Help with understanding and improving relation of the roast profile to flavors

Hello dear community. I’ve been roasting for a couple of days now and I’ve been enjoying it so far. My last roast was what I call a “perfect profile” (I may be totally wrong) and you can find it here:

Here are my notes:
Roasted 11th Jan Evening, tasted 13th Jan morning.
In Latte, Kristina is loving it.
I extracted the espresso ground at 1.5, first drop at 7th second, got 32g of coffee in 21 seconds. In espresso I can taste quite high acidity, but also bitterness. Kristina says its too bitter. The roast was not too dark, took out at ±215 IBTS, color-wise it’s around Full city.

Any ideas? What I am most perplexed about is, how can it taste acidic and bitter at the same time - but not the pleasant way, what people could call “balanced” but more like annoying.

Could it be not enough time to de-gas? Does that change the flavour so much? Or do you think the development time is too short?

BTW, I know the FC came at ±208C which is a bit late, but with these beans “Honduras premium blend” it’s very consistent. I tried some other ones and they were around 200-205C, so I don’t think anything’s wrong there.

Thanks a lot for any thoughts.

If you’ve tasted an espresso on 13th, then you needn’t worry about anything else. It’s much too early. A fair comparison would be trying to make coffee with sparkling water. It would most certainly taste foul. Once they pulled a shot for me from the coffee roasted for the World championship, which had about 36 hours of rest. They were saying it was OK. It wasn’t.

1 Like

What’s your delta?

2:19min, 23%

Did you compare that with your average for good tasting batches yet?

That curve looks dope.

But yeah, more to it than just the curve.

Delta is better explained by @braca19452f9m , he is kind of the legend in that regard. He can stead you straight. I’m just learning it myself.

1 Like

I am not yet sure what’s a good tasting batch :D. This is supposed to be the best, but tasting it too early truly has it’s negative impact on the taste. I tried it today and is SOOO much better. It’s been ±2.5 days since roasting. It’s still a bit too bitter, despite the roast level (Full City). Maybe it’s the extraction time, but that is almost spot on (8 seconds first drop, 25seconds to get ±32g of coffee).

Anybody has any experience with having the development phase RoR close to 5-0? And just keep it there? The reason I ask is that my FC on this bean is around 208C, then when dropping at around 215C yields to a bit bitter taste and maybe more than Full City roast, balancing it a bit more would mean I’d have to take it out at around 213C or even less, which would mean lower development time. SO if I want to keep the development time extended, my RoR must be close to 0 after the FC.

Thank you so much. This is in fact true… I am newbie to tasting coffees, but I can taste a significant difference waiting 24 more hours after roast.

I still taste a bit of a “marzipan” which may indicate some level of rawness? Not sure what can I attribute it to. It’s not unpleasant, but I dont usually taste it when buying commercial grade coffee. Any idea what could it be?

Thanks a lot for any help!

May I make an observation about your approach?

For most roasters in the past, it has been a sensory job. Using these graphs and data weren’t a part of the general industry 20years ago.

We can now sort of gamify roasting, and get guides on “what’s my RoR?” And all of that, but if yo are roasting to stats and graphs, how does what you do going to be better than A.I. roasting?

I suggest you spend time using the try-er a lot while your roast test batches. Get a good light so you can see color changes.

YES, using the tryer a lot will likely ruin your batch, but it will also allow you to roast with your senses, which is something that machines don’t currently do.

Smelling the roast as it roasts is the best teacher. Build that intuition and that conversation.

We’re at a geek phase in roasting where it attracts us nerds that want to use data and make things perfect, but if you roast that way, in a few years, you won’t be better than a robot.

Data is hugely useful, and so is technique and guides, and for many of us that is the enjoyment, even more than tasting the coffee.

Where are you at? Are yo a data nut or a sensual coffee drinker?

I tend to be more about the coffee, and data is a means. That is why I love sharing and learning about roasting with others, but I’ve noticed on this forum so many people discussing numbers…. Which are relative. It is weird, but hard numbers are almost less specific than generalities and theory of intention.

