A synth patch arrives at the mix bus carrying every frequency you gave it, and usually several you did not want. Designing a patch with the arrangement in mind produces a mix that is easier to balance, because the source is not competing with elements that have not entered yet.

The crowded band

The problem with unconsidered synth design is almost always the same. A default patch uses a full spectrum from the sub up through several kilohertz, and it sounds good in isolation. In an arrangement, it occupies the same region as the bass, the guitar and the vocal, and the mix has no space to work with.

Nothing about the patch is wrong. It is simply a complete sound competing against other complete sounds, and the mix becomes a negotiation between three things that all want the same frequencies.

Design for subtraction

The classic approach is to build a sound with fewer harmonics and then remove what you do not need. A saw wave at a low cutoff is warm and focused. Adding resonance at a chosen frequency gives character in a specific place. Opening the filter a little at a time and listening at each stage keeps the design intentional.

The alternative is additive design, building a sound from a small number of deliberately chosen partials. This gives very precise control over where harmonic content sits, and it produces patches that occupy a narrow band rather than a wide one, which is considerably easier to mix.

Neither approach is correct. What matters is knowing which parts of the spectrum your patch is using, and whether that is where you want it.

Designing with the arrangement in mind

1

Decide the patch's job before you design it. A pad that fills a frequency range, a bass that owns the low end, a lead that needs to be heard over everything. A patch without a job tends to occupy too much of everything.

2

Start with the filter more closed than you think. Open it until the character you want is present, then stop. The instinct to open it further comes from hearing the patch in isolation, where extra brightness sounds like detail rather than like competition.

3

Give the patch a gap somewhere. Every patch should leave at least one part of the spectrum to other elements. A pad with a scooped lower-mid leaves room for a bass, and a lead with a rolled-off bottom leaves room for the kick.

4

Unmute it in context, repeatedly, during design. A patch designed in isolation will need EQ work later, because it was never tested against the thing it has to coexist with.

Harmonic content and mixing consequence

ContentSounds likeMixing consequence
Full saw, open filterBright, aggressiveOccupies everything, hard to place
Full saw, closed filterWarm, mellowSits well, limited presence
Pulse, narrow widthFocused, hollowLeaves room at top and bottom
Square, full widthReedy, presentCompetes with vocals for midrange
Few partials, additiveClear, syntheticNarrow band, easy to mix
Resonant filter peakCharacterfulSingle focused emphasis, mixable

Frequency space for the mix

A useful check is to compare the patch's spectrum against everything else playing. If two sources both have significant content in the same region, something will have to give, and it is usually decided by level rather than by a considered decision about which one should occupy that region.

Making that decision at the patch stage is far cheaper than making it with EQ after the fact. A pad that was designed to leave a gap can simply stay where it is. A pad that occupies everything needs to be carved back, and carving back a full-spectrum sound is harder than designing one with a gap.

Retuned mixes inherit your patch choices

A retune moves the fundamental of every source down, which moves the entire harmonic structure with it. Patches designed to occupy a narrow band will occupy a narrow band at the new tuning too, and will need no more adjustment than they did before.

Full-spectrum patches become slightly more crowded after a retune, because their content has moved down into a region that is also carrying everything else's fundamentals. This is where 432Hz MASTER is most likely to need help from a high-pass filter, and it is another argument for designing patches with a gap rather than avoiding retuning.

With RASTA 528Hz, where intervals are redistributed, the harmonic relationships between patch and arrangement change rather than sliding, so patches that sat well before may not sit as well after even if their bandwidth is narrow.

The bottom line

Design the patch for its job, not for its demo. A sound that leaves room for the rest of the arrangement needs almost nothing done to it later, and that advantage compounds when the mix is retuned.

Frequently asked questions

Why does my synth disappear in the mix?

Usually because it occupies the same frequencies as other elements and is losing the level competition. The fix is usually to find it a gap in the spectrum, either by designing the patch with one or by cutting the competing sources rather than the synth.

Should I design patches in isolation?

Only for the initial character. Any decision about where a patch sits in the spectrum should be made with the rest of the arrangement playing, because the same patch can be well-judged in one context and completely wrong in another.

What does subtractive design mean in practice?

Starting with a harmonically rich waveform and removing the parts you do not want with filters. A saw at a low cutoff is warm and focused, and opening the filter adds brightness along with competition for space. Additive design is the opposite, building from chosen partials for precise control.

Do narrow-band synths mix better after a retune?

Yes, generally. A retune moves a source's whole harmonic structure down, so a narrow-band patch remains narrow and just moves. A full-spectrum patch becomes denser in a region that is also carrying other fundamentals, which is where it will need more attention after retuning.

More from the atixUniverse blog