Recent cups¶

A running log of what actually got brewed on the xBloom — the dial-in, cup by cup, newest first. Every cup: a rinsed paper filter and filtered water.

2026-08-26 — THE BARN — Bishan Wate · 58 · 15 g → 240 ml (1:16) · 94 °C · 45 s bloom (60 ml) + 100 ml + 80 ml · 120 RPM · Omni¶

First pour-over: generally good, with a slightly bitter end note.

First filter brew on this bag — the recipe was written but untried, and the body and sweetness clearly landed, so this is a tail problem rather than a broken dial. Bitterness confined to the finish on an otherwise good cup is the signature of the drawdown pulling too far, not of under-extraction, which would have made the whole cup thin and sour instead. Two of the starting numbers are candidates and both were calibrated guesses rather than measurements: grind 58 was chosen to split the two dialed Ethiopians in the log (57 and 60), and 94 °C was a deliberate step above the house 92-93 band on the theory that a dense light roast needs the heat. Temperature is the weaker extraction lever but the stronger BITTERNESS lever, and 94 sits above the house sweet spot, so it moves first and the grind stays put — the cup does not need less extraction overall, just less of the bitter tail. Bag is 34 days off roast, which is worth noting but not worth acting on: dial to the cup, not the calendar.

➡️ Next: Drop 94 → 93 °C, grind unchanged at 58. If the finish is still bitter, then coarsen to 60 (the bean's own second lever).

2026-08-25 — THE BARN — Guava Banana · 60 · 15 g → 240 ml (1:16) · 93 °C · 80 s bloom (60 ml) + one pour (180 ml) · 120 RPM · Omni¶

Better at 60 — the fruit turns up. Dialed.

Second brew, one lever moved: grind 62 → 60. The sourness and the hollow middle of the first cup are gone, which confirms the diagnosis was under-extraction rather than the bean being quiet. Worth recording why 62 was wrong in the first place — it was never measured for this lot, it was inherited from the closest dialed anaerobic natural in the log, and it sat at the coarse end of a range whose median is 60. The bag was also 33 days off roast, and aged coffee extracts faster, so a setting still stopping short that late was too coarse from the start. Temperature stayed at 93 throughout, which is the roaster's own figure for this lot; going cooler would have lowered extraction and pushed a sour cup further the wrong way.

➡️ Next: Nothing. Leave it here.

2026-08-25 — THE BARN — Guava Banana · 62 · 15 g → 240 ml (1:16) · 93 °C · 80 s bloom (60 ml) + one pour (180 ml) · 120 RPM · Omni¶

Sour and thin — the guava never arrives. Under-extracted.

First brew on the starting recipe, and the first real feedback it has had. Hollow middle, sharp up front, no body — the classic stopped-short profile rather than anything muddy or drying, so this is not the over-extraction that anaerobic naturals usually fall into. Two things point the same way. The grind of 62 was never measured for this lot — it was borrowed wholesale from another anaerobic natural — and it sits at the coarse end of this log, where the median dialed filter grind is 60. And the bag is 33 days off roast: aged coffee is more brittle and normally extracts faster, so a setting stopping short at day 33 was almost certainly too coarse from the start. Temperature stays at 93, which is the roaster's own guidance for this lot; cooler water would reduce extraction and push a sour cup further the wrong way.

➡️ Next: Grind finer: 62 → 60. One change; leave the ratio, the temperature and the two-pour shape alone.

2026-08-06 — El Cafetero — Kostaryka Hacienda Sonora · (as currently dialed)¶

OK — clean and drinkable, but not especially expressive. Brewed at ~22 days off roast (roasted 2026-07-15), which is the right window for a natural, so this reads as the coffee rather than under-rest. Left dialed as-is.

2026-07-27 — Area 51 — Ethiopia Yirgacheffe · Worka Nenke · grind 60 · 93 °C · 15 g at 1:16 (240 ml) · 45 s bloom then three spiral pours · 2:45–3:15¶

Landed perfectly — delicate and balanced, the recipe works as dialed.

