Describe your mixing process

[Kinh]

Get the mastering track on first, boost level to max and RMS as close to -4 as I can. If I cant then I do parallel compression for most of the tracks, squash the dynamics on master until I'm at -4 and the loudness matches some guide track.

Then I fix all the distorted tracks with eq and what not till overall level and RMS is right.

This is horrifying. Music deserves to breathe.

I know, it sucks but everything is loud now. The alternative is your tracks sound weaker than everyone else's.

[recursive one]

The alternative is your tracks sound weaker than everyone else's.

When listened through crappy iphone speakers

[AdvancedFollower]

Get the mastering track on first, boost level to max and RMS as close to -4 as I can. If I cant then I do parallel compression for most of the tracks, squash the dynamics on master until I'm at -4 and the loudness matches some guide track.

Then I fix all the distorted tracks with eq and what not till overall level and RMS is right.

This is horrifying. Music deserves to breathe.

I know, it sucks but everything is loud now. The alternative is your tracks sound weaker than everyone else's.

Services like Youtube and Spotify automatically turn down the volume of loud tracks (around -14 LUFS or more). So if you squash your tracks they'll actually sound weaker on those services since they will be the same loudness as other tracks, but with less dynamic range. You should also leave at least 2dB of headroom when compressing to lossy formats like mp3.

[Kinh]

Get the mastering track on first, boost level to max and RMS as close to -4 as I can. If I cant then I do parallel compression for most of the tracks, squash the dynamics on master until I'm at -4 and the loudness matches some guide track.

Then I fix all the distorted tracks with eq and what not till overall level and RMS is right.

This is horrifying. Music deserves to breathe.

I know, it sucks but everything is loud now. The alternative is your tracks sound weaker than everyone else's.

Services like Youtube and Spotify automatically turn down the volume of loud tracks (around -14 LUFS or more). So if you squash your tracks they'll actually sound weaker on those services since they will be the same loudness as other tracks, but with less dynamic range. You should also leave at least 2dB of headroom when compressing to lossy formats like mp3.

Hmmm.. interesting but if you play any mp3 from any new artist back in your daw and analyze it you can clearly see there's no headroom and it's been compressed like hell.

[Burillo]

I'm no authority, but here's my take on it.

1) Throw a lowcut EQ around 50 Hz (sometimes lower or higher) on every track.

yes, that's generally first thing i do after i have the basic sounds in place. hi/lo cuts to cut away anything that isn't necessary.

2) Get a rough levels balance across the tracks. Usually I start with the drums/percussion (do a submix of these).

start with what you want to be a central element, and build everything else around it. starting with drums is not always a good idea - you may very well want to start with vocals, or with your synth lead, or something else.

3) Compress tracks which are spiking enough to get master bus overloads

while that's part of what compression is for, that's not the entire story.

for example, if you have a bass guitar that is quite dynamic and you feel that you're some times suddenly losing low end because the bassist picked a dead fret - that's when you need a compressor. it's not just to prevent overshoots, it's also to even out the dynamic range, so that the track doesn't stick out or drown. the macro stuff can be fixed with volume fader automation (especially on highly dynamic material such as vocals), but the microdynamics are also what compressors are for.

compressors can also be used to shape the sound, not just make it sit better. a classic case is the snare compression - you don't just make the snare sit better, you're actually changing how it sounds, sometimes by a lot (it's not unusual for me to pile up 15dB of compression on a snare). drum bus is similar scenario - you can not only prevent spikes on the master, but actually shape the entire drum section to "breathe" and groove better (or slam it into oblivion if that's your kinda thing).

Unfortunately, I'm not sure how to tell when instruments are clashing in frequency areas and therefore when to use EQ to sort that. Any tips here?

that's easy just listen closely.

if you find yourself trying to figure out which track is playing at a particular frequency at any given moment - that's when you need to make room. a classic case of this is having two independent but equal vocal tracks (i.e. not just doubling - think duet) - inevitably, you'll have to space them out in a way that leaves them both audible.

another common case is the 4k+ range, where an acoustic guitar's pick attack might be mixed with your vocalist's diction (which is also around there). there's no way around it - one or the other will have to make room, because there's clearly a conflict there. over time, you'll simply start hearing more of this stuff - like, oh, this guitar's distortion is hiding my synth pad sometimes, i better do something to ensure that both are heard at all times.

note that not all clashes need fixing. transients are also something you need to take into account - if there's a clash in transient-heavy material vs. a sustaining material (e.g. a hihat vs. a noisy synth pad), then generally speaking such conflict does not need to be fixed unless you screwed something up - transients catch our attention and mask out underlying sound, and that's the point.

i should also note that not all such things are fixed (or fixable) with EQ. you have a wide range of tools at your arsenal - EQ, volume, panning, sidechain compression/dynamic EQ, and lots of other stuff. you may use something like MedlaProduction's MMultiAnalyzer to see which tracks are clashing - it has such a feature. also look into WavesFactory TrackSpacer or MedlaProduction's multitude of auto/dynamic/spectral effects, and of course sidechain compression/deessing. for a newbie, TrackSpacer is just the ticket.