Also, a lot of roasters are on the autism spectrum. It is a job that people on the spectrum do well. I do think it is sad that I won’t be able to hire a roasting staff in the near future, as we get undercut by machines, but at least we can help each other enjoy home roasting.

So I hope you get into your senses more, and yeah, blow some money on wasting a few batches using the tryer more. Do a slow batch and a fast batch. Take them dark and smell and take notes along the way. See how things change first hand. Then make coffee from one successful batch (where you didn’t use the tryer) and just drink that each day and see what da it tastes better on. Some people like day1, some like day3, and others like day 7 or 11. For espresso I know baristas that like day 14. By day 14, much of the “volatiles” have settled down, but there is still enough aroma and freshness and any oils have not gone stale or rancid. Generally pour over crowd likes day3-day11, and espresso people tend to like day7-day18. But new school people have all types of preferences. Many new school roasters have developed roasts and palettes for day 1 brewing. But by the sound of it, you have more traditional tastes, so a good rul for you would be to test your espresso on day7 or later. But do your cupping on day 3 to get used to detecting differences between your roasts. It is a lot easier to taste the difference between roasts in a cupping. That said, you might struggle to understand how the flavors from a cupping bowl translate to espresso, at first, but don’t worry that takes time. Main thing is to be able to detect differences and trends between roasts, and look for things you do and don’t like. If one bowl has more paper taste and feels rough on the tongue, you may flag that as less desirable, and when you pull espresso, you can reference the notes and go “oh no, yeah, nasty!” And then look for that in future cuppings and see if the trend holds up.

Roasting is a slow going. Has always been a job where 10 years is a blink. But all this new data and community is a great way to get up to speed much faster. As long as we can access the data in meaningful ways.

You asked about RoR? But did you ask what flavor or express your intention? If you are afraid of doing something the wrong way, then you know different RoR has different effects, so why just ask about RoR without expressing your desired result?

The answer is yes, 0-5 ROR IS OK, if that is what gives you the flavors.

Look, most people want to have an easy brewing coffee, but what if you prefer the flavors of a coffee that isn’t easy to brew? What if you light roast and need a longer brew of a much higher dose of coffee?

We aren’t even talking about variables of the roasting environment yet.

The insane amount of “yes ands” in this type of thing require that you have some sort of intention to give context to what you are striving for.

Traditional espresso? 3rd wave espresso? 4th wave espresso? And if traditional…. HOW traditional? Haha. Like really old school espresso is different from what we get today at most places. I suppose you can still find it in Europe, but the US, Australia, and most of the places I visited in Asia are not quite old world espresso anymore. But still decidedly more traditional than say the grassy light stuff that we can get in very specialized places.

So, I doubt I answered any question you had, but you know, give yourself a little time with this and explore it with your senses, and then look at the data and trends. Because there isn’t a one way. Just like you might ask how to cook a Steak and one person will tell you only use olive oil and salt, another will say cook it to 145F internally, and others will marinate and others will rest it, and others will souve, and and and and, and then every rule you started with gets broken. But what would be the point if you don’t know how it should taste or if you even like it?

When you finish a roast, crack open a bean and look and see how even the inside and out are. You will likely see lighter color inside and possibly a slight char near the “germ” of the bean. This will give you both bitter and sour flavor in your espresso. Sometimes it can be nice though, but that is such a fine line, it is generally not good. Uneven beans can be the best of both worlds, but usually are the worst of both. So, if you are prone to gamifying things, aim for even beans both inside and out, without char or burn marks.

Alright, time to go for a walk. Take care, happy roasting, hopefully this has some usefulness, and yes, we will share tons of numbers and rules with you in the coming months, but each, spend some time trusting your nose and take a break from the forum for a week or two.