A natural Yirgacheffe at 55 days off roast (roasted 2 June) still reading clean and balanced rather than muddy or flat — the finer-than-washed grind at 60 is doing its job of pulling clarity out of the fruit, and the 93 °C is lifting it without tipping into boozy.

➡️ Next: "Leave the recipe alone. If it starts tasting flat over the next couple of weeks, that's staling rather than the dial — don't chase it finer." || logging that cup and marking the bag dialed

2026-07-26 — BeMyBean — Sumatra Jingki Roda · 15 g · 1:16 · grind 61 · 92 °C · 60 ml/80 s bloom + single main pour¶

Amazingly fruity and very enjoyable, with just a touch of bitterness.

A nicely long bloom with a lot of foam afterwards — an anaerobic natural still shedding CO₂ at 19 days off roast (roasted 7 Jul), which is likely a decent chunk of the slight bitterness. The grind-61 bump (up from 59) was made precisely to skip the harsh first cup that El Recreo and Rwanda both served, and it looks like it worked.

➡️ Next: Leave the recipe alone — it should soften on its own over the next couple of weeks. If it is still nagging in a week, the lever is temperature, not grind: 92 → 90 °C, which keeps the body and fruit.

2026-07-16 — BeMyBean — Rwanda Cyato · ~23 d off roast now (roasted 2026-06-23); dial slot C — grind 58–59 · 93→92 °C · 60 ml/80 s bloom + single main pour · 1:16¶

🍃 The gooseberry (agrest) + florals never showed brewed young (~2 wk) + a touch bitter — the lesson: naturals need ~3 weeks' rest.

Reframes the earlier cup at 15 d off roast: the light bitterness + closed fruit (linden · peach crumble · black tea · gooseberry) reads as UNDER-REST, not a grind/drawdown fault — a young natural extracts unevenly (gassy → faint bitter) and its volatile fruit hasn't developed. Now ~23 d it should open. Lesson: give a delicate natural ~3 weeks' rest before judging or re-dialing.

➡️ Next: Give naturals ~3 weeks before judging or coarsening grind; confirm the agrest actually shows at ~3 wk before any dial change.

2026-07-13 — TESO — Rwanda Gicumbi · 15 g · 240 ml · grind 58 · 120 rpm · 60 s bloom + single main pour¶

Turning bitter as it settles — over-extracted

TUI telemetry: ran 3:25 (205 s) vs the 2:45–3:15 target — pours done ~168 s, then a ~26 s slow drawdown tail; the bed is flowing slow so the cup over-extracts, hence the bitter edge. Light washed peaberry at ~24 d off roast extracts freely.

➡️ Next: grind coarser 58→60 to speed the drawdown back into the target window

2026-07-13 — Savora Café — Filtrado Exótico · grind 59, 15 g / 1:16 / 240 ml, 120 RPM, 92→90 °C (xBloom Auto Medium, 4 pours)¶

Perfect — the fruit really comes through now; expressive and juicy 🍓

The 62 → 59 grind change fixed the previous cup's under-expression — fruit reads clearly now. Recipe lands here; any residual softness on the florals is the ~7-week-off-roast age, not the dial.

2026-07-12 — Savora Café — Filtrado Exótico · medium-auto · grind 62 · 15 g / 1:16 / 240 ml · 92→90 °C (slot C)¶

Good, sort of 'luxe' but under-expressive — lacks the floral lift of the Aromático; wants more taste

Aged ~7 wk off roast (roasted 27 May → aromatics fade with age) AND reads a touch under-extracted for an expressive/fruit-forward palate. Not a delicate-washed 'lacking' — it's a natural that can give more. Lever: grind finer to raise extraction (sweetness, body, aromatics).

➡️ Next: grind 62 → 59; if still flat, 57 and/or +1 °C. Some floral loss is the bag's age, not the dial.

