Sensory Thresholds: Beyond the Numbers
Finding ways to do better with the information we have
Dec 4, 2023
Sensory thresholds, sometimes referred to as recognition thresholds, are the minimum concentrations of a particular flavor compound that can be detected by the human senses (for our purposes, smelling and tasting). Within beer, sensory thresholds can help us assess quality and better understand flavor profiles. Beyond beer, they are used to set safety standards for hazardous substances or to develop new food flavors and fragrances in household products, cosmetics, and more. We use sensory thresholds in our own work to help us determine flavor standard concentrations for attribute training as part of our sensory program. We also use these methods to set the baselines for an aroma when evaluating new yeast strains.
How are sensory thresholds determined?
Several methods are commonly used to determine sensory thresholds, but one of the most common methods involves decreasing the concentration of a stimulus (like a specific flavor compound) by small amounts until an individual can no longer detect it. This method is repeated with many different tasters, and the average concentration at which the flavor is no longer detectable is considered to be the sensory threshold. Conversely, another method involves repeating exposure to the flavor compound at increasing intensities and the taster indicating the first detectable sample. Those managing these panels will then average the proportion of detectable and undetectable responses, and determine a threshold where the compound is detectable with a probability of 0.5 by 50% of the group. These tests are often conducted in studies and then used to frame the experiments conducted throughout the study.
Common yeast-derived flavor thresholds
| Compound | Sensory descriptors | Threshold |
|---|---|---|
| Acetaldehyde | green apple, solvent, paint, pumpkin | 10 ppm |
| Diacetyl | butter | 30 ppb |
| Ethyl acetate | nail polish, solvent | 25 ppm |
| Isoamyl acetate | banana, circus peanuts | 1.6 ppm |
However, breweries shouldn’t be expected to conduct threshold testing to understand their staff’s strengths, and threshold testing is meaningless if panelists don’t know what they’re looking for. Regular flavor training should be supplemented by difference and recognition tests, and panelists’ test results can be tracked to help understand where panelist sensitivities lie and where further flavor training is required. It’s worth noting that training at threshold levels can be pretty difficult — most panelists will have a hard time recognizing a new flavor at such low concentrations. It’s often best to flavor train at concentrations anywhere from two to six times the sensory threshold. For example, if a given flavor’s threshold is 20 ppb, tasters will be more likely to be able to discern a difference between the altered sample and an unaltered, control sample the first few times at 40 – 120 ppb, and that concentration can be reduced later as they become familiar with the flavor and more capable of recognizing it.
Triangle and tetrad tests are two commons methods of validating sensory panelists’ training. In both tests, tasters are tasked with identifying differences between a group of three or four samples. Typically one sample group will represent a control, while the other differs in some way.
Variables that affect sensory threshold perception
Individual differences
The sensory threshold of a particular flavor compound varies depending on the individual taster. Two people may experience the same amount of a flavor compound entirely differently — taster A may find a given flavor to be incredibly subtle while taster B perceives it at a moderate intensity. In general, however, there is a range of concentrations within which a particular flavor compound can be detected. If the concentration of a flavor compound is below its sensory threshold, then it will not be detectable by most, if any, tasters, whereas if the concentration of a flavor compound is above its sensory threshold, then it will be detectable by most tasters.
Adaptation
Adaptation is a trick of our brains that keeps us from being constantly overstimulated. Much in the way that we eventually stop noticing a scented lotion, or how we don’t notice the lingering smell of the dinner we cooked until we leave the room and come back, our brains are built to make us notice what’s new and different, rather than what is constantly surrounding us.
If a taster is exposed to a particular flavor compound for a long period of time, their sensory threshold for that compound may increase. This means that they may need a higher concentration of the compound to be able to detect it. One great example of adaptation is the lightstruck compound, or 3‑methyl-2-butene-1-thiol (3MBT), a skunky-smelling compound that arises when beer is exposed to light. 3MBT has an incredibly low sensory threshold — just 2ng/L (that’s 0.002 ppb!) is enough to be detectable — but it is also incredibly easy to adapt to, and some tasters may find it difficult to revisit a sample with 3MBT and detect it as easily as they did on the first pass.
Context
The context in which a flavor compound is experienced can also affect its sensory threshold. This is also referred to as the matrix effect: the changes observed in the detection or quantification of an analyte (in our case, beer flavors) when other substances are present in the sample.
