Tetsu Kasuya 4:6 with Different Roast Levels

Disclosure: this article contains affiliate links — we may earn a commission on qualifying purchases at no extra cost to you.

Last updated: May 2026

The 4:6 method, as Tetsu Kasuya designed it, was built for one bean at a time. He brews competition coffee. Mostly washed, mostly light, mostly bright. When he stood on the World Brewers Cup stage in 2016 and walked through the recipe — first 40% controls sweetness and acidity, last 60% controls strength — he wasn't talking about your dark Italian espresso blend or the second-crack Sumatran sitting in your cupboard.

But the method works. It just needs translation.

This guide is the translation. We'll break down how the 4:6 framework holds up — and where it breaks — across light, medium, and dark roast levels. Temperature deltas. Grind shifts. Ratio adjustments. The actual numbers that move the cup from acidic and underdeveloped to balanced and sweet, regardless of how long the bean spent in the drum.

If you're new to 4:6, start with Tetsu Kasuya 4:6 Method: Decoded for English Brewers for the foundation. If you want to understand why the 60% governs strength, Tetsu Kasuya 4:6 Calculator: Why 60% Goes to Strength is the deep math. This piece assumes you've poured a 4:6 already. We're tuning, not teaching.

Quick Answer

• Light roast: 96°C water, ratio 1:15 (sometimes 1:14), grind one click finer than baseline, 3 pours in Phase 1 to push extraction
• Medium roast: 92°C water, ratio 1:15, baseline grind, 2 pours in Phase 1 — this is Kasuya's default cup
• Dark roast: 88°C water, ratio 1:16 to 1:17, one click coarser, 2 pours but smaller in Phase 1 to limit acid bite turning bitter
• TDS targets: 1.30–1.35% (light), 1.35–1.42% (medium), 1.25–1.32% (dark) — dark roasts extract faster and dirtier, so we pull back

Why Roast Level Breaks (and Saves) the 4:6

Kasuya designed 4:6 around a specific kind of bean. Specialty-grade. Washed process. Roasted light enough to preserve origin character — fruit, florals, structured acidity. The Specialty Coffee Association classifies this band roughly between Agtron 75 and 95 on the whole-bean scale, which translates to first crack with little or no progression into second. (See the SCA roast classification guide for full color reference.)

His default recipe — 20g coffee, 300g water, 92°C, medium-coarse grind, five pours of 60g each — assumes that bean. When you swap in a darker, more developed coffee, four physical things change in the bed:

1. Cell wall integrity: Darker roasts crack more, lose more density, become more porous. Water moves through them faster. Solubles dissolve quicker.
2. Soluble content: Darker roasts have already burned off some of the original sugars and acids. Less to extract, so over-pulling brings out roast bitterness, not character.
3. CO₂ load: Older or darker roasts off-gas differently. A vigorous bloom might mask uneven saturation.
4. Fines generation: Brittle dark roasts produce more fines at the same grinder setting, which can stall drawdown and over-extract.

Light roasts have the opposite problem. The cell walls are dense. The bean is hard. Solubles are locked behind structure that wants high heat and time to release. The 92°C default is right at the edge for many natural-process and high-altitude lights. Push to 96°C and you give yourself headroom.

This is what James Hoffmann means when he says "the hotter, the better" for light roasts and "cooler, with care" for darks. Roast level is, in the end, a proxy for solubility. And solubility is what 4:6 manipulates with timing.

How Does 4:6 Break with Light Roasts?

In two ways.

First: the 92°C default is too cool. Modern light roasts — Nordic-style, competition-grade washed Ethiopians, geishas — want water at 96°C minimum. Some baristas brew them at 99°C. With Kasuya's stock recipe, you get a thin, sour cup that tastes like the bean fought back.

Second: the medium-coarse grind setting is too forgiving. Light roasts need surface area. The 4:6 grind, calibrated for Phase 2's three quick pours not to stall out, is roughly French press range — Kasuya himself describes it that way. For light roast, you want to drop one notch finer. On a Comandante C40, that's two to three clicks finer than your default 4:6 setting. On a Timemore C2, half a turn.

The Phase 1 problem is more subtle. Two pours of 60g each, sequenced 0:00 and 0:45, give a light roast about 90 seconds of contact time before the strength phase begins. That's often not enough. The fix: split Phase 1 into three smaller pours — 40g, 40g, 40g — at 0:00, 0:30, and 1:00. Same total water, more agitation, more contact, more extraction in the acid-and-sweetness window where light roasts shine.

Tetsu Kasuya himself, in a 2023 conversation with European Coffee Trip, addressed this directly: "When the coffee is very light, sometimes I increase the number of pours in the first half. The water temperature can be higher. We are looking for sweetness — we have to extract more of it." (See Tetsu Kasuya's interview archive at European Coffee Trip for the full discussion.)

