Fast vs Slow Roasting

I want to start a dialogue surrounding fast vrs slow roasting and the effect on body and acidity in the cup, caused by the speed of the roast. Your opinions please would be great on the two below points. I have always felt slower roasts mute perception of acidity, therefore helping embellish body. Contrary to this, I have always felt faster roasts embellish acidity perception, muting the perception of body. Any thoughts on these two points below? Are these two statements correct?

1) Sensory effects of “slow” roast
      a. Decrease body, can be perceived as thinner  

2) Sensory effects of “faster” roast
      b.Increased body, perceived as more mouth feel

In asking a few of my friends this question, the replies all made sense from a logical standpoint, however, Paul Songers reply is more from a science and experience working with panels who test this stuff, standpoint. Paul also has included references to his reply. Read Paul’s response. It's absolutely educational to say the least. The Delwiche study is online at and discusses this topic. I'm interested in having some discussion about these two points above from your personal perception standpoint:

On 6 Feb 2010 at 11:30, Paul Songer wrote:
Perception of body is not as cut and dried or as measurable as, for example, acidity (which measurably decreases depending on the type of acidity, the temperature to which it is exposed, and the time of exposure to the high temperature). First, one must have an adequate definition of the sensation. As detailed by the Illy organization, what we have been calling "body" is only one aspect of mouth feel, defined as the physical or tactile feeling of the liquid in the mouth (as opposed to a chemosensory sensation such as taste or aroma). Other tactile aspects one can experience in coffee include oiliness, creaminess/butteriness (related to perception of oils), syrupiness (related to the perception of particular carbohydrates), granularity, etc. (Navarini et al on espresso body, see reference below)

The tactile perception also appears to be affected by some chemical aspects. The presence of carbohydrates, especially mono and di-saccharides, affect perception of thickness in a Newtonian liquid (a Newtonian liquid is a liquid that flows downhill as a plumber would expect it to at normal temperatures). It is a phenomenon noted for decades in sensory that one can elicit a response of "heavier body" by adding sucrose (a di-saccharide) in less-than-taste threshold amounts (one reason why so many foods contain miniscule amounts of sugar). Astringency (the drying and tightening of mouth tissue and reduction of salivary capacity that are the result of certain chemical compounds) affect how one perceives the "wetness" within the mouth cavity, along with the "fullness" of the liquid. These phenomena (the interrelationship of acidity, sweetness, astringency, and body) are especially studied in wine and beer research. (Delwiche, Jeanine and Langstaff, Guinard, et al)

Experientially, mouth feel is kind of a liquid friction resulting from pressure. When one feels the liquid in the mouth, the experience is based upon the sensation of the liquid in contact with the mouth tissues, including when the liquid is moved around the mouth. Tongue movement pushing the liquid against the inner mouth tissues has a lot to do with the perception.

One can measure the degree of acidity according to pH and titratable acidity to reasonably predict what a taster's response to the stimulus will be, but standard means of measuring viscosity (such as specific gravity or resistance to sheer pressure) does not regularly correlate to perception of body in coffee or espresso. We are talking about a solution of less than 1.5% solids (espresso at about 5%, which is why the subject was of such interest to the Illy organization). Swishing it around the mouth involves other liquids present (saliva), resulting in further modification of tactile perception over time. At some point in the process, the cupper rates "body".

In terms of relating roasting to body, the heaviness or thickness/viscosity sensation appears to be most related to the solubility of the coffee cell wall (polysaccharides, especially arabinose and galactans) and availability of oils to dissolve during the brewing process. A reasonable conjecture is that this is related more to the physical structure of the bean resulting from roasting than the usually considered chemical changes. If so, this is directly a function of roast time. When a bean is roasted faster, it expands more and loses density more quickly than in a slow roast, allowing quicker penetration of coffee particles by water during the brewing process and more easily dissolveable cell walls and surface oils. Coffees of higher density in the green stage require more heat penetration (by whatever roasting means) to develop higher body (at the same time, they retain more acidity since the heat does not penetrate as quickly, as noted by Jim).

However, roasting time also has an effect in the sweetness to astringency continuum and this depends on the innate green bean quality. The body will be emphasized in a sweeter coffee and perceived differently in an astringent coffee. As the sugar browning deepens, sweetness is sacrificed to aromatics (through carmelization, Maillard reactions, and Strecker degradations). Roasting affects levels of astringency in terms of chlorogenic acid breakdown (darker, more astringent phenols balanced by more quinic acid, but other astringent aspects are lessened through roasting). Other considerations are how sensory attention is drawn (more acidic coffees are usually rated as having lower body) and how the brewed solution of coffee evolves over time. Bitterness can also play a role, so the degradation of chlorogenic acids (which by themselves are usually bitter and sometimes astringent) that have astringent phenols and various acids (most notably quinic, which is sour/bitter) as their degradation by products.

