Dear all,

I wonder if there was already a discussion conclusion on the influence of (pre/post) blending and roasting on the "simplicity" of espresso blends.

Assuming different green beans result in different roasted beans in terms of density and size I assume that one not only needs different grinder settings to achieve the intended extraction behavior but also that these changes differently over time.

Now for single origin roasts that are not blended this seems not to be a problem. But isn't shot repeatability for blends influenced by the exact 7g of beans that are "selected" by the grinder per shot. If now this "sample" is taken from a inhomogeneous blend in terms of bean size and density I could imagine that the particle size distribution of the resulting portion of ground coffee is differently from shot to shot making it harder to dial-in and keep the shot extraction behavior consistent.

Is there a way to pre/post blend by size/density or adjust roast profiles to achieve a "simple blend"?

Cheers,
Marko

If you're blending post roast, my answer is quite simple - yes.

For the first HB homeroasting competition, I roasted 3 coffees; Sumatra, Brazil and Guatemala, all to similar levels but requiring different profiles to reach those levels. Once roasted, I blended in different ratios for 2 to 3 doubles' worth of beans. My submission was a 2:1:1 ratio and I tested by using 8g as the baseline. This allowed me to grind 32g and pull 3 to 4 singles of a given blend for tasting purposes.

I lifted the idea of ratios from cooking and baking, and this made the blending and final calculations much easier. Given that the submission limit was 200g, I wanted to limit the number of coffees to maximize the chance that the ratios would be preserved. My thought process was that blending 3 different beans in larger percentages is a lot easier to keep track of than blending 5 beans to smaller percentages.

Blending by density is an interesting idea, but I'm not sure of the merit; my primary concern is taste, and density doesn't really come into view in this case. If you're looking for a "simple" blend, I believe that the classic Italian blend, as referenced in HB's Aficionado's Guide to Blending and Thoughts on an Italian Blend can be achieved by the following recipe:

1) Base coffee: This is your canvas, providing strong roast and chocolate notes but not overpowering the cup. > 60% of the blend. A decent Brazilian coffee will hold up fine and provide backbone. Roast darker to bring out roast notes and background the varietal notes.

2) Accents: < 40%. Add lightly roasted African coffees for floral and acid notes, Central Americans for some sweet and complex fruit notes, or anything else that would otherwise dominate the cup if pulled as a SO. Roast lighter to preserve varietal characteristics.

I like to tinker when I blend, so I will roast the coffees individually, and blend on a shot-by-shot basis.
I also think that if you're going to bring any funky coffees, such as aged sumatra, monsooned malabar, or robusta into the mix, you should definitely try a shot of each ingredient as a single origin.

An SO shot of robusta will make your hair stand on end....but you'll be able to pull the robusta notes out of blends that you taste later.

Also note that some coffees like more rest than others; I've pulled some very nice shots of Brazil on 7 days rest combined with Monsooned Malabar on 14 days rest. Almost Guinness Stout like in the chocolate/malt/nutty character.

Thanks for your answers and comments. My question arose exactly when I roasted 3 different beans with different (hopefully appropriate) profiles for post blending. The first one was actually a robusta with exceptionally large beans which I roasted slow but dark, the second a pre-blend made out of an Indonesia and Brazil roasted darker and some small African coffee for floral an acid notes roasted lighter. Exactly as you propose. Then I tested all three as SO shots and observed that I needed to adjust my grinder for each of those SO to achieve optimal extraction and taste.

Now I wondered what the effect is of blending those 3 together on finding the optimal grind setting for the result and if such an optimum at all can exist in this case. I guess that all 3 result in different particle size distributions with the same, but even more with different, grinder settings as they vary largely in size and density. This is further complicated by the fact that the grinder picks a slightly different ratio of the components from shot to shot (less of a problem for blends from 3 components than those of 5 and more components as Sherman noted correctly).

My thinking was triggered by the discussion on the remaining problems in optimal espresso extraction which seems to focuses on grinding (see the recent talk by on wet grinding by David Walsh <http://www.jimseven.com/2011/07/13/tamp ... vid-walsh/>).

Blending by density is an interesting idea, but I'm not sure of the merit; my primary concern is taste, and density doesn't really come into view in this case.

Isn't taste be connected to grinding and especially particle distribution and as such also to bean density and its variation within a blend?

Shouldn't a roaster at least try to provide "Simple Blends" by carefully selecting greens and profiles to make the component "fit" to help a barista to improve on consistency and taste?

Cheers,
Marko

MaKoMo wrote:Then I tested all three as SO shots and observed that I needed to adjust my grinder for each of those SO to achieve optimal extraction and taste.

Absolutely yes. Within the context of a SO, it is not surprising that grinder adjustments would be necessary. To further confound the issue, grinder adjustments may be made across the usable life of the roasted coffee. There are equipment differences - flat vs. conical burr grinders, lever vs. pump, the list of considerations can be staggering from a roaster's perspective.

Isn't taste be connected to grinding and especially particle distribution and as such also to bean density and its variation within a blend?

If the batch were placed into a container and then agitated with enough force to promote movement for enough time, it stands to reason that the denser beans would migrate toward the bottom, altering the physical composition. If the batch were then loaded into a grinder, there would certainly be differences in consecutive shots. The grinder, if not adjusted, would play a part in this as well. Still, that's a lot of ifs.

I have only anecdotal evidence that commercial roasters roast to their equipment; that is, they are using a specific grinder and machine, so they tune the roast and blend to those machines. To account for different types of equipment would be a rather large undertaking. Still, there are roasters that do this. I think that one of our members here, romanleal - a commercial roaster himself and one of the guys behind Evocation Coffee - has put out a "49mm" blend tuned for lever espresso machines. During the last Chicago Get-together, we had a chance to briefly talk about it, and he mentioned that testing was done on his La Pavoni.

My approach to blending considers taste as SO to be the primary influence for component selection. Mixing fewer components is a strategy to minimize the possibility of flavor differences due to less of A, more of B that is ground for a given shot. I have been able to access identifiable basic flavors (stronger acidity from a lighter roasted accent, good body and caramel base from the darker roast) in batches of up to 300g without considering bean density thus far, and in sampling the batches I have found that the rough percentages hold up.

In practice, I haven't found density to be of any significant influence with regard to bean selection. I can taste the components, and that is an acceptable level of precision at this point. I can hit the broad side of a barn more often than not .

Based on your earlier post, have you found this to be consistent with your experiences, or is there a significant difference?

Based on your earlier post, have you found this to be consistent with your experiences, or is there a significant difference?

For now I only tasted the components and observed that the robusta (with the large beans) needs a way tighter grind than the other two that are closer together in that respect. I now wonder if the robusta component is mixed with any of the other, if there is a way to choose an optimal grind. Not tried yet.

This might be an argument for post-grind-blending

I remember discussions that voted that a certain blend from Hausbrand (Northern Italy) is a very forgiving blend and I made the same observation some years ago (before I had to consume my own roasts). There were also discussions at that time that indicated that the batches you could buy from Hausbrand in your local store were roasted up to 2 years before. There might be a connection.

Also note that some coffees like more rest than others;

Yeah, I believe Robusta is one of those that needs more resting.

Marko

Excellent thoughts on this subject. I like the idea of blending post-grind. I don't know how to accomplish it best, thought.

Also, with beans of different size and density in a blend, I have been concerned with Sherman's note "that the denser beans would migrate toward the bottom".