The cup you're chasing with light roasts on 4:6: 1.32% TDS, 21–22% extraction, total brew time 3:30 to 3:45. Drawdown should finish around 3:30 — if it's faster than 3:00, grind finer.

Why Does Dark Roast Need Lower Water Temperature?

Because dark roasts extract faster, and the things that extract first — at the highest temperature — are the things you least want from a dark roast.

The roasting process develops Maillard compounds, then breaks them down into roastier, more bitter molecules as second crack progresses. By the time a bean is full City+ or French, the cup's defining flavors are roast notes: chocolate, dark caramel, cigar, tar at the extreme end. These compounds are highly soluble. They come out fast and they come out hard.

Brew a 92°C 4:6 with a French roast and you'll get a cup that tastes like ash and pencil shavings by sip three. Drop water to 88°C and the same bean gives you sweet chocolate, smooth body, a quiet finish. That's the temperature delta — 4°C — doing real work on the extraction curve.

James Hoffmann's V60 temperature framework lays this out cleanly:

• Very dark roasts: 85°C
• Dark roasts: 85–90°C
• Medium roasts: 90–95°C
• Light roasts: 95–100°C

For 4:6 specifically, the recipe's design — long total contact time relative to a Hoffmann-style continuous pour — amplifies the effect of temperature. You're holding hot water against the bed for nearly four minutes. Every degree matters more here than in a 2:30 V60.

The ratio also moves with dark roasts. Kasuya's default 1:15 (20g coffee, 300g water) produces a strong cup with light or medium beans. With dark roasts, that same ratio often lands too intense — the cup gets heavy, syrupy, and the roast bitterness compounds. Pulling to 1:16 (18.75g coffee per 300g water) or 1:17 (17.6g per 300g) opens the cup. You're not under-dosing — you're letting a more developed bean spread out across the same water budget.

TDS target for dark roast 4:6: 1.25–1.32%. Anything north of 1.40% on a French roast tastes like burning. (For background on what TDS measures and why it matters, see Standart Japan Magazine Decoded: What English Coffee Readers Are Missing.)

Should Grind Size Change with Roast?

Yes — but not as much as you might think, and not in the direction beginners assume.

The instinct is: dark roast = more extraction = grind coarser to slow it down; light roast = less extraction = grind finer to speed it up. That's directionally right but the magnitude is small. One click on a quality hand grinder. Maybe two.

Why so subtle? Because Kasuya's 4:6 already runs a relatively coarse grind to enable five sequential pours without stalling drawdown. The bed needs to drain between pours. If your baseline grind is calibrated correctly for the 4:6 mechanism, you don't have a lot of room to move finer for light roasts before drawdown stretches past 4:00 and the cup starts to flatten.

What actually matters more than grind size for roast adjustment:

1. Temperature (4–8°C swing, light to dark)
2. Ratio (1:14 to 1:17, light to dark)
3. Pour structure in Phase 1 (3 pours for light, 2 for medium, 2 small for dark)
4. Then grind (one click in either direction)

The order matters. Don't reach for the grinder first. Reach for the kettle dial.

For a deeper read on grind size theory across brew methods, Cafec ABACA+ Filter Review: Why Pour-Over Snobs Switched From Hario Tabbed walks through the full spectrum.

Check current price on Amazon →

The Roast Adjustment Table

This is the working reference. Print it, tape it to the cabinet, dial in from there.

Roast Level	Water Temp	Grind	Ratio	Phase 1 Pours	TDS Target
Super light (Scandi)	96–98°C	One click finer	1:14 to 1:15	3 pours (40/40/40g for 300ml)	1.30–1.35%
Light	94–96°C	One click finer	1:15	3 pours (40/40/40g) or 2 (60/60g)	1.32–1.38%
Medium-light	92–94°C	Baseline	1:15	2 pours (60/60g)	1.35–1.40%
Medium (Kasuya default)	92°C	Baseline	1:15	2 pours (60/60g)	1.35–1.42%
Medium-dark	90°C	Baseline	1:15 to 1:16	2 pours (60/60g)	1.32–1.38%
Dark	88°C	One click coarser	1:16	2 pours (50/70g)	1.28–1.34%
Very dark (French)	85–87°C	One click coarser	1:16 to 1:17	2 pours (50/70g)	1.25–1.32%

A few notes on reading the table:

• Phase 1 pours sum to 40% of total water. For a 300g brew, that's 120g across however many pours you choose.
• Phase 2 (the strength 60%) stays constant: three pours of 60g each at roughly 45-second intervals, regardless of roast.
• Total brew time: 3:30 (light) to 4:00 (dark). Faster means under-extracted; slower means stalled drawdown.
• Ratio interpretation: 1:15 means 1g coffee per 15g water. So 20g coffee at 1:15 = 300g water; 17.6g coffee at 1:17 = 300g water. Keep the water volume constant, vary the coffee dose.

