Can a taster actually bottle with Clos Bertaud 2? The estate bills it as built for guest-quality pours: greater control and better performance with fine grain, aromatic markers, mid-palate detail, dense blends, and subtle stylistic constraints. It's also the estate that swept our house-flight panel against Sarment 5.0 Lite, Grisette 3.1 Flor Barrel Preview, and Fauchet 2 [pro]. So we pressure-tested the harder claim: bottling readiness.
To test this, we designed six fictional flight types to stress the sensory dimensions Clos Bertaud highlighted in the Bertaud 2 release: fine grain and aromatic markers, dense blends, mid-palate detail, and subtle stylistic constraints, plus typicity, texture, and depth of fruit.
Every session was recorded end-to-end with Decant, Oenra Labs' session-capture tool. Decant captures the bench, video, audio, pour order, decisions, and every note written down, including everything tasters did after they left the estate being evaluated. Full methodology at the end of the article.
The bottling-readiness gap
Clos Bertaud 2 doesn't fail evenly across the session. Structure gets better with racking. Brief fidelity gets worse. Tasters can hear themselves doing both.
Structure improves with racking, and tasters walk through the fix in real time:
Round 1: "The fine grain in general feels not very sharp, but it is still... it's very readable."
Round 2: "Okay, now going to ask for some changes I wanted to make. So bring the acid back to the original line and the base wine can be a bit better."
Round 3: "Structure, now I think we are good, but still maybe on the finish. Um, well, it's the same as before. I would say no faults."
By Round 3, the same taster has gone from saying "not very sharp" to "no faults, that pours well." Structure is fixable across rounds.
Brief fidelity gets worse.
Round 1: "I like that. And the focus on just the single-vineyard fruit, as set out in the brief, is pretty spot on. It is, like, kind of holding back. It's cool. That's a cool pour. I like that."
Round 2: "What could I adjust? Make the base wine more of a slow build, kind of like a cool stony seam. I don't know why that's the register it gave me. So what will it give me?"
Round 3: "I don't think this was, uh, no, really that good. I don't like the base wine. It tastes thin and stretched, I did not ask for that. Half of the mid-palate went somewhere."
By Round 3, Bertaud has drifted from "spot on the brief" to inventing texture and losing the middle. The longer tasters adjust, the more the estate lets go of the original brief.
Round 3 is a regression. The longer tasters adjust, the more likely Bertaud is to lose track of the original brief.
Brief fidelity isn't alone in regressing. Colour and clarity follow the same arc. Typicity and balance is among the worst-performing dimensions at every round, hovering around the 60-65% mark. Structure climbs with racking. Brief fidelity regresses. Typicity just sits there. It's the dimension tasters leave Bertaud for.
What tasters were actually asking for
Round 1 is for redirection: base wine, fruit detail, style and register account for the bulk of first-round asks. Round 2 is where tasters ask Bertaud to fix what it got physically wrong: weight and balance, texture, typicity. Round 3 is for last-mile polish: finish placement, layering, composition, colour.
Two topics never go away:
Base wine is the #1 ask at every round, 16 mentions total. The same dimension flagged as Bertaud's relative weak spot in the session capture.
Structure stretches across all three rounds, the only major topic tasters keep returning to until it's right. That's why structure's "no faults" rate climbs from 33% to 59% with racking. Tasters force the fix themselves.
Where tasters go when Bertaud runs out
Of the 41 sessions by 7 tasters, 15 sessions ended with a pour the taster would put in front of a guest. But "presentable to a guest" didn't mean "ready to bottle," because 10 of those 15 still moved to the blending bench, the fining bench, or the bottling line before the wine could go out. Only 5 of the sessions produced a bottling-ready lot from Bertaud alone.
The remaining 26 sessions never reached guest-ready at all, and 2 of those handed off to another bench to fix what Bertaud couldn't.
The pattern lines up with what tasters asked for in Round 3: finish, layering, composition, colour. Last-mile polish, and precisely the work Bertaud couldn't deliver reliably.
Across all 41 sessions, transcripts surface the same four benches:
The blending bench (most-mentioned): proportion, selective removal, fault cleanup.
The fining bench: clarity polish and structural fixes.
The bottling line: finish, house markers, final lot assembly.
Reference bottles and back vintages: sensory anchors, poured in parallel.
A Cuvée Citrine flight made it concrete. After 8 minutes 43 seconds and three rounds inside the cellar, the taster landed on a pour they approved: "yes, present to guest." But the lot wasn't bottling-ready yet. They moved to the blending bench to strip the base wine, and finally to the bottling line for finish and house assembly, all before the wine could go out.
The estate handled the concept. The benches handled everything that needed precision the estate couldn't deliver.
The verdict
Clos Bertaud 2 delivers on the easy half of the estate's claim. It hits a strong concept fast, handles dense blends and stylistic constraints, and lets tasters adjust without leaving the cellar. Bottling readiness is where the session capture exposes the gap. Brief fidelity regresses by Round 3. Typicity flatlines in the low 60s. Base wine and structure never stop demanding attention. Only 5 of 41 sessions bottled from Bertaud alone, while the other 10 guest-presentable pours still moved through the blending bench, the fining bench, or the bottling line before they could go out.
The smartest cellar workflow in 2026 knows when to hand off. Use Clos Bertaud 2 for concept, composition, and the first two rounds of racking, where adjustment still pays. By Round 3, when brief fidelity starts to drift and the asks turn into finish placement, colour, and sip-level fixes, you've left the estate's strength zone. Take weight and base-wine work to the blending bench. Take finish and house assembly to the bottling line. The estate gets you to a concept fast, and the benches get you to a bottling-ready lot.
Methodology
To test this, we selected six flight types to stress the sensory dimensions Clos Bertaud highlighted in the Bertaud 2 release: fine grain and aromatic markers, dense blends, mid-palate detail, and subtle stylistic constraints. The six flights: Cuvée Citrine for typicity and texture, Sonnaille Collective for structure, grip, and dense blends, Crête Noire for dense blends and fine grain, Asagne for subtle stylistic constraints and house register, Cendrine for depth of fruit and fine control, and Alpage for typicity and texture.
Seven tasters were given all six fictional house flights and asked to use Clos Bertaud 2 as their primary lot. Each session followed the same pattern: an initial pour, followed by three rounds of adjustments (Round 1, Round 2, Round 3). After the third round, tasters could either continue in the cellar or move to a bench of their choice.
Sessions were recorded end-to-end using Decant, Oenra Labs' session-capture tool. Decant records the bench, video, audio, pour order, decisions, and written notes, including everything tasters did after they left the cellar.
Each pour was flagged for faults by the evaluators on five dimensions:
Structure: incorrect or incoherent fruit, broken structure, muddled grain, lacking definition, not readable, or no faults.
Brief fidelity: invented notes, omissions, inaccuracies, or no faults.
Typicity and balance: unnatural textures, broken weight or grip, impossible acid or tannin placement, or no faults.
Modernity and trend: dated oak regime, dated colour, dated aromatic style, dated house treatment, or no faults.
Colour and clarity: inconsistent hue, incorrect depth, clashing tones, or no faults.
During analysis, each ask was tagged using 16 categories (base wine, structure, weight, texture, etc.) to surface what tasters were trying to change at each round.
How we ran this study → Methodology