2026-07-11 — KLARO — Colombia Geisha · 15 g · 1:16 · 240 ml · grind 60 · 93→90 °C taper · 3:13¶

Clean peach & citrus but a touch bitter on the finish — mild over-extraction.

TUI curve: pours finished ~2:42 but the cup dripped on to ~3:13 — a ~30 s slow drawdown tail = extended contact, the source of the bitterness. Beverage 209 g (retained ~31 g). The brew hit spec; the grind was just a hair fine for this bed.

➡️ Next: Grind coarser 60 → 62 to speed the drawdown and lighten the finish; keep the water cool (~93 °C) as the roaster asks.

2026-07-11 — TESO — La Leona · 16 g · grind 59 · 95 °C · 110 RPM · 1:15.5 (dial slot C, bag near the end)¶

👍 Much better — more expressive with more taste, at grind 59 (bumped manually on the machine).

Grind bumped straight to 59 on the machine itself (past the earlier 56 → 57 end-of-bag correction) and it's a clear improvement — more flavour, less of the muted/bitter edge. Recipe updated to match what's actually being brewed. Bag is running low.

➡️ Next: Happy as-is for the rest of this bag.

2026-07-09 — BeMyBean — Kolumbia El Recreo · 15 g · grind 59 · 92→91 °C · bloom 60 ml/80 s + single main pour (2-pour, Hedrick rebuild) · 120 RPM · 1:16 · agitation after bloom (dial slot A)¶

🙂 A bit bitter — first taste of the Hedrick 2-pour rebuild.

First brew on the 2026-07-06 restructure (long 80 s bloom, single main pour, grind held at 59). Same bitterness pattern as the sibling BeMyBean Rwanda Cyato recipe on this same Hedrick template — reads as the grind/drawdown needing one more coarsen step, not a re-rebuild.

➡️ Next: Coarsen grind 59 → 60 (per fix_guide: bitter/boozy → coarser). If still edgy after that, drop the main pour 92 → 91 °C or 91 → 90 °C.

2026-07-08 — TESO — Rwanda Gicumbi · 15 g · grind 58 · 93→92 °C · bloom 60 ml/60 s + single main pour (2-pour) · 120 RPM · 1:16 (Hedrick re-dial v2)¶

🙂 More flavourful than v1 — the re-dial worked; just not a personal favourite (bean preference, not a dialing fault).

Second brew, on the Hedrick re-dial (cooler 93→92 °C, 2-pour, coarser 58). From the live scale: 240 g water in → 210.3 g in the cup (~30 g retained), total 3:23 — pours done ~2:48, then a clean, even drawdown to ~3:14. Noticeably more flavourful and cleaner than the bitter/flat v1 (95 °C, 4 pours), so the fix landed — it's simply not a favourite here, and that's a bean-taste thing, not a recipe problem.

➡️ Next: Recipe is good — don't chase it. A bean-preference miss; drink the bag down as a milk coffee if black doesn't grab you.

2026-07-08 — BeMyBean — Rwanda Cyato · 15 g · grind 58 · 93→92 °C · bloom 60 ml/80 s + single main pour (2-pour) · 120 RPM · 1:16 · agitation after bloom (dial slot C, first brew)¶

🙂 Good — but a tiny bit bitter.

First cup on the Hedrick-clarity rebuild (grind 58, 93→92 °C, 2-pour, big long bloom). From the live scale: 240 g water in, 209.7 g in the cup (~30 g held in the bed), total 3:41 — the pours finished ~3:08 (water hit 240 g), then a slow ~30 s drawdown to the end. That long a draw on a delicate natural, at 92–93 °C, is the likely source of the light bitterness (tipping just past the clarity window). Close, not a rebuild — a small nudge.

➡️ Next: Coarsen grind 58 → 59 (speeds the drawdown AND eases bitterness — Hedrick: bitter → coarser); if still edgy, drop the main pour 92 → 91 °C. One change at a time.

