wish to send you a complete session folder. But since this requires that you both use the same DAW (and since session folders often contain the bloat of many unused audio files), this is usually not ideal. Instead, request either a two-track rough mix, a set of stems, or a complete set of consolidated, zero-bounced files that you can import into your DAW. Ask for something resembling the final arrangement as closely as possible, to ensure that your part makes sense in context. Make sure you’re given the session’s sample rate and bit depth, and set up your session to match.¹ If the track uses programmed drums or was recorded to a click, obtain the exact tempo marking from the producer. Ask that the consolidated files start at timestamp zero of their session (even if this means several empty measures at the top) to ensure grid alignment in your own.

Danger! The “HK instrumental rough mix” to which I’ll be tracking isn’t fully aligned to the left of my session. If I were to overdub to this and then export files that extended all the way to the beginning, those parts would play back slightly later than I intended when imported into my collaborator’s session. Double-check this early and often!
Danger! The “HK instrumental rough mix” to which I’ll be tracking isn’t fully aligned to the left of my session. If I were to overdub to this and then export files that extended all the way to the beginning, those parts would play back slightly later than I intended when imported into my collaborator’s session. Double check this early and often!
Much better! Notice that the “HK instrumental rough mix” is now aligned to timestamp zero (or bar 1, beat 1).
Much better! Notice that the “HK instrumental rough mix” is now aligned to timestamp zero (bar 1, beat 1).

Once you have what you need, create a new session in your DAW. Import all of the audio into your session, making sure it’s aligned to the very start of the grid at timestamp zero. Set your session’s tempo to the one provided by the producer, so your grid reference will match theirs (allowing you to set up your own click track, if necessary). You’ll occasionally be given a track where the song starts right at the beginning of the session, leaving no space for a count-in. In this unfortunate case, you may wish to move ALL audio files rightward one or two measures on the grid, but do not forget to move everything (including your newly-recorded part!) back to its original location later!

When ready to add your part, create a new track and immediately (before recording anything) name the track with something clear, concise, and descriptive (e.g. “Rhy Gtr L,” “Bridge Shaker,” etc). This tells your DAW what to name the associated audio files. We want to avoid ending up with a bunch of files with names like “Audio_01.02.wav” or “Audio_10.14_L.wav,” as such file names will lead to a frustrating exercise in forensics down the road (particularly if something comes up missing).

Now that your session is set up, you’re ready to arm your track and hit record!

+ Learn more on Soundfly: Record pro-quality demos and songs from your bedroom studio with our free home recording course Demo Recording 101!

Audio Hygiene — Getting Sound

We’ve come a long way since the early days of home recording. Contemporary prosumer interfaces, while not totally comparable to bona fide professional equipment, are getting better all the time. The most popular units usually have solid A-to-D conversion and usable mic preamps which are, at the very least, unlikely to be the biggest obstacle to meeting expectations. Most of what you’re able to accomplish sonically will depend on your skill and experience as an engineer, so let’s leave aside the aesthetics for now and focus on avoiding the kind of technical problems that could render your work unusable.

Let’s handle the low-hanging fruit first: Double check that your project’s sample rate matches that of your collaborator’s. Sample rate conversion has gotten better over the years, but it still always comes with a subtle, though audible compromise. Even if your collaborator isn’t fidelity-obsessed, unnecessary sample rate conversions are an entirely avoidable degradation.

“A Quick Guide to Microphones” from our free course, Capturing and Warping Samples in Ableton Live

Once you’re getting signal to your mic preamp or DI, set your gain/level appropriately and conservatively. With 24-bit digital, there’s no reason to risk coming anywhere close to clipping, as the dynamic range of the format is extremely wide. However, as a courtesy to your client or collaborator, it’s probably best to avoid excessively low levels, too. Lots of gain applied later requires an extra step, and can sometimes come with a noise penalty, so shoot to have most of the activity right near the middle of the meter’s range. This is not exactly set-and-forget — during the course of your recording, keep an eye on the meters and clip indicators, as it’s not uncommon for players and singers to get a bit louder as they warm up.

Speaking of noise, most home recording environments are quite noisy, both in terms of ambient rumble (HVAC, appliances, room noise) and interference (EMI/RFI from cell phones, bad power, fluorescent lights, etc).  Take steps to minimize this as best you can, as these sorts of problems will become exacerbated at mix time. Remember, the noise floor of multi-tracked layers compounds with each additional stack (and any compression will bring up the noise as well), so close those windows, turn off the A/C, and kill the fluorescent lights.

