Need advice on mastering a song

Hi. I have a 3 1/2 minute song which has 2 separate vocalists, a midi piano track with the main music, and a guitar track to supplement the midi piano. Can anyone give some practical advice on how to go about mastering this? I have Sonar Cakewalk Produced (currently Sonarx1b, need to install the latest patch) and Adobe Soundbooth CS4. In the past I've taken clips into Soundbooth to clean up breaths and background noise, then brought those back into Sonar. Adjusted all the volumes (since Cakewalk handles multitrack much better than soundbooth) and eventually exported a WAV of the whole mix to touch up / add effects in Soundbooth (which I find easier). I'm aware that Cakewalk is a much more powerful program, however, and would welcome some advice on how to do everything within this one program - and any other general tips about approaching mastering. I know you're only supposed to compress things once to avoid artifacts, and I didn't know if adding VX64_Vocal to my vocal tracks would mess things up if I then did a final compression on the Master Bus. Anyhow, all suggestions welcome . . . Thanks

for mastering work at around 84 db in the room
use a hi pass filter and cut everything off below approx30 hz before the compressor or limiter. frequently listen to pro recordings of other artists on the same monitors to not get "out there" in some wierd tangent that won't sound good on a different system, this just for starters.

  Thanks for the replies. 1) Not sure where I picked up that compressing twice was an absolute no-no, but glad to hear I was wrong (with the caveat of compressing gently and being alert for distortion). This will make things much easier since I can now get each track roughly into working shape and then touch up in the Master Bus. 2) Just used volume attenuation to reduce loud breaths in a couple of songs - easy, quick, and excellent results. 3) Will have to read up on hi pass filter and limiters again. Appreciate the help and happy to hear other advice / opinions

mosk


  Thanks for the replies. 1) Not sure where I picked up that compressing twice was an absolute no-no, but glad to hear I was wrong (with the caveat of compressing gently and being alert for distortion). This will make things much easier since I can now get each track roughly into working shape and then touch up in the Master Bus. 2) Just used volume attenuation to reduce loud breaths in a couple of songs - easy, quick, and excellent results. 3) Will have to read up on hi pass filter and limiters again. Appreciate the help and happy to hear other advice / opinions

Volume attenuation? How was this accomplished? If through the Process>Audio>Gain menu you may end up with a very abrupt volume differential during audio. IMO the best method to knock down unwanted breathing noise is to use volume automation like bitflpper suggested. That way you can create a volume slope that sounds natural. A clip gain envelope will also work fine. The important thing is to get that slope between loud and quiet volumes.

 Well, this is where theory and practice always seem to diverge. The song I'm working on at the moment has a guitar track and a bass guitar track both of which are fairly quiet and lowish decibels (not totally clear on the difference between volume and gain aside from need to avoid cranking gain up so don't get distortion whereas not sure if the same with volume). Anyway, I tried adding Boost11 to both the guitar and bass guitar tracks to get their volume closer to the midi piano (which I can make quieter through lowering volume) and the vocal tracks. I also added VX64 to the vocal tracks, Sonitus Reverb to a bus which I routed several vocals to, and another Boost to the MasterBus to tidy things up at the end. The volume from several tracks now wobbles up and down abruptly and unpleasantly. Didn't know if this is because my approach is flawed or the effects are chewing up too many computer resources (though I'm on Windows 7, 64 bit system running 64 bit Sonar Producer x1b (haven't installed c patch yet), with 24GB RAM. Do I need to render the effects somehow for smooth playback? I tried peeling off a few of the effects, and have tried using Sonitus Compressor in some areas, but not having much luck. Is there a stepwise fashion I should proceed in? Or something obvious that I'm messing up? Thanks

Compression is akin to painting - you get much better results by applying several thin coats than a single thick one.

Thanks for the continued advice. Any specific advice on Panning? In most of my songs I'll have two vocal tracks singing identical parts, as well as responsive parts in between the main vocals. I'll also have 1 piano track, one guitar, and possibly a bass guitar. I guess it would make sense to separate the main vocals from the 'responsive' vocals (which may answer questions or give commentary that addresses the lyrics sung in main part) Beyond that, however, I don't have any real sense on what to do (other than trial and error and a completely subjective choice about what sounds best). Thanks