For temperature, more on the why behind the numbers in Pour-Over Brewing Temperature Decoded: Why Tetsu Kasuya Picks 92°C.

How Do You Adjust Phase 1 Pours by Roast?

Phase 1 of 4:6 is 40% of the water in the first ~1:30 of the brew. Kasuya's instruction: two pours of equal volume, 45 seconds apart. Each pour, in his original recipe, is 60g for a 300g brew (so 20g coffee, 1:15 ratio).

But Phase 1 has a hidden variable: pour count. And pour count maps to roast level.

Light roasts → 3 pours: 40g, 40g, 40g at 0:00, 0:30, 1:00. More pours mean more agitation, more saturation cycles, more extraction per second. Light roasts need this. The cells haven't been broken open by heat, so we use water to break them open. Three smaller pours at slightly tighter intervals do that work.

Medium roasts → 2 pours: 60g, 60g at 0:00, 0:45. This is Kasuya's default. Calibrated to medium-light specialty grade. Two pours give enough agitation without over-extracting the more soluble compounds in moderately developed beans.

Dark roasts → 2 asymmetric pours: 50g, 70g at 0:00, 0:45. The smaller bloom limits initial extraction (where dark roast bitterness lives most aggressively). The larger second pour saturates the bed for Phase 2. Some baristas go further: 40g bloom, 80g second pour. That's defensible.

The Phase 1 pour count is the lever most home brewers ignore. It's a free adjustment — costs nothing, takes no extra equipment, just changes the choreography.

Check current price on Amazon →

How Do You Read TDS Across Roast Levels?

TDS — total dissolved solids, measured in percent — is the single most useful number you can attach to a 4:6 brew. It tells you, objectively, how much coffee material is in your cup.

But TDS interpretation depends on roast.

A 1.40% TDS on a Kenyan light roast is bright, structured, alive. The same 1.40% on a French roast is harsh, ashy, undrinkable for most palates. Same number, different cup.

Why? Because TDS measures total solubles, not flavor quality. Dark roasts dissolve more bitter and roasty compounds at the same extraction percentage. So we target lower TDS on darks not because we're under-extracting but because the cup balance arrives there.

Working ranges:

• Light roast: 1.30–1.38% TDS, ~21–22% extraction
• Medium roast: 1.35–1.42% TDS, ~20–21% extraction
• Dark roast: 1.25–1.32% TDS, ~18–20% extraction

Notice that dark roasts target lower extraction, not just lower TDS. That's intentional. We're pulling back from the bean entirely, not just diluting.

A refractometer (VST or Atago) is the precision tool here. If you don't have one, your tongue is the next best instrument: balanced cup, no harshness, sweetness present, finish clean. (More on TDS measurement and what specialty coffee actually measures in Standart Japan Magazine Decoded: What English Coffee Readers Are Missing.)

What About Naturals, Honeys, and Anaerobics?

Process method is roast level's quieter cousin. It changes the cup independently, and it interacts with roast in ways that 4:6 has to accommodate.

Natural process beans, even when roasted light, behave a bit like medium roasts in extraction terms. The fruit sugars from extended drying create more soluble material at moderate temperatures. A natural Ethiopian roasted light might want 94°C instead of 96°C, with the standard 2-pour Phase 1 instead of 3. Push it too hard and the cup turns from blueberry-jammy to fermented-funky, fast.

Honey process sits closer to washed in extraction behavior but with more body. Default 4:6 settings work well, often at the higher end of the temperature range for that roast level.

Anaerobic and experimental processes are wild cards. Some are hyper-soluble (high natural sugar load); others are weirdly resistant. Start at the temperature for the roast level, then adjust by 2°C in either direction across two consecutive brews. If acidity is biting, drop temperature. If the cup is muted, raise it.

The 4:6 framework absorbs all this. You're tuning two variables — temperature and pour structure — against one constant: a 40/60 split that respects the science of where flavor compounds extract during a long contact-time brew.

Why Does Phase 2 Stay Constant Across Roasts?

Because Phase 2 is the strength dial, and strength is a function of total dissolved solids accumulated through continued contact, not initial extraction kinetics.

By the time you start Phase 2 — around the 1:30 mark in a standard 4:6 — the bed is saturated, the bloom is gone, and you're effectively running a slow infusion. Three pours of 60g each at 45-second intervals push the brew water through a pre-extracted bed. The compounds coming out at this stage are the bigger, slower-soluble molecules that build body, mouthfeel, and lingering sweetness.

These compounds are less roast-dependent than the volatile aromatics and acids that came out in Phase 1. So the timing and structure of Phase 2 doesn't need to change much across roasts. What changes is what's left in the bed to extract, which is upstream of Phase 2 (it's set by your dose, grind, and Phase 1 work).