2026-07-06 — ROST — Etiopia Abeba · 15 g · grind 57 · 93→92 °C bloom+main pour · 120 RPM · 1:17 (dial slot C, first brew)¶

⭐ Soft and clean — neither bitter nor sour, tea-like. Brew landed right in the 3:00–3:30 target window.

First cup on the recipe rebuilt today around the Lance Hedrick clarity method (fewer/gentler pours, long 80 s bloom, 93→92 °C, coarser + wider ratio to manage this Ethiopian's fines). Landed clean at ~3 min — soft, balanced, no bitterness or sourness, clear black tea character coming through. Flipped to dialed.

➡️ Next: Keeper as-is.

2026-07-06 — BeMyBean — Kolumbia El Recreo · 15 g · grind 58 · 93→92→90→88 °C 4-pour taper · 120 RPM · 1:16 · agitation after bloom (dial slot C, first brew)¶

🙂 Good, tiny bit bitter — lots of CO2 foaming after the bloom, even extraction, less floral than the Geisha (expected, different bean).

First cup off the starting recipe. Heavy foaming on the first pour (fresh anaerobic natural off-gassing hard) but the extraction itself was even. Only a tiny bit bitter — nudging grind 58 → 59 (coarser, per fix_guide) to smooth it out.

➡️ Next: Taste at grind 59. Still bitter → 60. If it swings sour/flat, back to 58.

2026-07-05 — BeMyBean — Kostaryka Geisha · 16 g · grind 60 · 94→92→88→85 °C 4-pour taper · 120 RPM · 1:15 · agitation after bloom (dial slot B)¶

👍 Improved, definitely less bitter — grind 60 is the right dial.

Confirms the coarsen from 58→60 fixed the bitterness the temp taper alone didn't clear. Recipe now dialed on both fronts — temp taper for the burned/scalding issue, grind 60 for the bitterness. Flipped dial_journey's "pending taste" entry to confirmed.

➡️ Next: Happy as-is.

2026-07-05 — TESO — San Pedro · 16 g · grind 57 · 95→95→94→93 °C 4-pour taper · 120 RPM · 1:16 (dial slot A)¶

⭐ Awesome, balanced cup — the researched starting recipe nailed it first try.

First brew off the researched Pink Bourbon recipe (finer + hotter + high RPM to pull the molasses/tangerine, gentle taper across the 4 pours) — landed balanced with no sour/bitter. Flipped to dialed. Nothing to change.

➡️ Next: Keeper as-is.

2026-07-04 — BeMyBean — Kostaryka Geisha · 16 g · grind 58 · 94→92→88→85 °C 4-pour taper · 120 RPM · 1:15 · agitation after bloom (dial slot B)¶

😐 A bit bitter — burned/scalding is gone (taper worked) but over-extraction crept back in.

Temp taper held (burned character stayed fixed), but at grind 58 it's reading a touch bitter today. Matches the bracket already on file (55 = slightly bitter, 58 = clean & sweet) — the cooler tail pours may be extracting a little differently than when 58 was first dialed, or the bag's just resting further. Per fix_guide: bitter/astringent → grind coarser. Nudged 58 → 60.

➡️ Next: If still bitter at 60, go 62. If it swings sour/flat instead, back off to 59 (one step at a time).

2026-07-03 — BeMyBean — Kostaryka Geisha · 16 g · grind 58 · 94→92→88→85 °C 4-pour taper · 120 RPM · 1:15 · agitation after bloom (dial slot B)¶

👍 Good — the steep taper fixed the burned/scalding cup, no sourness or thinness crept in.

First brew on the steepened 9 °C taper (up from a first, too-shallow 4 °C attempt) after the flat-95 °C recipe was coming out "burned/scalding" straight off the machine. Fronting the heat on bloom + Pour 1 (94/92 °C) protected the earlier sourness fix, and dropping hard on Pours 2–3 (88/85 °C) killed the scorched character without under-extracting the back half — the bed's residual heat likely blunted the raw temp drop. Grind held at 58 throughout (temp was the right lever, not grind). Confirmed — keeping this taper.

