I am a relative newbie and am trying to learn, how to adjust the various variables by taste. Jim Schulmans Espresso 101 and the diagram on adjusting dose and grind setting by taste have been a big help to me.



I am wondering though, how the degree of roasting affects this diagram. I am using a dark roast on the Strega and am having problems with bitterness. According to my understanding of this diagram I was grinding finer, thinking higher solubles yield would give me more sweetness to balance the bitter/sour. This didn't work out and lately I was given the advice to grind coarser, which has improved things (I am still working on the temperature and will maybe get a PID).

Reading Jim Schulmans "Some aspects of espresso extraction" written in 2007 got me thinking. Jim talked about two bitter components, the light maillards and the dry distillates. Maybe my problem are the dry distillates dominant in dark roasts. These very slowly dissolving flavors can seemingly be reduced in the shot by lowering the solubles yield. For a given setup (espresso-machine, grinder, basket ...) this means coarsening the grind and adjusting the dose accordingly. My cup would then still contain the fruit acids, maillard compounds and caramels.

So maybe in dark roast coarsening the grind and reducing the solubles yield gets you sweeter shots, because the bitter (dry distillates) is reduced and in lighter roasts tightening the grind and increasing the solubles yield gets you the sweeter shot by increasing the caramels. Bear in mind, that I am a newbie theorizing here, so if I'm way off track or am misunderstanding things please help me.

You generally want high extraction and yield, but if the beans are over roasted you can't do that without getting roast flavour in the cup. So to compensate you tend to underextract. I.e. Grind coarser, stop the shot sooner, use lower brew temps, etc.

Some thoughts from Matt Perger on the topic. http://www.baristahustle.com/the-downwa ... -of-death/

Thanks for the link. Interesting read. I think generally, we are trying to find the right balance. If I miss this balance and make the wrong conclusions I go off course and can get into a "downward spiral of death". I'm not a roaster, so I can't take influence on that part of the process. I am trying to influence the taste-profile of a given bean or blend to get the maximum out of them according to my taste. The solubles yield is one parameter with (as far as I gather) a big influence on taste. I'm not sure I should always aim at a maximum solubles yield, though it should keep in a certain rage. If I'm dialing in a dark roast from a reputable roaster and it tastes bitter, my first assumption is, that I'm using the wrong settings for this bean/blend. So what variables do I change? In this case grinding somewhat coarser seems to improve the taste, by reducing bitterness. But this wouldn't be the conclusion I would come to studying the diagram above, hence my speculation, things could be different for dark roasts.

Another factor to take into consideration is the increased amount of fines in grinds of dark roasts, as they are said to be more brittle (unless they are oily). These fines would also make it necessary to grind coarser to keep up the flow rate and would probably be a cause of increased bitterness.

Are there any theories why the flow-rate influences the bitter/sour balance?

I created a revised version of this figure b/c I found it easier to think about starting with a defective shot rather than a balanced shot. You can find it in the same thread as the original. The principles are unchanged. If your shot is too bitter, you have an over-extraction problem which you fix by lowering extraction by grinding finer (harder to extract stuff from bigger particles) or dosing higher (harder to extract stuff from more particles). If I'm reading your original post correctly, you attempted to fix an over-extraction problem (excessive bitterness) by grinding finer. This will tend to lead to more extraction (smaller particles are easier to suck stuff out of) so that should have made your taste problem worse. So that's my practical response.

On a theoretical note, I approach extraction as the same process no matter if the roast is dark or light. Dark/light roasts tend to contain different flavors, however, and we typically want to extract fewer of a dark roast's bitter and more of the bean flavors from a light roast. But I think it's worth considering if chemical or physical differences in types of roast do impact the extraction process in some fundamental way. I don't know enough to have an opinion, and I suspect that from the view of practice it probably makes little discernible difference. But it's an interesting question.