Post by kiernan on Mar 9, 2021 16:44:58 GMT -5
Hello. Thought I would inquire about this in case it is something standard you have encountered. I have roasted a few batches of beans where I get a distinct flavour when sampling the roasted bean that I associate with glue (not that I regularly snack on glue 
). It's a flavour that goes away if I roast more. Also, despite the glue taste, the chocolate itself has come out fine. I've gotten this with two completely different beans. Is this just coincidence that these beans happened to have those flavours? Or is it standard sign of underroast, overroast, or some other roast or processing error?
On a related note, I am curious about peoples thoughts on sampling beans during the roast. I've seen it encouraged and I've read a few posts about bean roast tests. Is there any actual benefit to doing this? I understood that flavour profile can change a lot within even a few minutes at the end. Also, I thought that the flavour profile continues to develop after roast for several hours, if not days. Seems like sampling during the roast might be quite inaccurate and, in any event, probably not give quick enough feedback to be useful/practical, except perhaps to detect a significant underroast. Am I missing something?
). It's a flavour that goes away if I roast more. Also, despite the glue taste, the chocolate itself has come out fine. I've gotten this with two completely different beans. Is this just coincidence that these beans happened to have those flavours? Or is it standard sign of underroast, overroast, or some other roast or processing error?On a related note, I am curious about peoples thoughts on sampling beans during the roast. I've seen it encouraged and I've read a few posts about bean roast tests. Is there any actual benefit to doing this? I understood that flavour profile can change a lot within even a few minutes at the end. Also, I thought that the flavour profile continues to develop after roast for several hours, if not days. Seems like sampling during the roast might be quite inaccurate and, in any event, probably not give quick enough feedback to be useful/practical, except perhaps to detect a significant underroast. Am I missing something?