➡️ Next: Happy as-is. If it ever reads thin/sour instead, split the difference (94→91→89→87) before going back to flat.

2026-07-02 — Savora Café — Filtrado Aromático · 15 g · grind 60 · 92→90 °C · 120 RPM · 1:16 · agitation after bloom (dial slot A)¶

👎 Bitter af — the sourness fix overshot into over-extraction.

Brewed from the dial (slot A) on the untested aged-bag tweak: grind dropped 65→60 AND bloom agitation turned on at the same time to chase the earlier sourness. Both levers push extraction the SAME direction, and stacked together they overshot straight past clean into bitter — exactly the failure the recipe note flagged as a risk. Lesson: don't stack two extraction-increasing changes untested: one lever at a time.

➡️ Next: Turn OFF bloom agitation, keep grind at 60, re-push slot A, re-taste. If still bitter, coarsen grind 60 → 62 next.

2026-07-01 — ROST — Kolumbia Tolima Decaf · 16 g · grind 62 · 90 °C · 100 RPM · 1:16 (~256 ml)¶

A bit bitter — mild over-extraction.

Brewed from the dial (slot C). This EA-sugarcane decaf is porous and quick to extract, and it's well-rested (~2 months off roast) so it pulls even faster now — the bitterness reads as over-extraction on the base recipe, not a roast fault.

➡️ Next: Coarsened grind 62 → 64 and re-pushed slot C. If still bitter, drop temp 90 → 88 °C (one lever at a time).

2026-06-30 — TESO — Rwanda Gicumbi · 15 g · grind 56 · 95→93 °C · 120 RPM · 1:16 · agitation after bloom · 2:51¶

👍 Better — the rough/drying edge eased on the SAME recipe; cleaner and more enjoyable. Likely the extra rest opening it up.

Re-brewed the stock recipe (grind 56, 95→93 °C taper — the finer/hotter "more flavour" tweak was NOT applied yet) and it's noticeably better than the rough first cup. The brew record showed a smooth, even staircase and steady drawdown at 2:51 — extraction's healthy. The improvement reads as rest: now ~11 days off roast, and washed peaberries tend to open up ~14–18 days, exactly as flagged. So the earlier roughness was partly under-rest, not just the recipe. Still headroom to push the fruit further.

➡️ Next: Happy → keep as-is. Want MORE fruit/expression → grind 56→54 + hold 96 °C (don't taper to 93) now that it's more open.

2026-06-29 — Savora Café — Filtrado Aromático · BLEND ½-caf · 10 g Savora + 6 g decaf · Savora medium-auto preset (grind 65 · 120 RPM · 92→90 °C) · 1:16¶

👎 Blend flopped — the two just muted each other into a flat, blurred middle. Single origins win; drink them solo.

Tried a Savora-led half-caf with the ROST sugarcane-EA decaf (Colombia × Colombia, both medium, so on paper compatible — winey fruit + the decaf's vanilla/molasses sweetness). In the cup it didn't add up: the preset's coarse grind 65 / 120 RPM over-extracts the porous EA decaf while damping the Savora's winey fireworks, so instead of "fruit + sweetness" you get a dull, blurred middle. Confirms the rule that two distinct, good single origins tend to cancel rather than stack. The half-caf idea is fine — the execution isn't worth chasing.

➡️ Next: Drink Savora and the decaf separately. Half-caf itch? Go Savora-dominant + tighter: 12 g Savora / 4 g decaf, grind ~62, single 91 °C.

2026-06-29 — TESO — Rwanda Gicumbi · same recipe + milk¶

🥛 Better with milk — the rough edge rounds right out. A proper milk coffee; went down a treat at home.

