My question is ... am I missing something? Are these really the slim margins people are working on? Do we just suck at tasting?
To me, the profiles between the good and the baked batch are so similar, I can't imagine how hard it must be to roast perfectly.
Is "baked" flavor an additional "baked" taste, or is it a dulling of other tastes (floral, fruit, etc.)?

-----------------------------------------

Ok. So, I thought that I was doing ok with my roasting progress and heard about the Rao defect kit. https://www.scottrao.com/blog/2018/6/3/ ... fect-kit-2

The basic idea is that professional roasters (deliberately or accidentally ) roast 3 batches of the same bean. One roast is an ideal development profile. Another roast is an underdeveloped roast. The final is a baked roast. The hope was to be able to identify these roast defects in my own roasts, so that I can learn to avoid them.

I emailed Regalia and got on their waitlist for the next batch to be available. After a couple of weeks, I got an email saying they were available. They also provided the temperature/time roast profiles.

I received the 3 100g samples of beans, a Colombia Huila washed caturra from the Familia Rodriguez. The roast profiles are here:


My friend and I cupped all three 10g/180mL 200F water. 4 minutes break the crust. Slurp periodically as they cool.
We also made aeropress cups of each. 20g/250mL water, 1.5minutes.
All cups were tasted blindly.

The underdeveloped coffee was obvious. You can see in the profile that it is way different.
The baked and the good version, though, were very very similar. Honestly, we couldn't identify which was which. The differences were so subtle, that it was basically a 50/50 shot.

My question is ... am I missing something? Are these really the slim margins people are working on? Do we just suck at tasting?
To me, the profiles between the good and the baked batch are so similar, I can't imagine how hard it must be to roast perfectly.

Instead of encouraging. This lesson was quite depressing. I know 100% that my heatgun/breadmachine could NEVER work on these margins. I've roasted what I thought was some pretty nice coffee, though.
I'm getting a Huky in a couple weeks, so I hope that improves my situation. And ultimately, if I can't taste the defect, than it doesn't really apply to me.

*2020-2-13 Update* I went to Paolo's shop and he spent a lot of time helping me learn about tasting these defects.
Jump to post Rao Defect Kit impressions and frustration

All I can add is thanks for the post. I'll try and emulate them.

Rickpatbrown wrote:My question is ... am I missing something? Are these really the slim margins people are working on? Do we just suck at tasting?
To me, the profiles between the good and the baked batch are so similar, I can't imagine how hard it must be to roast perfectly.
Is "baked" flavor an additional "baked" taste, or is it a dulling of other tastes (floral, fruit, etc.)?

-----------------------------------------

Ok. So, I thought that I was doing ok with my roasting progress and heard about the Rao defect kit. https://www.scottrao.com/blog/2018/6/3/ ... fect-kit-2

The basic idea is that professional roasters (deliberately or accidentally ) roast 3 batches of the same bean. One roast is an ideal development profile. Another roast is an underdeveloped roast. The final is a baked roast. The hope was to be able to identify these roast defects in my own roasts, so that I can learn to avoid them.

I emailed Regalia and got on their waitlist for the next batch to be available. After a couple of weeks, I got an email saying they were available. They also provided the temperature/time roast profiles.

I received the 3 100g samples of beans, a Colombia Huila washed caturra from the Familia Rodriguez. The roast profiles are here:
image

My friend and I cupped all three 10g/180mL 200F water. 4 minutes break the crust. Slurp periodically as they cool.
We also made aeropress cups of each. 20g/250mL water, 1.5minutes.
All cups were tasted blindly.

The underdeveloped coffee was obvious. You can see in the profile that it is way different.
The baked and the good version, though, were very very similar. Honestly, we couldn't identify which was which. The differences were so subtle, that it was basically a 50/50 shot.

My question is ... am I missing something? Are these really the slim margins people are working on? Do we just suck at tasting?
To me, the profiles between the good and the baked batch are so similar, I can't imagine how hard it must be to roast perfectly.

Instead of encouraging. This lesson was quite depressing. I know 100% that my heatgun/breadmachine could NEVER work on these margins. I've roasted what I thought was some pretty nice coffee, though.
I'm getting a Huky in a couple weeks, so I hope that improves my situation. And ultimately, if I can't taste the defect, than it doesn't really apply to me.

The example graph they provided online shows a much more significant deviation in graphs between good and baked imo. It seems possible they failed to make it as "baked" as they were supposed to. Sounds like a really hard experiment to ship...

From what I remember from tasting his defects, he tries to roast so the defect is the main difference.

IIRC, the differences between baked and regular were not as pronounced as underdevelopment, as you say.

How old were your roasts? In my limited experience with my roasts, the baked taste note has become less obvious as the roast ages.

The file that they sent for the roast profile indicates December 27th as the roast date.

So "baked" flavors are actually a flavor ... not a lacking of flavor?
Does it taste like the browning of bread?

Can you usually associate this defect with a specific feature on the roast profile curve?

I would consider neither of those "good" roasts to be good.

Yes, it is really hard to "roast perfectly". When you do, the result is unmistakeable. I'm drinking a pre-roast blend of a very pedestrian Brazil/Colombia, roasted to the brink of 2C, 415*F, 10:30 total roast time. It's a 10oz cup that you would swear had a teaspoon of sugar added. If it was roasted anywhere near those two "better" profiles, the sweetness would be gone.

Almico wrote:I would consider neither of those "good" roasts to be good.
If it was roasted anywhere near those two "better" profiles, the sweetness would be gone.

Wow. What is so wrong with the two better curves?

If it was roasted anywhere near those two "better" profiles, the sweetness would be gone.

Alan, what specifically would be different on your curves vs these? Can you post yours.

Hello:

Almico wrote: ... very pedestrian Brazil/Colombia, roasted to the brink of 2C, 415*F, 10:30 total roast time.
... would swear had a teaspoon of sugar added.

I'm currently roasting what would probably fit the same description of your pre-roast blend and rather struggling to get the roast just right.

Would you be so kind so as to post the curve so I can compare it to mine?
Of course, I understand the limitations of such a comparison.

Thanks in advance,

CIV

Almico wrote:I would consider neither of those "good" roasts to be good.

Yes, it is really hard to "roast perfectly". When you do, the result is unmistakeable. I'm drinking a pre-roast blend of a very pedestrian Brazil/Colombia, roasted to the brink of 2C, 415*F, 10:30 total roast time. It's a 10oz cup that you would swear had a teaspoon of sugar added. If it was roasted anywhere near those two "better" profiles, the sweetness would be gone.

I agree - the "good" roast is usually quite a bit better to show more divergence.

Baked tends to present as a slight lack of sweetness or lack of structure but can also kick up some cardboard or ash notes when you add water. This is all depending on your quality of extraction but wait for it to cool (like maybe ~16-20 min after initial pour) and you'll taste it pretty obviously.