A more detailed account
All mastering was done within Izotope Ozone with the exception of noise reduction in Adobe Audition and De-essing using dB Audioware. As this was for the benefit of people with limited budgets, I confined the tools used to easily accessible software.
The song was a good example for mastering as it exibited many typical problems.
The vocal was very sibilant, some background noise exibited, guitar parts too loud, vocal buried, Eq issues, slightly dry mix, uncontrolled levels in most frequencies. Additionally there were a lot of clicks in the vocal which was highlighted as the levels were increased.
I found the greatest challenge was to improve the high frequency response while controlling the vocals harshness. In the end I feel I compromised slightly with a master that is a biased to a top and bottom end exaggeration. Given the substance of the song this was probably the best approach to enhance it's musicallity.
Now you know what I was dealing with, here's how i did it.
- Converted to 96kHz/32 bit Wav. file
- Added 1 second leader to start of song
- Ran file through noise reduction software to negate the obvious hiss at end of the track.
- Raised the overall level with intelligent limiting, around 6dB with a medium release time to maintain some smoothness, particularly in the snare. Already a loud mix any further increass would have done too much damage to the depth.
- EQ'd the mix, reduced the bottom end to remove boominess, cut high frequencies considerably and also around 290hZ to improve clarity.
- Used tube enhancement on bass with a 3ms delay to increase punch. A little on 700hZ-10kHz for air.
- Multiband compression to controll various aspects of mix. Tightened bass. Expansion on lower mids and high frequencies. Clamped the upper frequencies and deep but minimal ratio compression on upper mids to help vocal and drum presence.
- Stereo imaging was reduced in lower frequencies and opened up in upper and low mids, this helped with clarity somewhat but applied quite minimally. The overall stereo image was quite wide to begin with but seemed acceptable in mono and out of phase so I didn't toy with it a lot.
- Mastering reverb was applied in VERY small quantity to help gell the mix and improve the drum sound. Around a 24ms pre-delay using a moderate sized room and fairly heavy damping. Reverb is dangerous o apply as it's easily overdone and can dull a mix. Rather than add an effect I was looking for a tightening of the overall sound without hearing the reverb.
- Dither was applied prior to converting out of Izotope. As bitrate was higher than 16 this is a must.
- De-essing the mix was next. I focussed on the overall sound during the earlier processes knowing that this step was required. It would have been impossible to control with EQ and compression without killing the overall sound. Around 24dB of cut using a triggered lowpass of 3.36kHz. Detect was around 6.2kHz and 2.6kHz wide.
- Click removal of vocal track was left till last as all previous actions determined which parts had to be fixed.
- Mix was then converted to 44.1 / 16 bit before saving as an mp3.
Mastering is similar to mixing in that it can be highly subjective at times. Although you are trying to create something that is balanced on all playback systems (impossible) there is a degree of decision making involved in respect to maximising the appeal of the song.
http://www.lightningmp3.com/live/file.php?fid=3345