How much does filtration influence mouthfeel?

The Question

“Charcoal filtered” is one of the oldest claims on a vodka label, usually presented as an unqualified good: more filtering, more smooth. We were skeptical in both directions. Does filtration meaningfully change mouthfeel? And is the change always an improvement — or does an aggressive filter take texture we’d rather keep?

Setup

One batch of our spirit, split five ways: an unfiltered control, a simple particulate filter, and three levels of activated carbon contact, from a light pass to a deliberately excessive multi-pass treatment. Everything else held constant — same proofing water, same rest time, same glassware. Then blind triangle tastings with everyone in the building we could recruit: three glasses, two identical, find the odd one out. It’s a humbling format and that’s why we use it.

Observations

The honest headline: several of the differences were harder to detect than we predicted. The control versus the light carbon pass produced plenty of confident guesses and an unimpressive success rate — with a spirit that comes off the column as clean as ours, there simply isn’t much for a light polish to remove. The heavy multi-pass sample was a different story. Tasters found it reliably, and the words they reached for weren’t compliments: “thin,” “flat,” “like it ends early.” A few tasters described the lightly filtered samples as marginally softer than the control, and we believe them, but the effect sat near the edge of what our panel could resolve. Aroma differences tracked the same pattern — negligible at light contact, clearly stripped at heavy contact.

What We Learned

Filtration is subtraction, and subtraction doesn’t ask permission — carbon takes texture and aroma compounds along with any roughness. For a spirit with real flaws, that trade can be worth it, which is why the technique is a century old. For a spirit that’s already clean, heavy filtration mostly removes character it took us effort to keep. It confirmed a design decision: our process puts the cleanliness in the separation, so filtration can stay what it should be — a light finishing step, not a rescue operation. “Filtered X times” is a number, not a virtue.

Next Time

Two threads worth pulling. First, temperature: cold filtration is standard practice in parts of the industry and may change what the carbon selects — worth a controlled comparison. Second, our panel kept using texture words we can’t measure yet (“round,” “ends early”). We’d like to pair future tastings with something more instrumented, even crudely. When your data is adjectives, better adjectives are progress.