For example, a heavily hopped beer will often have several non-hop flavors present, like malt flavors and fermentation-derived esters, but the intensity of the hops will likely overpower much of the other flavors, creating a context in which it is difficult for a taster to discern much else. Take for example a blonde ale vs. a dry-hopped IPA, both fermented with the same Thiolized yeast strain — the thiols present in the blonde ale, even if they are just at the threshold level, will seem more intense than in the IPA, where highly aromatic hops will meld and reduce the overall perception of thiols. The matrix effect is at work here, as the IPA has a stronger concentration of other flavors present that are not at play in the blonde ale. Brewers can use this information to write a recipe in which the various components work together to build a harmonious overall flavor profile.
From a biological perspective, we define a threshold as the minimum amount of a stimulant required to cause a nerve cell to fire and subsequently deliver some sort of information to the brain.
Do sensory thresholds matter?
Thresholds are important because they represent the bare minimum amount required for a compound to be perceived. When a flavor measures in just at threshold, that means it is at its lowest detectable level for the majority of tasters. For a flavor to be truly impactful in a beer, it will likely need to be above its minimum threshold to reach the widest audience.
Sensory thresholds are also important for brewers from a quality perspective. When running a quality panel, knowing what the threshold is for a given flavor compound can help serve as a baseline when evaluating beer, and the presence of those flavor compounds can be further validated through lab testing.
Sensory recognition in practice
While banana notes are common in a Hefeweizen, they may be out of place in other styles. Training on flavor compounds can help your team identify when banana belongs — and when it doesn’t. Image made using AI.
The sensory threshold for diacetyl, the classic buttery flavor, is very low — typically between 10 – 40 ppb — but individual tasters may have different sensitivities to diacetyl. Some people may be able to detect diacetyl at concentrations as low as 5 ppb, while others may not be able to detect it at concentrations as high as 100 ppb. Diacetyl tends to have a fairly high rate of anosmia among tasters, so it’s important to validate tasting panel members and note which tasters have the highest and lowest sensitivity to diacetyl.
In many small breweries, understanding taster perception of diacetyl can be an incredibly useful tool for controlling the cellaring process. Knowing who on the production team is the most sensitive to diacetyl (i.e., can pick it up at the lowest concentration) can help determine when it’s safe to crash the tank and prepare for packaging, since they’ll be the most likely to pick it up.
On the other hand, the sensory threshold for isoamyl acetate, a flavor compound reminiscent of banana, pear, and sometimes bubble gum, is relatively high — between 600‑1200 ppb — meaning that it takes a larger amount of isoamyl acetate to have a noticeable impact on the flavor of a beer. A classic German Hefeweizen will likely contain around 2000 – 3000 ppb of isoamyl acetate, which gives it that classic banana flavor. However, the yeast strains typically used to make Hefeweizen also tend to produce other flavor compounds like clove, which may mask the intensity of the banana flavors once combined with any malt and hop flavors present. By understanding sensory thresholds, brewers can create beers with a specific desired flavor and aroma profile.
Establishing quality thresholds
At this point, you should be ready to apply these principles to your in-house sensory panel. The staff is trained, you’ve identified your most sensitive tasters for each flavor attribute, and you’re ready to run your brands through panel testing.
Regular training and monitoring is the key to establishing quality thresholds in the brewery. Use this list as an outline for building your own quality monitoring regimen.
Let’s imagine that you have established a maximum quality threshold of 10 ppb diacetyl for your beers, and any beer that exceeds that amount is considered out of spec. During routine testing, your IPA and your weissbier have been flagged as containing diacetyl, so you decide to test those samples in a lab to determine the concentration. In the IPA, the diacetyl measures in at around 30 ppb, whereas in the weissbier it comes back with 10 ppb, and both samples were flagged by only the most sensitive of your tasters.
This is where context comes into play: what other elements are present in those beers that affected the perception of diacetyl? It’s possible that other batches of IPA were considered to be quality that had above the established quality threshold (10 ppb) but below the amount it took for the most sensitive tasters to perceive it. Perhaps it was the presence of different malts or higher levels of hop aromatics that blended the overall flavor profile together in such a way that the diacetyl became more difficult to detect.
Takeaways
Sensory thresholds are ultimately a benchmark to help better understand the point at which a given flavor can be detected. However, there are numerous factors that can influence flavor perception, so it’s important to develop your own in-house quality thresholds. Define what you’re hoping to achieve for your brewery, all of your brands, or even specific brands, and seek consistent results. What’s really important is to train your staff to recognize various compounds and understand how to spot them, in both a quality management setting and for overall flavor development.
Related Collections
