I have come to believe that the scorched beans are most likely to be those sitting on the metal ledge at the top of the (sides of the) perforated metal base of the bean container. You can see stationary beans at the bottom of the glass sides in the first minute or two, and it is unlikely that the beans at the bottom (on the perforations) would not move, as they are the ones most exposed to the air flow.
I am not keen on using a bean-mixing method that involves taking off the chaff collector and so I use tilting to dislodge the beans at the sides. Always roasting 113g at a time, at one minute I take a large pot holder to keep the chaff collector in place while I tilt the whole thing sideways to about 45 degrees. I do this in the N,S,E, and W directions. stopping each time once the air flow dislodges the beans at the sides.
I have not had any problem with uneven roasting with SO beans, although blended beans vary in the speed at which they become brown during the roasting process. This was particularly the case this morning when I roasted some SM Liquid Amber Blend. I certainly have not experienced individual burned beans. Possibly out of naivete, I think I have that issue in hand.
That said, I am struggling with identifying exactly when to stop a roast in order to arrive at a specific target - I drink only espresso, and so tend to aim for FC+, Vienna, or French Roast, depending on the bean, always going to SM's stated upper limit. I keep records and use a thermometer, the calibration of which is off - 4% low in boiling water. I do not think that matters as long as I am aware of it. Reproducibility is probably more important than the exact number. In my mind, I add 15 to the measured temperature in the target range.




