malachi wrote:As noted previously, a good argument for very large data sets.

Actually it's a good argument for always using the same equipment to make your measurements. Precision and accuracy are two different things. If you use the same readout device and the same thermometer (and thermocouples are pretty stable over time in the temperature range we're talking about here), your repeatability should be way better than the uncertainty limits that John mentioned. What's relevant, when we're talking about temperature measurement, is correlation between the measured temperature and taste. That's the important thing. If your measurement gear is stable over time, well taken care of, and you use the same gear all of the time, the relationship between measured temperature and what tastes best to you should stay pretty consistent.

In other uses, short term precision trumps accuracy. for example, comparisons of relative performance between two or more machines only requires short term stability, since the measurements can be performed over a couple of day period at most. And we're often concerned with temperature constancy over a variety of duty cycles, rather than what is the absolute best number.

-Greg

Good to see the discussion of metrology getting down to the real bits of it. Now ask yourself what is temperature anyway? Excedrin anyone?

gscace wrote:What's relevant, when we're talking about temperature measurement, is correlation between the measured temperature and taste. That's the important thing. If your measurement gear is stable over time, well taken care of, and you use the same gear all of the time, the relationship between measured temperature and what tastes best to you should stay pretty consistent.

perhaps we should start referring to "indicated temperature" and "true temperature" (or "absolute temperature") instead of just "temperature", in much the same way aircraft airspeed is noted as "indicated airspeed" and "true airspeed".

"degrees F-sub-i"

eg. 203Fi or 202Ft (or 202Fa)

The analogy to Indicated vs. True airspeed is, unfortunately not legitimate. The "errors" that separate the two have real causes, i.e. the density of the air.

Temperature on the other hand is absolute. It is what it is. EVERY reference to temperature we've ever seen here is an "indicated" or measured temperature. How close it is to reality is a function of the equipment and the procedures used. As some have pointed out, an ice bath calibrated thermocouple set up will get you very close to what the real temperature you are trying to measure is.

But the bottom line, in my opinion, is that we use temperature measurements here to help determine what makes good tasting coffee, and to troubleshoot equipment when it's not doing it. You could have a thermometer that reads to a hundredth of a degree, but that you KNOW is ALWAYS 5 degrees low and still use it for those purposes.

My guess is that 99.5% of the baristas in Italy don't know what a thermocouple is.

rfc wrote:The analogy to Indicated vs. True airspeed is, unfortunately not legitimate. The "errors" that separate the two have real causes, i.e. the density of the air.

Temperature on the other hand is absolute. It is what it is. EVERY reference to temperature we've ever seen here is an "indicated" or measured temperature. How close it is to reality is a function of the equipment and the procedures used. As some have pointed out, an ice bath calibrated thermocouple set up will get you very close to what the real temperature you are trying to measure is.

But the bottom line, in my opinion, is that we use temperature measurements here to help determine what makes good tasting coffee, and to troubleshoot equipment when it's not doing it. You could have a thermometer that reads to a hundredth of a degree, but that you KNOW is ALWAYS 5 degrees low and still use it for those purposes.

My guess is that 99.5% of the baristas in Italy don't know what a thermocouple is. :lol:

Actually, "close" is the whole issue here. In the context of this discussion, there is no such thing as "close enough."

There is, as noted earlier, really no practical (and affordable) equipment to allow us to truly know that the temp is always exactly 5 degrees low (or the like).

The fact that baristas in Italy don't know something or use something or do something has been overused (and, in fact, abused) - often to justify doing things that (honestly) result in inferior coffee or not doing things that would result in superior coffee.

rfc wrote:The analogy to Indicated vs. True airspeed is, unfortunately not legitimate. The "errors" that separate the two have real causes, i.e. the density of the air.

Temperature on the other hand is absolute. It is what it is.

the errors which separate indicated temp from true temp are real, too. positional errors, compensation errors, and other instrument errors. if we hold those errors as constant as we can, then we can ignore them and work with indicated temp, as we can usually work with IAS while flying, or we can account for them and work with true temp, as we sometimes need to work with TAS. if i'm working on one machine, or just for myself, then indicated temp is fine. if i want to spread the info around, then true temp is needed.


--barry "let's not forget errors induced by pilot tube angle of attack"