Use less, but use the right thing in the right order

Almost no amateur mix suffers from too little processing. They suffer from processing added to fix the previous processing, in an order nobody decided, at amounts nobody verified.

Three plugins doing 2 dB each is not subtlety. It is three chances to be slightly wrong, and three interactions to untangle when something sounds off.

How chains get bloated

It happens the same way every time.

You add an EQ to fix a resonance. The cut changes the balance, so the source sounds quiet, so you add a compressor with makeup gain. The compressor now reacts differently and pushes the top end forward, so you add a de-esser. The de-esser dulls it, so you add an exciter.

Four plugins. The original problem was one resonance. Every plugin after the first exists to compensate for the one before it.

The way out is not “use fewer plugins” as a rule. It is checking after each step whether you have solved the problem or moved it.

The one-sentence test

For every plugin in a chain, you should be able to say in one sentence what it does that nothing else does.

• “This removes the room resonance at 340 Hz.” — keep it.
• “This stops the level jumping between verse and chorus.” — keep it.
• “This is a bus compressor for glue.” — keep it.
• “I always put one here.” — delete it.
• “It sounded a bit better.” — bypass at matched level and check again.

That test alone removes most of the excess in a typical session.

Why louder always wins, and why that matters

Almost every processor makes things louder: EQ boosts, makeup gain, saturation, limiting. And a signal 1 dB louder sounds better in an A/B, regardless of what else changed.

So the standard workflow — add plugin, bypass, “yes that is better”, keep — is systematically biased toward keeping plugins that did nothing useful.

The fix is unglamorous: match levels before every comparison. Use the plugin’s output gain, or a gain utility. Then decide.

You will delete a surprising number of plugins in the first session you do this consistently.

Order beats quantity

Two plugins in the right order routinely outperform five in the wrong one, because most mixing processors react to what reaches them.

A cut before the compressor stops the compressor chasing energy you were going to remove. Saturation after compression stays consistent instead of changing with every loud passage. Reverb last, or on a send, so nothing processes the tail.

Getting that right removes the need for the compensating plugins. Effects chain order covers the reasoning in full.

Consolidate where one plugin does several jobs

Sometimes the honest answer is that a chain of three could be one.

A corrective EQ, a frequency-triggered compressor and a de-esser are three plugins performing level-dependent gain changes in three frequency ranges. A dynamic EQ does all three in one window, with one gain structure and one place to look when something is wrong — see dynamic EQ instead of EQ plus a compressor.

That is not a shortcut. It is removing two hand-offs where things could drift.

Fix the source instead

The most effective processing decision is often not to process.

• A muddy sound that needs three EQ cuts might be the wrong sample.
• A part needing heavy compression might need a better performance.
• Two instruments requiring constant ducking might need different octaves.
• A vocal fighting for space might need the guitar to stop playing there.

Changing the source or the arrangement solves the problem permanently and survives every playback system. Processing manages it. See ducking to create space for where the line sits.

How to strip a bloated chain back

When a channel has stopped making sense:

1. Bypass the entire chain. Listen to the raw source in the mix.
2. Name the actual problem in a sentence. One problem, the biggest one.
3. Enable only the plugin that addresses it. Set it from scratch.
4. Listen again. Is there still a problem worth fixing?
5. If yes, add one plugin. If no, stop.
6. Delete everything still bypassed. Not “keep it disabled just in case” — delete it. It will be in the project history if you need it.

Most channels end this process with one or two plugins and sound better than they did with six.

The signals that you have gone too far

• You cannot explain what a plugin is doing.
• Bypassing the chain at matched level sounds cleaner.
• The mix sounds worse than the rough mix from week one.
• CPU is a constant problem on a modest arrangement.
• You are adjusting settings rather than listening to the song.

That fourth one is worth taking seriously. A CPU problem caused by chain length is a structural message, not a hardware one. If the chains are genuinely needed and finished, printing effects to audio solves the load without deleting anything.

The point

Restraint is not an aesthetic preference. Every processor you add is another thing reacting to everything before it, another gain change to track, and another place for a small error to compound.

Fewer, better-placed decisions are easier to make, easier to hear, and much easier to fix at two in the morning when something is wrong and you cannot find it.