My top changes for the better (for me) over the last couple years, in no particular order:
Building a dedicated DAW for music only.
Switching from Mac to PC so that I could upgrade the parts easier and more cost effectively.
I take a template-based approach, possible because my music is for myself, only. Even Group Channels can be identical from song to song. So I decided to move my stereo buss processing, pre-mastering and mastering to a secondary computer. It’s a real-time chain via optical cables.
This was a huge step forward for me. Freed up CPU on my DAW, freed up space on my screens, Cubase projects are therefore smaller and load faster, provides good decoupled-ness between my project and the mastering chain.
Also, it means my Cubase mixdowns are free from stereo buss processing (buss compression, etc.) being printed into them, which gives me more flexibility when mastering or remastering later.
Going to K-14 for tracking / VSTi sound design. This has been the best thing, sonically, I’ve done. Making sure I don’t hit input-modeled plugins too hot has really open up my soundstage and avoided mud in the low-mids. It also allows me to sum without any limiting or compression so that my mixdowns have the most flexibility. This is because I actually have my final stems, at K-14, pulled back to -18 rms so that they can sum without any clipping.
Creating K-14, K-12 and iTunes spec’d LUFS -16 endpoints in my mastering chain, as well as the obligatory “Beatport Loud” DR6/DR4 level at the end, has allowed me to make better mixing decisions because I can make sure the mix sounds good loud as well as more dynamic. There are times when it would sound good loud, but fall apart when at K-14 or K-12. By being able to quickly monitor between all of these dynamic ranges, it’s made my mixes better. Bonus is that if I needed to have someone else master, I have K-14 mixes I’m confident about, ready to go.
Stereo Buss Compression goes to a pre-mastering stage, which goes to K-14, then is gain reduced to K-12, then gets lowered signal-level-wise (but kept at K-12 dynamic-range-wise) to the iTunes spec (LUFS 16), then gets leveled back up and slammed to the typical DR6/DR4 territory.
If I have to make things loud, it’s not going without a fight for wider dynamics. The first step of that is actually producing to those wider dynamics. The second step is releasing both mixes (one at K-14 or K-12, one at DR6/DR4) and letting the consumers choose which they want.
You can bring the horse to the water, but you can’t make it drink. The louder DR6/DR4 will still be the most downloaded, listened to, purchased, probably, but I’m at least moving in a direction to provide the option and to help educate the masses on why the loudness war is not a better listening experience.
Anyway, those are my top five.