Made a milk version and it's clearly better than black — the slightly rough/robust black-tea edge that made it "a coffee you'd add milk to" smooths into a balanced, friendly milk cup. So this bag lands as a milk coffee (latte/flat-white) rather than a black sipper. Recipe unchanged.

➡️ Next: Use it as the milk-coffee bag. Black & want it smoother? grind 56 → 58 or last pour 91–92 °C.

2026-06-29 — TESO — Rwanda Gicumbi · 15 g · grind 56 · 95→93 °C · 120 RPM · 1:16 · agitation after bloom¶

☕ Coffee-forward and serious with clear black tea — but a bit ROUGH: tastes like a coffee you'd add milk to, not a delicate sipper.

First brew on the researched 4-pour recipe, with machine agitation after the bloom enabled (the shaky icon). Brew 3:01; the extraction log showed a smooth, steady drawdown — so it's extracting evenly; the roughness is a flavour edge (mild astringency), not a stall. Reads as an everyday, milk-friendly coffee rather than a clean/floral sipper, with the packet's black tea clear. To smooth the rough edge: grind a touch coarser (56 → 58) or drop the last pour to 91–92 °C. Left it dialed for now (sound + drinkable), but there's a smoother version to try.

➡️ Next: If chasing smoother/cleaner: grind 56 → 58, or last pour 91–92 °C (one lever at a time).

2026-06-28 — ROST — Kolumbia Tolima Decaf · 16 g · grind 62 · 90 °C · 1:16 (256 ml) · spiral pours¶

⭐ Surprisingly good for a decaf — smells like ordinary coffee but tastes clean with real fruit. Like La Leona with a Colombian fruity twist.

First brew on the researched decaf recipe — 2:48, smooth steady drawdown. Decafs usually taste flat/hollow; this one actually has body and fruit (a Colombian twist on the clean, tea-like La Leona profile). The aroma is plain "average coffee," but the cup over-delivers. The researched starting recipe landed first try → flipped to dialed. Nothing to change.

➡️ Next: Keeper as-is.

2026-06-28 — TESO — La Leona · 16 g · grind 56 · 95 °C · 110 RPM · 1:15.5 · spiral bloom¶

🍵 Plain, nice and universal — a clean, easy crowd-pleaser rather than fireworks. Spiral bloom gave even saturation.

Second cup on La Leona, now with the spiral bloom (the fix for the dry-edge problem on a centered pour). Brew 2:42; the Studio's record showed a smooth, steady drawdown — even saturation confirmed. Reframed it: my earlier "lacking" was me expecting Geisha fireworks — but this is really an approachable, everyday cup (a friend tasted it and called it "good, like normal coffee"). Clean and delicate is the bean's nature, not an extraction fault, and that makes it a genuine crowd-pleaser. Recipe unchanged.

➡️ Next: Optional only — RPM 110 → 120 for a touch more floral lift; otherwise it's a keeper as-is.

2026-06-28 — Savora Café — Filtrado Aromático · ICED flash brew · 15 g · 1:10 · 150 ml · grind 58 · 120 RPM · 92 °C · 85 g ice¶

⭐ Very good iced — bright and juicy, leans sour, and I like it that way.

First flash brew off the new iced recipe, and it works. The brightness/sourness is partly the method (flash-chilling accentuates acidity) and partly a touch of under-extraction at grind 58 — but it's tasty and exactly to taste, so it stays. Keeper. If I ever want it rounder/less sour: grind a touch finer (58 → 56) or cut the ice slightly — one lever at a time.

➡️ Next: Happy as-is. Only if craving less sour: grind 58 → 56.

2026-06-28 — BeMyBean — Kostaryka Geisha · 16 g · grind 58 · 95 °C · 120 RPM · 1:15 · agitation after bloom¶

⭐ Perfect this morning — clean, sweet, balanced. The bloom-agitation experiment paid off.

