Preparing Beats and Instrumentals for a Vocal Session

I work with a bunch of hip-hop artists and a few R&B singers. Most of the time they bring their own instrumentals to the studio instead of having me write music for them. I usually charge $300 or so to write and produce instrumentals for artists and there are 3 zillion kids with FL Studio using the title producer that will put something together for free.

The problem with free beats is that most of the time the quality of the audio really sucks.  Most MC’s are downloading instrumentals off of the web or the beats are coming in over email.  These are always compressed files which lack accuracy and sound quality.  OGG Vorbis files, MP3’s, WMA’s and Apple’s M4P’s or AAC’s all can sound pretty bad.  If you are starting a recording project, you want to start with the best quality audio that’s possible.  The following guidelines are intended to help people avoid releasing crappy sounding music.  Mix down your instrumentals using the following suggestions as a guide.

  1. Use full-quality uncompressed digital audio like WAV or AIFF files.  At the very least, these files should be 16 bit 44.1Khz stereo files.  I prefer to work with 24 bit files at either 44.1 Khz or 88.2 Khz.  The quality of the audio is much better and is easier to manipulate.  Using uncompressed files is the best way of ensuring that your engineer will be able to make a great mix of your songs.
  2. If you must use a compressed file-format, use FLAC (the Free Lossless Audio Codec) <http://flac.sourceforge.net/>
    FLAC is great because it is lossless, which means that even though the files are smaller than uncompressed files, they sound just as good as uncompressed files.  By using additional processor power you can make FLAC files even smaller.  In a series of tests that I did with my colleague Connor Smith, we discovered that FLAC was capable of shrinking our test file of uncompressed audio at 5.3 MB down to 1.6 MB without loosing any audio quality at all. FLAC files are sometimes small enough for people to email if they are short.
  3. Give the engineer stems. Stems are separate stereo tracks for each of the instruments in the instrumental.  For instance, you would have separate files for the drums, the bass, the rhythm instruments, the keyboards, the samples.  When you give the engineer stems they are able to mix the different instruments with the vocals.  A lot of the time the instruments block out the vocals in a mix.  If you send stems, the engineer can lower the instruments without lowered the drums and the bass. If you don’t bring stems, the engineer can’t leave the drums loud if the instruments are getting in the way of the vocals.
  4. If you have to use compressed lossy files, use the best possible quality that you can get.  OGG Vorbis, MP3, WMA, and AAC/M4P all offer the option of making higher quality files that are larger in size or smaller files that sound bad.  Here’s the audio choices going from best sounding to worst sounding:Ogg Vorbis (.ogg) is Open Source, Free and Awesome <http://www.vorbis.com/>
    Microsoft’s Windows Media Audio (.wma) <http://www.microsoft.com/windows/windowsmedia/forpros/codecs/audio.aspx>
    Apple’s Advanced Audio Codec (.aac or .m4p) <http://www.apple.com/quicktime/technologies/aac/>
    Mp3 (.mp3) MPEG layer 3 (Motion Picture Engineering Group) <http://lame.sourceforge.net/>
  5. Use the highest bit rate that you can use with all of the above audio formats.  I recommend a minimum bit rate of 256 Kbps for Ogg, WMA and AAC, but a minimum of 320 Kbps for MP3 audio.  VBR or Variable Bit Rate can be a little squirrelly, so to be safe always choose the highest quality option available.
  6. Find out if the engineer has the same software that the beat was created in.  I have FL Studio XXL so I can get FruityLoops files with the loop bundle and mix the  instrumental with the vocals directly.  It’s very likely that your engineer has software that can work with your format.
  7. If the file was ever a compressed file, you can never make the quality better.  For instance, if a beat-maker emails you a beat as an MP3 and you then convert it to a 16bit 44.1 Khz WAV file, it will never sound better than the MP3 file.  Never try to burn a CD with MP3 versions of the music.  You are just making the problem worse.

Please don’t hesitate to ask questions about file formats.  I can also help you to get great mixes either with advice or you can send me your projects to work some magic.

4 Responses to “Preparing Beats and Instrumentals for a Vocal Session”

  1. Aaron Says:

    So have people sent their beats using fl studio to big time producers such as timbaland,dr.dre or the dream or can they send them?

  2. Hendrik Gideonse Says:

    I don’t have personal information about what types of files have been sent to the “big time producers” but I suspect that they use their own tools to produce their own instrumentals. What I find all the time is that clients send their music to me as FL Studio Files with the samples bundled into an archive. The two most common software tools for stand-alone so-called beat production are Reason and FL Studio (FruityLoops). If I were Dre and was working with an outside client that wanted to use their own instrumentals, I would just have them sent to an assistant engineer to be dumped into whatever format I wanted.

  3. majstor Says:

    I was wondering about “stems”… which volume and panning I should apply when sending to engineer, and do I need to cut the part that repeats, or should I leave it (with blank parts) with length of actual/complete beat?
    Should I export each instrument/channel separately?

  4. Hendrik Gideonse Says:

    Usually stems are stereo tracks (or a pair of mono tracks) for each of the main groups of instruments. Drums, Bass, Chordal Instruments, Lead or Melody Instruments, Backing vocals and lead vocals. Often times with hip-hop the snare and the kicks are made separate in case the mix engineer wants to bump them up or down a dB or two. When I am mixing I prefer to have every instrument as a separate track (either mono or stereo) so that I can have as much control as possible with the mix. Most of the time you should send each track or stem at full length. Don’t assume that an engineer will know where to loop sections. If you’ve asked for a remix then everything is up for grabs and can be changed.

Leave a Reply