Oils (lipids) can also have different weights and qualities, ranging from more liquid (like regular cooking oil) to more waxy (harder fats, like butter). This can be a function of coffee age. Fresh green coffee has few free fatty acids, but these develop over time especially in the presence of higher temperature and moist conditions. (Speer, et al)

There are also present in coffees emulsifiers that enable the integration of oils, liquids, and gasses (hence crema) under conditions that are dependant to a certain extent on how the coffee is brewed (pressure is usually necessary) but is not as well studied as far as I can tell from a point of view of how these components are affected by roast. (Navarini et al on crema)

It is an interesting subject. My personal conclusions (based on research, personal experience, experience in dealing with panels, and measurement where possible)  is that (1) when coffee is roasted to a certain degree, the cell walls gradually become more soluble the darker the coffee is roasted, up until the point that carbonization (complete reduction into basic carbon organic components) occur, at which point the tactile perception becomes more of granularity rather than body. At the same time, as the coffee is roasted darker and density is reduced, oils become more available and soluble into solution at typical brewing temperatures, especially when brewed under pressure as with espresso. These are the most basic perceptions of "body" and other tactile senstions. Development of "body" from this veiwpoint is a physical phenomenon that depends on the initial state of the green bean (density, composition of the cell wall, amount and types of oil), the rate of bean expansion during the roast, and its final physical condition upon brewing (grind and so forth).

(2) The chemical aspects of astringency and carbohydrates (and sometimes bitterness) add to or subtract from this basic experience, not always in predictable ways. From a roasting perspective, these aspects can be maximized or minimized, but much depends on the initial chemical aspects of the bean. At darker roasts, chlorogenic acids degenerate into quinic and other acids and (astringent) phenols. The darker the roast, the more degradation of sucrose occurs (but at the same time, more cell wall material becomes soluble).

(3) There are also present in brewed coffee emulsifying aspects that allow gasses, water-based liquids, and oils to combine (this is especially true in espresso preparations regarding "crema") that contribute to the overall mouth feel sensation. These may have a "sweet spot" in roasting where they balance particularly well, depending on the blend of coffees and which components are actually present in the green coffees. This has been a carefully monitored aspect in much espresso research.

Finally, and most important in any reasonable discussion of the subject, (4) the perception of body must be well defined experientially by cuppers. My personal experience is that a panel may agree on relative levels of sweetness, acidity, balance, quality of aromatics, etc., but be completely at odds on the subject of body. It appears that it is not as well defined as we think.


Navarini et al, Espresso Coffee Beverage: Classification of Texture Terms, Journal of Texture Studies, vol. 35 (2004), p. 525–541.

Navarini, et al, Physicochemical characterization of the espresso coffee foam, 21st ASIC (2003), p. 320-327

Langstaff, Guinard, and Lewis, Sensory Evaluation of the Mouthfeel of Beer,

Delwiche, Jeanine, The Impact of Perceptual Interactions on Perceived Flavor, Journal of Food Quality and Preference Vol. 15 (2004), p. 137–146

Speer, et al, The Lipid Fraction of the Coffee Bean, Brazilian Journal of Plant Physiology, vol. 18 (Jan.-March 2006),


 Thanks Paul for sharing

 Thanks Paul for sharing this, and to Sherri Johns for directing me to your blog. 

Paul Songer's response is an informative, thoughtful tour-de-force - typically wonderful! I have absolutely nothing to add (and much to learn!) from a technical or scientific perspective, but a few comments from a couple of decades of roasting on all kinds of roasters and tasting the output. 

My experience with drum roasters has been consistent flattening of acidity for roasts taking longer than 14 minutes. For classic full city roasts (no second pop) the ideal seems to be around 11-12 minutes. 

I've done most of my roasting on Probats and the like ranging from 35 to 120 kilos, but started my roasting career on a half-bag Sivetz and have often had small (5 lb.) fluid bed roasters available for roasting specialty lots. In general I've found coffees roasted on fluid bed equipment to have slightly brighter acidity and lighter body than drum-roasted coffee taken to the same degree. "Back in the day" (nearly 30 years ago) when you could routinely roast a Kenya double A to 35 on the Agtron scale and still taste blackcurrant I would've said the drum roaster was the way to go, but given the general quality and acidity levels of what is available today I see a lot of advantages to fluid bed roasters, heretical as it may be to say so in some circles. 


Thanks again for the great blog!

Post new comment

  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options