If you’re recording a DI signal with an amp simulation, make sure you print and provide the sound of the sim. If you like, you can print your performances to two tracks — one with the raw direct sound and one with the simulated amp — for safety. Choose the simulation carefully and make it sounds exactly like you want. Your playing dynamics will unconsciously respond to what the virtual amp is giving you, so changing it later doesn’t always work well. If you’re monitoring effects that you feel are integral to the part, you may wish to print those, too. It can be safest to print effects like reverb and delay to a separate track (reverb is seductive and tends to make us want to go overboard). But if you’re confident something is an essential part of the sound (like spring reverb on a guitar, or a delay pedal that defines the part) it’s often best to take the risk and print what’s working.

+ Read more on Flypaper: “The Ins and Outs of Recording Your Guitar ‘Direct In'”

Session and File Hygiene Redux — Getting Finished

Once the part is recorded, we have to ensure that your collaborator can import it into their session without tearing their hair out in frustration. If you’ve done any punch-ins or edits, make sure to go through and finesse the edit points carefully, placing crossfades at each one to avoid pops or tics. Check each crossfade, really listening analytically (headphones are a plus) to make sure there are no unnatural decays, double breaths, chopped-off sustains, or ghosted notes. If you hear anything amiss, adjust the edit point until it’s perfect — you’re about to commit to these edits forever! If a clean edit proves impossible (and it wasn’t a once-in-a-lifetime performance), it’s probably best to re-record that section and do a better punch. For this reason, I actually like to check punches and edits as I go (rather than saving the clean-up for the end), but you’ll determine your own best workflow.

Here’s what I mean:

Be careful with applying fade-outs to tracks you’ve recorded. When in doubt, let the mix engineer handle the tops and tails. If you do choose to do this as I’ve done here, err on the side of longer, later fades—once exported, your collaborator could always “chase” your fade with a shorter or earlier one, but they’d be unable to go the other direction. Always leave as much information as possible, within reason.
Be careful with applying fade-outs to tracks you’ve recorded. When in doubt, let the mix engineer handle the tops and tails. If you do choose to do this as I’ve done here, err on the side of longer, later fades — once exported, your collaborator could always “chase” your fade with a shorter or earlier one, but they’d be unable to go the other direction. Always leave as much information as possible, within reason.
If comping from multiple takes or cleaning up punch-ins, find good edit points and apply short crossfades. Lengthy crossfades increase the likelihood of flams, ghosts, and audible chorusing of the noise floor, so I like to keep mine as short as reasonable.
If comping from multiple takes or cleaning up punch-ins, find good edit points and apply short crossfades. Lengthy crossfades increase the likelihood of flams, ghosts, and audible chorusing of the noise floor, so I like to keep mine as short as reasonable.
Listen to make sure the edit point is completely inaudible. If multiple mics were open in the room at once, as with this horn section, the resulting bleed or spill will mean it usually sounds the most natural to edit and crossfade in groups.
Listen to make sure the edit point is completely inaudible. If multiple mics were open in the room at once, as with this horn section, the resulting bleed or spill will mean it usually sounds the most natural to edit and crossfade in groups.
You can see within the crossfade a lump of audio representing the slight ghost of the singer’s breath. Listening carefully in headphones, you could just barely hear the flam of two faint breaths, which sounded very unnatural. If I had consolidated the file with this error, it would’ve been permanently embedded within the track.
You can see within the crossfade a lump of audio representing the slight ghost of the singer’s breath. Listening carefully in headphones, you could just barely hear the flam of two faint breaths, which sounded very unnatural. If I had consolidated the file with this error, it would’ve been permanently embedded within the track.
Moving the edit point slightly later on the timeline solved this problem— there’s now one clear, natural-sounding breath, and the edit point is inaudible, even when soloed in headphones.
Moving the edit point slightly later on the timeline solved this problem — there’s now one clear, natural-sounding breath, and the edit point is inaudible, even when soloed in headphones.

+ Read more on Flypaper: “10 Tips for Making More Realistic MIDI Music”

Comp together your best raw performance and leave any processes like vocal tuning to the producer. However, if you do give in to your ego and try to sneakily pitch-correct, never apply multiple tuning passes to one file, as this degrades fidelity severely, and can start to sound very unnatural. Always go back to the original raw file and start again.

