Is wine made in the vineyard or the winery?" is one of those questions that crops up time and again in wine circles. The widely held consensus is that the best wines come from the finest grapes with minimal intervention in the winery. It is never going to be possible to create liquid magic if the fruit is not up to the job.

Far from suggesting that winemaking is easy, I believe it is an art, one that is one-third skill, one-third hard work and one-third intuition. Vineyard visits are always a wonderful way to see at first hand what it takes to make good wine.

Just as this year's harvest was starting, I was taking part in the 35th Sandeman Port Seminar in Oporto and the Duoro valley in Portugal. It was an eye-opening few days and a trip I'll not be quick to forget.

Whether you are the man in the vineyard, the winemaker, the master blender or the final consumer, it seems to me that port is the procrastinator's nightmare.

Let's start at the beginning with the vineyard. Frighteningly, in excess of 80 grape varieties are authorised for making port. The reality is that substantially fewer varieties are used. The following gang of five are the most commonly found: touriga naçional; tinta barroca; touriga francesa; tinta roriz and tinta cão.

No vines last forever and there is a continuous programme of grubbing up and replanting with new vines. It is an expense and a gamble. What varieties will do well in which plots? Will climate change result in differing characteristics?

What will the consumer be demanding in terms of taste and styles of port in ten, 20 and 30 years' time?

When they're deciding what and where to plant, the individual qualities of each variety have to be taken into account - in relation to current and future markets, the growing conditions and the need to create a consistent product.

A quick tasting of the 2006 vintages of each individual grape variety brought home as clear as day the different qualities of each of the varieties.

Touriga naçional was robust, intense and that it is a variety high in sugar was perfectly evident.

By comparison, touriga francesa was much more aromatic and delicate. It demands very hot sites to deliver respectable yields.

Tinta Barroca is a real tough-nut and will thrive in the arid, north-facing slopes. The wines it gives are rich in colour and have good texture. Tinta roriz is the Portuguese name for tempranillo and what was so striking in the Duoro was the intensity of the colour in comparison to its Spanish cousins.

I fell for tinta cão. It is less obviously attractive for all sorts of reasons. It is paler in appearance, it doesn't yield as generously and it has a tendency to oxidise. That said, it was really aromatic and seemed to be more refined than the other four. You could see that it would add that bit of something to the final wine.

Port is, of course, a blend. Not just a blend of grape varieties but of different vintages too. A house such as Sandeman isn't just making one wine either. With over ten wines in the range, the winemaker's got his hands full. Consider every morning having to get your head around white ports, tawny ports, ruby ports, LBV (Late Bottle Vintage) and vintage ports.

As the new vintage wines arrive in the cellar you need to decide if it's been the sort of year that warrants being declared a vintage. That aside, you still have to choose which wines will go into which ports. Will it be part of a 40 year-old tawny, or is it destined for the Founders Reserve'? The latter is made from a selection of the finest lots' of each vintage and then aged for five years.

The decision is unlikely to be immediate. You will be watching closely to see how the wine develops and assessing it's ageing potential, its colour, what aromatic qualities it has. It is a never-ending, forever changing challenge.

Then there's the final blend. Becoming a port master blender isn't the job you get at the end of a four year degree.

No, it is the role you move into after years of experience and learning from those who have been blending before you. In fact, it is less of a job and more of a craft - a calling even.

Sandeman kindly bedecked us in master blender white coats and presented us with a glass of port they wanted us to recreate out of the three vintages set in front of us. Yes, only three! It was a humbling experience.

I tasted the first - the one we were seeking to emulate. It had fresh, vibrant fruits. The tannins were evident but in no way overbearing. It was perfect. It was the combination of 2003, 2004 and 2005. The latter was the gutsiest of the three, with the most pronounced tannins.

The 2004 was, it seemed to me, à point and had better integrated fruit. 2003 was elegant and the closest in colour to the wine we were endeavouring to replicate.

Armed with lots of empty glasses and a pipette, I got started. It was impossible! My first attempt was too sharp and lacking in fruit. The second was all fruit and no structure. I kept sniffing and mixing and still seemed no closer.

In the end I settled for a blend that didn't seem too far removed from what we were looking for.

Master Blender, Carlos Silva, was only moderately impressed. It is true that I had managed to get pretty close to the taste but I had used too much 2003 - and that was unrealistic given dwindling stocks.

If I had added more 2005 it would have been that bit fruitier and more commercially viable. It seems that the choices aren't solely about taste but about economics too!

I am not surprised the winemakers in Oporto are such a happy, relaxed and welcoming bunch.

Clearly, it is because they are as enthusiastic about drinking it as they are about making it. The only decision now, is which port to try.

I recommend the following: Tawny Port (£7.25): Nutty, elegant with sweet spice.

Founders Reserve Ruby Port (£8.75): Well integrated fruit with refreshing black fruits.

Late Bottle Vintage (£8.75): Powerful, black, caramel-coated fruits with well-integrated tannins.