re: Data shift… take a look at the IBTS and the Bean Temp. They run fairly parallel, drawing a little closer to one another as the roast progresses (till FC when they tend to move very close together). It’s the simultaneous difference between IBTS & BT that I was referring to. The difference tends to get smaller as the IBTS gets dirty.
These 2 screenshots were taken with the mouse at about 3:00 minutes into the roast. Check the data at the upper right of each. Bean Temp is approximately the same value in each, but Drum Temp (IBTS) is about 9F° less in the lower shot.
The 1st screenshot was taken 9 roasts before the 2nd shot. In just 10 roasts the IBTS/Drum Temp indicates 9° less. If left uncleaned, the contamination keeps building till Drum Temp is less than Bean Temp at the end of the roast.
If you don’t want to deal with cleaning the IBTS lens (the clear ‘glass’ over the sensor), then you can use Bean Temp exclusively for managing the roast and ignore IBTS. But that presents a problem during Preheat. Since Preheat set point is determined by IBTS value alone, that means a roaster with a dirty IBTS as actually Preheating to a higher temp than the same roaster with a clean IBTS. You will have to guesstimate what adjustment you have to make to keep Preheat where you want it otherwise the starting temp when you charge the beans will get progressively higher and higher.
re: hearing FC & SC… I’ve got that problem too. I wear hearing aids and I’m not sure if that helps or hinders. The issue, of course, is the crashing sound of the beans obscuring the sound of the crack. For a given variety of green beans, FC is fairly repeatable. I find I can make out the crack if I watch the temp… I convince myself that the sound is a little different. I don’t try to pick out the very first cracks, but once it gets going a little I’m pretty sure I’m getting it right.
You can also pick out FC to some extent by watching RoR- the beans change from an exothermic reaction to an endothermic reaction (giving off more heat than they are absorbing) which causes the increase in RoR. I can see it clearly in hindsight (it’s very obvious with the profile I use), but I’ve never trusted myself to identify it properly without sound for confirmation.
As an alternate to direct listening, one Bullet owner here devised a ‘listening tool’ using a mechanic’s stethoscope. It’s intended to help a mechanic listen to the sound of an auto engine while it’s running. He replaced the long thin probe with a short length of metal tube and some hose that fit that tube. The metal tube was sized to fit the trier opening. He said it worked great for his purposes: just removed the trier and inserted the stethoscope adapter into the trier and listened for the crack. Do some digging with search and you should find his post with a photo.