When all editing is accomplished to your satisfaction, consolidate the file (do a save as or new playlist, if you wish), making sure that your file starts at timestamp zero. If you had to move the audio rightward earlier to give yourself a count-in bar, now’s the time to move it all back to the original zero point!

Finally, it’s time to export. If your file is a mono source (like a vocal), export it as a mono .wav file. If it’s a stereo file (like a stereo keyboard), then export it as a stereo .wav file. Never export or bounce mono tracks as stereo files, as this creates confusion and extra steps for the producer or mixer. After the files are exported, check once more that everything is clearly and logically named (you might wish to remove any auto-numbered appendages from the end of the file names).

Note that the “acoustic” track here (purple) extends all the way to the very beginning of the session, at bar 1, beat 1. This file, when exported, can be seamlessly imported into my collaborator’s session. The four horn tracks above it, however, do NOT extend all the way to the beginning of the session, so if I export them as-is, my collaborator would have a frustrating time trying to make everything line up.
Note that the “acoustic” track here (purple) extends all the way to the very beginning of the session, at bar 1, beat 1. This file, when exported, can be seamlessly imported into my collaborator’s session. The four horn tracks above it, however, do NOT extend all the way to the beginning of the session, so if I export them as is, my collaborator would have a frustrating time trying to make everything line up.
Zooming out, we can also see here that the horn tracks consist of a few different clips, with some edits and some empty space. We don’t want to export these clips individually, so we need to consolidate them into whole files that extend all the way to the beginning of the session.
Zooming out, we can also see here that the horn tracks consist of a few different clips, with some edits and some empty space. We don’t want to export these clips individually, so we need to consolidate them into whole files that extend all the way to the beginning of the session.
In this image, I have consolidated the four horn files so that they extend all the way to the beginning of the timeline, just like the “acoustic” track (purple). These can now seamlessly be exported and imported with sample-accurate timing.
In this image, I have consolidated the four horn files so that they extend all the way to the beginning of the timeline, just like the “acoustic” track (purple). These can now seamlessly be exported and imported with sample-accurate timing.

Always send the highest resolution file you can (usually a 24-bit .wav file), and never send a lossy format like an .mp3! The .wav files are typically too large to send via email, so use free services like Dropbox, Hightail, Google Drive, or WeTransfer to deliver them.

While you’re waiting hear the final mix, you can feel good that you’ve spared your collaborators the frustrating agony of needless detective work and salvage missions. A little care on your part allows them to focus their energy on making your collective work sound the best it possibly can. I promise it’s worth it.

Notes

¹ If, for some reason, the master session is 16 bits, you may wish to match the sample rate and record at 24 bit, either dithering down to 16 bit if you know how, or delivering a 24 bit file (leaving it to your client to apply dither).

Get the top Flypaper articles delivered straight to your inbox once a week.

Brad Allen Williams
Brad Allen Williams

Brad Allen Williams is a NYC-based guitarist, writer/composer, producer, and mixer. He has worked with artists such as Cory Henry, Bilal, Kris Bowers, and José James on stages around the world. As a leader, his 2015 album Lamar received critical acclaim in numerous mainstream outlets including The New York Times. In addition to being an in-demand session player, he collaborates with many artists as a writer, producer and mixer, working primarily from his home base The Bunker Studio in Brooklyn, a two-suite facility he helped to build. [Photo ©2017 Deneka Peniston]

  • Roupen

    In the middle of a recording, if there’s clipping, do you record the whole track from beginning with a lower gain, or you just adjust the volume of the already recorded part and continue with lower gain?

    • John Hull

      Hey Roupen. You could either turn down the pre and rerecord the whole passage, or turn down the gain and try to punch in.

    • brad allen williams

      It depends somewhat on what’s going on– If there’s a single clip in the middle of a stellar performance, I will listen and decide if the distortion is audible and/or severe enough to warrant even fixing it. If the clipping is extremely transient and I can’t hear it–and again, it really is just ONE spot–I might just reduce the level of the entire recording once finished, and let it pass.

      If it’s audible and bothersome, I would be inclined to re-record, or another solution would be to try to punch in that area and move back off the microphone or lighten my touch in that area, to see if I can avoid the clip but still sound natural as though part of the same performance.

      If there’s consistent clipping throughout the track, I’ll usually go back and start again. That’s a real problem with gain staging that needs to be addressed, and level-matching is generally more trouble than simply re-recording.