Flavours in a wine

Training on Olfactory characteristics of a wine in Tasting Panel of Vila Viniteca

What are flavours in a wine? Where do they come from? How can we learn to identify them? How are they classified and how can we reproduce them on a wine base? These and many other doubts came out during the training of the olfactory characteristics of our Tasting Panel.

Our Tasting Panel evaluates the flavours found in a wine according to different trials, which consist in answering the question about if the analysed sample contains or not a specific characteristic (yes/no). This has been trained to differentiate both positive and negative flavours. We mean by positive flavours those which mainly come from the variety of a wine elaboration or from vinification and aging process. Negative flavours, also need defects, are the result of any interaction during the fermentation or aging process and use to generate unwanted flavours. Initially we pretended that the Panel differentiate between 21 families of positive flavours and 9 defects.   .

What are wine flavours?

Wine flavours are, after their appearance, their letter of presentation. Flavours are, mainly, volatile compounds that by its composition are released easily and are fast captured by the receptors located in the nasal cavities and pituitary gland, in charge of the olfactive sense.

There can be found a lot of flavours In a wine belonging to different families as flowers, citrus, tropical fruit, aromatic herbs, spices or toasted and chemicals. It was decided to not get pleased with the 21 families which had been chosen initially and train them in detail differentiating the specific flavour in each family, in order to obtain more precise information of each wine which the Panel would analyse. For instance, for the family “tropical fruit” the panellists would be trained to identify the flavour of pineapple, mango and passion fruit. This was an important challenge because in this way, the Panel would be able to differentiate a total of 54 descriptors of positive flavours (24 specific of white wines, 16 specific of red wines and 14 found in both), plus the 7 defects.


In order to start training the panellists in different flavours, it was thought that the most appropriate was to present the natural products from where originally came the flavours, for instance orange skin, coffee beans or elderflower. In this way we reinforced the olfactive memory of the panellists, due to the fact seeing the product was easier to associate a smell or an image.

The challenge was to reproduced then all the flavour on a wine base to make the reference patterns. At the beginning we decided to carry out macerations of the natural product with wine during several hours, according to the product nature. A method which resulted effective, but very complex. In order to speed up the process, it was chosen to search for the aromatic molecules of each flavour. Being advised by different providers it was concluded that it wasn’t the most appropriate manner to reproduce the flavours in a wine due to the fact its aromatic structure is much more complex than only one molecule.

The second option was to reproduce the flavours on base of natural extracts in liquid stage to dilute them lately in a wine base. We contacted different providers and we tried to use this method from 44 of 59 reference patterns of the positive flavours (on different moments the patterns serve for training the same descriptor).

Related to the defects, we contacted a recognised company in oenological products which facilitated us the prepared molecules in order to reproduce them in a wine bottle.

Initially, so much as to positive flavours as for negative flavours the panellists were trained in a border of perception named “consumer level”, that is to say that the average of population can perceive. Later we decrease the concentrations and the panellists were trained with patterns with a minor intensity with the objective to reach the named “expert level” which allows them to be more precise.


Because of different factors, we find that some natural extracts don’t work on a wine base either because the chemical compound wasn’t well dissolved, due to a low concentration and it was unnoticeable, or because the flavours were very far from the searched descriptor. Finally, we only validate 33 natural extracts from 44.

Related to the 26 patterns of the positive remaining flavours, we chose to reproduce 18 by means of macerations and eliminate 8, because we do not find the manner to reproduce them accurately in a wine, taking into account that, the family to which they belong is already represented with other validated flavours.

We are not enough satisfied with the result of 1 of the 9 defects and we decided to discard it. Anyway, it is one of the less common defects in a wine.


Finally, we trained our panellists in order to be able to differentiate between 46 different flavours and 8 defects. To achieve it, we created 52 patterns of positive flavours (some patterns serve to train a same descriptor) and 8 patterns of defects. The patterns are reproduce don a white and red wine base, in some cases we consider that some flavours can be found in both wine types, for instance, vanilla flavour.

We detail the families and descriptors of positive flavours used in our Panel, as well as the patterns with which we train the panellists and their elaboration manner, as follows:

We detail the descriptors of negative used flavours in our Panel, as follows:

To sum up:

We do periodic reminders during the Panel sessions, as the flavours are some of the characteristics with more difficulty of memory retention, as well as reliability sessions which allow us to check that all the panellists know how to identify them precisely.Flavours training has been one of the most ambitious challenges we carried out in the Panel and for which we had to confer more time. More than half of the training hours have been allocated to flavour characteristics. A total of 35 hours – of 60 hours- were carried out during these 2 years and a half.

Other sensorial characteristics related to the flavours and evaluation methods.

Flavours and defects are evaluated with differentiating tests, which consist in answering if the wine has or not some of the descriptors of which they have been trained.

Then it is also evaluated its intensity in different moments and conditions by means of linear scales of 10 points.

We evaluate:

  • Flavour intensity at still glass: olfactory ability of the wine without being moved.
  • Flavour intensity at swirled glass: olfactory ability of the wine after stiring it inside the glass.
  • Defect intensity (if any is detected): defect ability of being perceived when smelling. We can distinguish if, analysing this parameter, in case of high intensities of some defects, a wine should be considered as not eligible for its consume.
  • Flavour intensity on the palate: flavours abilities of a wine to be perceived when the wine i son the palate.
  • Global persistence of flavour or detected flavours: related to the ability of flavours in a wine to remain latent, once it disappears from the palate. It is calculated with caudalies, 1 second is the equivalent of 1 caudalie. This parameter is analysed by using a chronometer.

Finally, in order to help our panellists, we create a wine tasting protocol to learn how to improve the identification of different flavours (Annex 1)

Annex 1

Evaluate the samples in olfactive manner:

  1. Introduce the nose inside the glass, at still glass (without stirring its content), taking short breaths of about 3 and 4 seconds in order to keep the air moving inside the nose and not going into the lungs. We take away the focus of the smell meanwhile we are evaluating it.
  2. If necessary, we will smell our skin in order to let resting our sense of smell.
  3. We stir slowly the wine in the glass
  4. We smell again, this time introducing less the nose inside the glass so that the evaporated alcohols do not affect us in the flavours detection. This second breath should not last more than 5 seconds.
  5. Repeat the process the second time if necessary.

Hits: 1360

Leave a Comment

Your email address will not be published. Required fields are marked *