Check the track start. Make sure there are no unwanted pauses or noises before the music begins, but avoid trimming the track so tightly that every breath or sound before the music starts is cut away. Generally, the music waveform should start 50 - 500 ms after the beginning of the sample. This start is usually shorter for pop/rock and longer for classic tracks.
Check the track end. Make sure it isn't cut too early, especially if there's some ambience or delay at the end, and use a nice fade-out even if it's only for the background noise.
On many pop/rock tracks there is a longer fade out on the chorus. It's always better to do this at the mastering stage and not the mixdown, otherwise you might get digital noise at the end of the fade out.
With fade ins and outs, exponential and logarithmic curves sound more musical than linear ones. Use inverted S-curves for longer fade outs.
Adjust the track volume and L-R balance in relation to the other tracks. Check that the L and R channel have about the same peak and RMS levels and that the track has the correct loudness in relation to the others.
You should never normalize every track on an album. By doing this, the level of a quiet track would be boosted much more than a loud one, and the quiet track would end up sounding louder/closer than the loud one.
Adjust subtle differences in balance between the different frequency ranges with a mastering EQ (see EQ tips below).
If the track sounds too narrow or too wide, adjust the stereo width using a stereo imager or a psychoacoustic processor.
If you're having trouble refreshing a dull recording, use a bass/treble enhancer or exciter carefully, but try with the EQ first.
If you want to remove undesired tape hiss, try a denoiser, possibly with "fingerprint" function to identify the exact spectrum of the noise to be removed. Check carefully that you don't cut important high frequency parts of the signal. It's better noisy than dull.
Adjust the dynamic range if it doesn't fit the final medium and/or the final listening environment. You want to avoid the listener having to adjust the volume constantly as some parts are too quiet and some too loud.
Compare the overall sound also with other productions, such as reference CDs, to make sure you're within the range of possible variations.
If you can't get your master to the volume you want, try scooping out some of the low mid ranges. It can be really hard to get loud if you're heavy between 400-800Hz. Try cutting on individual layers throughout your mix and think about all that added headroom you can create with small, eﬀicient cuts.
Watch out for troublesome areas when you master. Subtle EQ dips around 250Hz, 1000Hz, 4000Hz and 12000Hz can make a huge diﬀerence in how clean your track sounds when played loud.
Using a reference is important, but using a reference for speciﬁc parts of the frequency spectrum is even more important. Only solo the parts of a reference that you're trying to match, then use all of your mixing and mastering tools to recreate that section in your track. To get the best result, solo your reference below 200Hz, then do the same to your track and keep working until they sound the same.
If your master is lacking sparkle, try boosting the sides above 7000Hz in your master. This will make the track feel bigger and brighter without being as hard on the ears as a traditional EQ boost. You can also cut the lows from the sides below 150-200hz to clean up muddiness.
Automate the volume going into your drop to help increase the punch and impact of the drop. Try automating the master volume before the limiter during buildups. Try a combination of volume, EQ & reverb automation on your build. Use volume dips, low cutting with EQ, and automating the reverb wetness up.
Excitement is very important when mastering. As well as exciting the high end, try exciting the lows for some added warmth. You can even use a mid/side exciter to have ultimate control over your master. Exciting the sides can be done much more aggressively than the mids in the higher bands.
Always check your master on every sound system you can find. Take exports/bounces to your car and really judge your master with some detailed note taking. It can often take at least 2 or 3 attempts before you get the results you want.
Equalizer (EQ) Tips
When adjusting the low frequencies, too little makes recordings sound thin and weak, while too much makes them sound boomy and distorted on most speaker systems. To get the right results, you can use a mild low shelving filter or a wide peak parameter EQ centered around 50 Hz.
On the other hand, when adjusting the high frequencies, too little makes recordings sound dull and unclear, especially at low volumes, while too much makes them sound harsh and unpleasant, especially at high volumes. Again, you can use a mild high shelving filter or a wide peak parameter EQ centered around 16 kHz.
Check the balance between the bass frequencies and the high ones. In rock/pop/electronic music, the usual difference between the loudest bass frequencies (50 - 100 Hz approx.) and the high frequencies (approx. 10 kHz) is 30 - 40 dB. If the spectrum analyzer shows a difference less than 30 dB, the track can sound quite aggressive, harsh and thin. However, if it shows a difference more than 40 dB, the track can sound quite warm, but also a little muddy and dull.
If you want more power, try boosting 16 - 60 Hz, but listen with a system that responds down to 16 Hz, or with good headphones, otherwise you might push it too far! Remember that energy in this range can "chew up" a lot of the available dynamic.
