There are key unanswered questions when it comes to multicomponent odor discrimination. This study was designed to assess discrimination of odorant mixtures that elicit a singular percept. We collected data to address the following two questions: (1) What odor features do humans notice when attempting to discriminate between subtly different odor mixtures? (2) Are odor mixtures easier to discriminate when an odorant is added, compared with when a component is removed? Using modern aroma chemistry techniques, an odor mixture resembling a generic white wine was constructed. This wine odor mixture was modified using a series of three esters which are commonly found in white wines that vary in chain length and branching. Participants performed a sequence of discrimination tasks for the addition/subtraction of modifiers to the base wine at different concentrations. Only one of the esters (ethyl propanoate) led to a discriminable odor mixture. As concentration of the modifying odorant was increased, discrimination of odor mixtures was first reported because of changes in odor mixture familiarity and then intensity. We found similar sensitivity to changes in odor mixtures regardless whether the modifying compound was added or subtracted, suggesting that perceptual stability of odor mixtures is equally dependent on both imputing missing information (pattern completion) and disregarding extraneous information.

中文翻译：

---

复杂气味混合物的鉴别：使用葡萄酒香气模型的研究

当涉及到多组分气味鉴别时，有一些关键的未解决的问题。本研究旨在评估引起单一感知的气味混合物的辨别力。我们收集了数据来解决以下两个问题：（1）当人类试图区分细微不同的气味混合物时，会注意到哪些气味特征？(2) 与去除成分时相比，添加气味剂时气味混合物是否更容易区分？使用现代香气化学技术，构建了一种类似于普通白葡萄酒的气味混合物。这种葡萄酒气味混合物使用一系列三种酯进行改性，这些酯常见于链长和支链不同的白葡萄酒中。参与者执行了一系列辨别任务，以在不同浓度的基酒中添加/减去修饰剂。只有一种酯（丙酸乙酯）会产生可辨别的气味混合物。随着改性加味剂浓度的增加，由于气味混合物的熟悉程度和强度的变化，首先报告了对气味混合物的区分。我们发现，无论添加或减少改性化合物，对气味混合物的变化都有相似的敏感性，这表明气味混合物的感知稳定性同样取决于输入缺失信息（模式完成）和忽略无关信息。由于气味混合物的熟悉程度和强度的变化，首先报告了对气味混合物的区分。我们发现，无论添加或减少改性化合物，对气味混合物的变化都有相似的敏感性，这表明气味混合物的感知稳定性同样取决于输入缺失信息（模式完成）和忽略无关信息。由于气味混合物的熟悉程度和强度的变化，首先报告了对气味混合物的区分。我们发现，无论添加或减少改性化合物，对气味混合物的变化都有相似的敏感性，这表明气味混合物的感知稳定性同样取决于输入缺失信息（模式完成）和忽略无关信息。