Post roasting analysis by smell/fragrance?

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
Posts: 1623
Joined: 17 years ago

#1: Post by farmroast »

Right after roasting, cooling and putting the beans in a bag or jar can anything be learned by the smell/fragrance of the roast?
LMWDP #167 "with coffee we create with wine we celebrate"

Posts: 136
Joined: 16 years ago

#2: Post by Frost »

(true confessions) I am a habitual post roast sniffer Ed, so I have some experience with this. I put the roast in small mason jars after cooling (usually the 12oz jars or about 3 batches to a quart) then I crack the seal and sniff the dry aroma maybe once a day. Not to be too analytical, but gotta release the pressure and I can't drink it all day. It's a small but intense pleasure. Most of my roasts will just smell like roasting coffee (a bit rough/smokey) for the first few hours and don't develop a clear aroma for about 12 hours. From here out I think the aroma is key to a good roast, but does not tell all. I've had some heavenly aroma make a sour or astringent shot, but a flat lifeless aroma I can't recall making a remarkable espresso. Right out of the roaster I can't tell much but I have done the crunch test sometimes when I'm culling stinkers to see if I'm tossing anything good.

Ken Fox
Posts: 2447
Joined: 18 years ago

#3: Post by Ken Fox »

farmroast wrote:Right after roasting, cooling and putting the beans in a bag or jar can anything be learned by the smell/fragrance of the roast?
In the distant past, on one of Jim Schulman's earliest visits, he stuck his schnozolla into the path of the smoke exiting my sample roaster and I had the impression that he used that (the smell of the smoke) as an indication of when to terminate a roast. We both agreed shortly thereafter that this approach really didn't work with a drum roaster, especially if one had something more accurate (thermometry) to guide the roast in process. I doubt that Jim does this anymore with his air roasting as he uses a PID controller to control his roasts.

One thing that you can do that might yield some useful information is to weigh the green beans before roasting and then the finished product after roasting. If you divide the final weight over the pre-roast weight, you will then be calculating the weight loss that occurred from roasting. I roast predominantly dry processed Africans, and the green beans lounge around in my cool but dry basement before they get roasted, so the typical weight losses I get (around 15.5 to 17.5%) may not be typical of what you would get with the beans that you roast, and also are effected by the final roasting temperature (e.g. roast level).

Nonetheless, I find that by weighing each batch, or nearly every batch, that I have a point of reference that allows me to check my roast profiles and my thermometry. Each of the beans I roast comes out differently, even if I roast them all to about the same final temperature which is several degrees before 2nd crack would begin. A given bean is typically within 0.5 percent of its "usual" weight loss each time, usually closer than that. This is to say that there is a consistency within a particular bean taken to a typical temperature that, were it not repeated on a given batch, would give me reason to consider whether I'd screwed up in some fashion. Most of the time when I think that I have deviated too much from my desired profile, it is reassuring to find that the weight loss after roasting remains typical for what I usually get with that bean taken to that final temperature, and later tasting of the finished product confirms that the roast turned out well.

By all means, smell. But I'd invest a few bucks in an accurate gram scale and I think you will get more from that then you will get from sniffing the results just after roasting.

What, me worry?

Alfred E. Neuman, 1955

User avatar
Team HB
Posts: 13820
Joined: 19 years ago

#4: Post by another_jim »

I think the aromas right after roasting can be deceptive.

As Ken says, I used to end roasts by smell. However, this requires both very small batch sizes and a spot where one can pick up the aromas without the soot and burnt oil particles. Since I started the coffeecuppers gig, I've found the best predictor of cup quality is the aroma during the first crack, just before the point where oils and smoke mask the aromas. You will smell funk and ferment quite clearly at this point. You will also get an indication of the high end, acidic aromas. However, there will be no clue about the quality of the caramel and roast flavors until you actually drink (I sometimes miss the little fresh roast where I could smell those clearly and nail them perfectly at the end of the roast).

This means that the smell at first crack is a window on the bean quality, but no indicator on how to profile the roast. However, if you compare the impression you got of the coffee at the first crack with that you get drinking it, you'll know whether you nailed the profile or whether it needs work.

This rule doesn't work the same way for Harars and Yemens. I have an Ismaili from last year that smells dreadful at the first crack; but makes an excellent mocha style shot roasted to a rolling second. If you do get a natural that stinks at the first crack, you might still get a very good cup when you take it to a rolling second and roast out the ferment.
Jim Schulman