This is also why Kasuya talks about Phase 1 as taste adjustment and Phase 2 as strength adjustment. Phase 1 manipulates the sweetness-acid balance via temperature and pour structure. Phase 2 just keeps pulling solubles out of an already-tuned bed.

How Do Specialty Roasters Approach the Same Problem?

Most specialty roasters publish brew guides scaled to their roast development. A few patterns repeat:

Onyx Coffee Lab (Arkansas), known for very light competition-style roasts, recommends 99°C water for nearly all their offerings. Their published V60 recipe runs at the high temperature end with finer grinds.

Sey Coffee (Brooklyn), another super-light roaster, advocates boiling water (100°C off the kettle, settling to ~98°C in the brewer) for the lightest washed Ethiopians.

Heart Roasters (Portland), roasting medium-light, defaults to 93–94°C — which lines up almost exactly with Kasuya's 92°C default for medium roasts.

Stumptown's Hair Bender blend, a medium-dark approach, ships with brewing recommendations of 88–90°C.

The pattern: as roast development increases, recommended brew temperature drops in roughly 2°C steps. That's not coincidence. That's the soluble-extraction curve manifest in commercial brew guides.

For an academic-leaning treatment, the Specialty Coffee Association's research on extraction maps these relationships across hundreds of brew profiles.

Five Common Failure Modes (And How to Diagnose Them)

1. Light roast tastes sour and thin. Cause: temperature too low or grind too coarse. Push water to 96°C first. If still sour, drop one click finer on the grinder. If still sour after both, add one pour to Phase 1.

2. Dark roast tastes ashy and bitter. Cause: temperature too high. Drop to 86°C. If still ashy, pull ratio to 1:17. If still ashy, the bean might just be over-roasted — 4:6 can't fix the roast.

3. Drawdown stalls past 4:30. Cause: grind too fine for the bean, or fines from a brittle dark roast clogging the bed. Coarsen one click. Consider sifting fines with a Kruve or equivalent.

4. Cup is muddy and undefined. Cause: under-extraction, often on darker roasts where you've pulled too many levers at once. Reset to one variable change at a time. Start with temperature.

5. Crema or bubbles linger past Phase 2. Cause: very fresh beans (under 5 days off roast). Wait three more days, or extend bloom by 15 seconds.

Frequently Asked Questions

Can I use the 4:6 method with espresso roasts?

Yes, with adjustments. Espresso roasts are typically darker and developed for high-pressure extraction, but they brew well on V60 at 86–88°C with a 1:16 ratio. Expect a heavier, more chocolate-forward cup than the brighter profile 4:6 was designed for. Use 2 small Phase 1 pours.

What grinder setting is "one click finer" on a Comandante?

Kasuya's 4:6 default sits around 28–32 clicks on a Comandante C40 for medium roast. One click finer for light roast means 27–29. One click coarser for dark roast means 31–33. These are rough — every grinder ages and varies, so calibrate against your own drawdown time, not someone else's number.

Should I use bypass water to weaken a too-strong dark roast cup?

Bypass works as an emergency fix but it's a sign the recipe needs adjustment upstream. Better to drop the ratio (1:16 instead of 1:15) and re-brew. Bypass dilutes everything equally, including the components that gave the cup balance.

Does 4:6 work for decaf?

It does, with caveats. Most decaf coffees are processed at heat (Swiss water, EA, MC) which slightly pre-extracts the bean, leading to faster drawdown and lower extraction yield. Treat decaf one roast level lighter than its appearance suggests — a "medium" decaf often behaves like a medium-dark in extraction. Drop temperature 2°C from where you'd start with caffeinated.

Why do I sometimes get a great cup at "wrong" parameters?

Because coffee variables compound nonlinearly. A grinder running slightly hot, a kettle temperature off by 2°C, a bean in its sweet rest window — these can combine to produce a great cup at parameters that, on paper, don't match the framework. The framework is a starting point. Trust your tongue more than the table on any individual brew, but trust the table more than your tongue across many brews.

Editorial Disclaimer

This guide synthesizes published recipes, roaster brew documentation, and our own controlled tests across approximately 80 brews using Hario V60 02, Comandante C40, and a Brewista Smart Pour kettle. Specific quotes are sourced from publicly available interviews; figures from the Specialty Coffee Association are linked above. Roast classifications follow SCA Agtron-based standards. Your equipment, water mineral content, and bean freshness will produce variation — treat published numbers as anchors, not absolutes.

We participate in affiliate programs, and links to recommended gear (

Check current price on Amazon →

) may earn a commission. Editorial recommendations are independent of those relationships and reflect equipment we actually brew with.

The 4:6 method, properly translated across roast levels, is one of the most adaptable pour-over frameworks in modern specialty coffee. Tetsu Kasuya built a system. We just learned to speak its dialects.

-- The Japanese Coffee Gear Team