The test cup for the "agitation after bloom" change turned on last time — and it landed: a genuinely perfect morning cup, no sour/bitter, with more sweetness/body than before. Caveat, honestly: hard to fully separate the agitation from it just being a great morning coffee for my palate — but no reason to touch anything. Recipe stays as-is (dialed, agitation on).

2026-06-27 — TESO — La Leona · 16 g · grind 56 · 95 °C · 110 RPM · 1:15.5¶

🍵 Clean, balanced and tea-like — no bitterness or sourness; grind 56 is on the money. A delicate, elegant cup. Could use a touch more aroma/expression.

First cup off this bag (~22 d off roast). Extraction is dialed — no sour/bitter means grind 56 is right, so don't touch grind. The "muted / no smell / wants more" is partly the bean (a washed Geisha is inherently delicate, clean, tea-like — not a loud natural) and partly expression to push via RPM/ratio. Flipped the site recipe to dialed after this cup.

➡️ Next: RPM 110 → 120 for more floral lift (then ratio 1:15 for more body).

2026-06-27 — BeMyBean — Kostaryka Geisha · 16 g · grind 58 · 95 °C · 120 RPM · 1:15¶

👍 Less bitter, tastes ok — the bracket closed (70 sour → 55 bitter → 58 clean). Dial landed.

Change made (untested): turned "Agitation after" ON for the bloom — the next cup is the test (does it add sweetness/body, or push back toward bitter?). Grind 58 alone was already good, so this is a bonus experiment, not a fix — easy to revert.

2026-06-26 — Savora Café — Filtrado Aromático · medium-auto preset · 16 g¶

⭐ Juicy and sweet — strawberry, winey fruit, caramel. The built-in medium-auto preset nails this medium natural/anaerobic with zero changes. A standout.

Favourite cup so far. Paired with a Kinder Bueno 🍫 (hazelnut-choc against the caramel + winey fruit — great match). ~24 d off roast and still singing — drink it up before the winey fruit fades.

2026-06-26 — BeMyBean — Kostaryka Geisha · 16 g · grind 55 · 95 °C · 120 RPM · 1:15¶

🙂 Better — sweeter, the sour/muted gone, but a slight bitterness crept in. So 55 went a touch past; aim ~58.

Brew time 2:24. Bracketed it: 70 = sour (under) · 55 = slightly bitter (over) → sweet spot ≈ 58–60. The 2:24 time is set by the 3.0 ml/s flow schedule, not grind — chase taste, not time.

➡️ Next: Grind 55 → 58 (then temp 92–93 °C or ratio 1:16.5 if still bitter).

2026-06-26 — BeMyBean — Kostaryka Geisha · medium-auto preset · 16 g¶

😐 Sour and muted — the medium-auto preset under-extracts a light Geisha. Needs its own recipe: finer + hotter.

Clean A/B vs the Savora (same preset, different bean). Diagnosis: under-extraction of a light filter roast — sour + muted + a bitter/drying finish once cooled (that's astringency, not over-extraction). ⚠️ The generic "bitter → grind coarser" rule does not apply to an under-extracted light roast — grind finer. Lesson banked: light/filter roasts want a finer grind (and ideally their own preset); a "bitter finish on a light roast" = under-extraction/astringency → finer, not coarser. Research note: no published xBloom recipe for this bag; light Geisha + raisin-honey → hot water ~93–96 °C + finer grind + ~1:16–17. The medium-auto preset under-extracts on two fronts (grind too coarse and temp too low, ~88–91 °C) → use a light preset, not just a grind nudge.

➡️ Next: Custom recipe — grind finer ~3–5 steps, optionally RPM 120; drink it hot.

2026-06-26 — Savora Café — Filtrado Aromático · medium-auto preset · 16 g¶

👍 Pretty good — a solid first brew straight off the auto preset (first cup on the new Studio).

Fruit-forward natural; to brighten it, grind a touch finer (~2–5 steps) and/or RPM toward 120 — but "pretty good" on the auto recipe is a perfectly fine place to live.