I haven’t even finished reading your message being so impatient to reply. I love the thoughts. You got me right, I am very much invested in numbers, I love to calculate and measure anything and everything, and your thought about this approach being easily replaceable by AI shortly is spot on.

The reason why I am this way at this point, is simple: I am having hard time associating final result (what I taste) to all those variables that come into place. Some of which are: bean origin, quality and characteristics, roasting attributes - length, power, RoR, development time and only then the way you extract the coffee. I need to lock some of these things first and play with only a few variables to be able to attribute those tastes to changes I did.

The problem with that is also, that I do not yet have the ability to know what taste to look for in the coffee. Is it about personal preference? Balance between acidic and bitter taste? And even knowing that there should be a balance, I just made a cup of coffee that is acidic at first and bitter at last… does that mean there is a balance or it’s totally off, because those tastes come at different times?

I am just learning all that and trying to grasp it as many of us here I assume, seeking help from more experienced coffee drinkers/roasters.

Thanks for all the insights :slight_smile:

1 Like

This video is quite helpful too. (For other newbies out there)

The bare minimum for espresso is 3 days, although it should be at least 7. For example, my Sumatran beans taste the best long after the roast. Like 3 weeks or a month after. I roast light to medium, depending on the bean. No need replacing the original flavours with coal and rubber. It might be a robusta thing, but low quality robusta. I’ve found that Sumatran coffees have a much harsher flavour the darker the roast is. Now, back to the thing at hand.

I think you’ll find that the SCA protocol is a much better aid in that regard.
Protocols & Best Practices — Specialty Coffee Association (
Tasting coffee early should only be done in cupping. I mean, you could brew a fine Turkish with fresh beans, because the method is much similar to cupping. In the protocol, you’ll find the grind size, the correct water ratio, but I’m failing to understand why they require a medium dark roast. Do not follow that one, unless you’re into something like that, or want to do a test drive hitting all the parameters on the list.

The more you think about it and the more data you get, the more it will set you back. Go with simple things. The farmer knows what he produced. The one who sold you the coffee knows what he has bought and has surely told you. You need to get that out of the coffee, nothing more. How? The more you roast, the more you learn. For example, next time you might want to try going slower and see how that looks like. Do not be afraid because the coffee will never be undrinkable, unless you dump it before FC or burn it beyond recognition. The most important thing, don’t listen to me or anyone else. Think about it and make your own decisions.

1 Like

The reason why I am this way at this point, is simple: I am having hard time associating final result (what I taste) to all those variables that come into place. Some of which are: bean origin, quality and characteristics, roasting attributes - length, power, RoR, development time and only then the way you extract the coffee. I need to lock some of these things first and play with only a few variables to be able to attribute those tastes to changes I did.

Sounds like you have the right idea about changing as few variables as possible. And you’re right that the extraction method matters, you can get more or less acid by changing the temperature.

The problem with that is also, that I do not yet have the ability to know what taste to look for in the coffee. Is it about personal preference?

If you’re roasting for you, yeah, personal preference, why not? :slight_smile:

Roasting for others, trying to be like the pros? Got any fancy cafes in your area? That’s how I keep up with what’s good. I like drinking my own roasts, but when I go to the cafe I’ll get a pour over of some single origin bean that they also have for sale in bags.

Usually theirs is (much) better than mine, and usually I come away with an idea of what it is I like about theirs or don’t like about mine.

If you want to be methodical about tasting, check out the flavor wheel

1 Like

Wow! Thank you for that thoughtful reply! That is a specific view I hadn’t fully considered, but really informs your other questions, and helps me know better if I have any useful info how to share it. (I hope I have useful tips, haha!)

And thanks for making your journey here, you ask questions that I may not have thought of yet, so I learn as you explore too.

1 Like

Thank you all so much for pointers and thoughts and sharing your own experience. It’s super helpful. I’ll keep trying and hopefully growing my expertise.


1 Like

This might be the most important post I have read here in many months.
Thank you…

It is quite scary to write that way for me. Thank you for your kind words!