Cristal, Louis Roederer 2004

The 2004 Cristal is nuanced and classy right out of the gate. Hints of lemon confit, marzipan, tangerine oil, spice and chamomile are all suggestive of a Champagne that has arrived at its first inflection point of early maturity. The interplay of slightly more mature notes with a good deal of freshness makes for an incredibly delicious wine to enjoy now. As it turns out, I tasted the 2004 again the following day and found it even more vibrant than the bottles we served at this lunch. This remains one of my favorite vintages here. Two thousand and four was the highest-yielding vintage in Champagne at the time, yet the best wines have always been compelling.

The 2004 Brut Cristal has put on quite a bit of weight since I first tasted it earlier this year. It is a powerful, structured Cristal layered with considerable fruit. Chardonnay seems to play the leading role in 2004, at least today. Cristal is often accessible young, but that is far from the case here. This is a serious, painfully young Cristal that will require considerable patience. Readers who are willing to spend some time with the wine today will find a super-impressive, complete Cristal. The 2004 Cristal is 60% Pinot Noir and 40% Chardonnay. This is Lot L033331E100008, disgorged January, 2010. Anticipated maturity: 2014-2034. My visit to these historic cellars earlier in the year was an eye-opening experience. I spent several hours with Chef de Caves Jean-Baptiste Lecaillon tasting through a wide range of 2009s vins claires. Readers may want to take a look at my feature on www.erobertparker.com for more on those wines. The visit was capped off by previews of the 2004 Cristal and the as-yet-unreleased Cristal Rose. Both were superb. Over the last six months the 2004 Cristal has come together beautifully and is shaping up to be a truly great, monumental Champagne. From top to bottom, this is an impressive set of wines. My only real criticism of Roederer is the estate's insistence on using lot numbers that resemble missile launch codes in their complexity. Surely something simpler must be possible.