If you want more BD punch, try boosting 50-60 Hz.
If the bass is too loud, try cutting 100-150 Hz.
If the sound is muddy, try cutting the 200-250 Hz range. Muddy sound is much more obvious if you're listening on small multimedia speakers or portable audio.
If the general sound lacks a warm, full quality, try boosting the 250-400 Hz range. Alternatively, you can cut this range if the general sound is muddy.
If the vocals are getting lost in the mix, try enhancing the range where the vocals have the most power (800 Hz - 1,5 kHz).
If the guitar sounds too sharp, try reducing the 2.5 - 4 kHz range.
If the mix is not clear (you want more tranparency), try boosting the 2-3 kHz range. Be careful though, because the human ear is most sensitive to the 2-4 kHz range. If you boost this range too far, it can become very hard to listen to or even cause hearing damage at high volumes.
If you want more focus or impact from the vocals, try boosting around 5 kHz but take care of the "S".
If the "S" and "T" are too sharp, try cutting the 6-7 kHz range or use a de-esser. However, it's better to use a de-esser during the mixing process, and even better not to position the microphone directly in front of the singer when recording.
If the percussion sounds harsh and metallic, try cutting 10 kHz and boost over 12-15 kHz a little bit. This produces a more elegant sound
If you want more spark and finish from a recording, try adding some air with a shelving at 16 kHz.
It's often better to use phase linear EQs. They won't modify the shape of low-frequency waveforms that would reduce headroom and/or cause clipping. However be careful using phase linear EQs, because they can sometimes cause a annoying pre-ringing, which is a percussive sound which quickly fades in before the hit.
If the overall dynamic range (the difference between quietest and loudest passages) is too wide, you can use a compressor at moderate compression ratio (1:1.2 to 1:1.6), relatively low threshold (-20 or even lower) and soft knee, to adapt the overall dynamic range of the recording to the final delivery platform, such as streaming or CD.
If the signal peaks are too loud, try controlling them using a compressor with high ratio (1:2 to 1:4), high threshold (-10 or higher), hard-knee and fast attack/release times. It's also an option to use a limiter instead.
If you want to adjust the level of different song parts such as the intro, use a volume curve instead of a compressor.
Classical and jazz tracks often use very little compression in mastering, especially if they already have compression on single channels and groups during mixing. Tracks such as pop, rock and dance often need some additional sum compression, but don't overdo it. If you do, it can sound terrible over the radio, where lots of compression can be added, and will lose all of its punch.
If you have a good analogue compressor and AD/DA, you can go in and out of your DAW, using that instead of your favorite plugin.
The mastering process wouldn't be complete without multi-band compression. This allows you to control compression across different frequencies in your track. You can set attack, decay, ratio and makeup gain across multiple bands. When used correctly, they help you reach a higher perceived volume without distortion.
They also be used instead of a traditional EQ to balance lows, mids and highs in a track.
The Limiter should be the last effect in your master bus chain. Use it to control short peaks and avoid clipping of the signal.
Set the maximum level as -0.1 or -0.2 dB FS, because some older DA converters cause distortion if the signal reaches 0 dB FS.
If you use limiting up to 2-3 dB limiting, this is virtually undetectable and will give you more loudness. However check that the loudest passages have no distortion.
To avoid audible pumping artifacts, set the release time as short as possible, however very short release times can sometimes cause unwanted distortion.
Avoid clipping on the master out. Don't try to boost your track's volume to the loudest level, but instead aim for the best sound quality at the loudest level without distortion or clipping.
If you're looking for an aggressive sound (such as drums) during mixing, it can be a good option to have clipping on single tracks while keeping other tracks (such as vocals) clean. In other words, single track clipping or even drum/percussion clipping is fine, but avoid master out clipping.
Moderate clipping is sometimes used instead of limiting when mastering through analogue effect chains. Although this can sound fine on high-end AD converters, it often sounds bad on cheap ones.
Mastering can't fix everything
As in mixing, remember there are some things that cannot be fixed by Mastering.
For example, if you have two instruments in the same frequency range and one is too loud or too quiet, there is very little you can do to rectify it.
If the bass is too loud and covers the bass drum, nothing can bring this bass drum back to life and make it punchy. However, if the bass drum is too loud and the bass is too quiet, you could try to compress the bass drum and increase the bass (low level) signal.
If the mixdown was already distorted or clipped, it's almost impossible to fix this later in Mastering. In this case, you'll need to redo the mixdown.