on the scale. For each training session, standards solutions were
freshly prepared in a dry Riesling wine from the Staatsweingut
Neustadt matching the vintage of the experimental wines. Each
tasting booth was furnished with a complete set of sensory
standards. During four training sessions, judges were familiarized
with the standard solutions which were presented solitary and in
binary mixtures of varying strength. Instant feed-back was given
after completion of each task by the computer program and
re-tasting was possible. Evaluation of bitterness was assessed as
‘‘bitter intensity’’ and ‘‘bitter persistency’’. To broaden the molecular
base for bitterness, panelists were trained with caffeine and
ethyl gallate standards which were presented in equi-bitter
concentrations and subsequent dilutions.
During descriptive analysis, six wines were presented in randomized
order following a Latin square design. Replicates were
served in different sessions as well as varieties. Attribute intensities
were rated on a 10 cm unstructured line scale labeled with
‘‘not noticeable’’ on the left end and ‘‘very strong’’ on the right.
Odor descriptors were assessed in a comparative set-up, rating
the intensity of one attribute in all wines before assessing the next
attribute. In contrast, orally perceived attributes were evaluated
monadically, assessing all descriptors in one wine before moving
to the next sample after a two-minute break.
Temporal dominance of sensations analysis
For TDS analysis, the four in-mouth attributes used during DA
(sweet, sour, bitter and astringent) were selected plus the hot sensation,
as requested by the panel. Retronasally perceived attributes
were omitted from TDS analysis due to the focus on taste-related
modifications and panel consensus. Three training sessions were
conducted presenting different standard dilutions to ensure that
each modality was recognized correctly. To familiarize panelists
with the time course of perceived sensory stimuli, judges listened
to an audio track consisting of two sounds (195 and 440 Hz) at
increasing and decreasing intensities. Judges were requested to
assess the dominance of one sound over the other and the correct
results were presented immediately after finishing the audio task.
Finally, TDS assessments of wines were done followed by group
discussion of the results.
Each TDS descriptor was represented by one button on the computer
screen. The sequence of attribute buttons differed from judge
to judge to compensate for order effects. However, each judge
worked with the same sequence of attributes to ease the search
for the appropriate button. Sample assessment was started by
clicking the start button parallel to sipping the wine and judges
were requested to click on the button, representing the currently