I´m working with olfactometry and came across some strange odour threshold results.

All four members in a dynamic olfactometer panel smelled a sample consistently with high and low concentrations but not the concentrations in the mid range. Concentrations ranging from approx. 15 000 to 4 000 x dilution were detected as well as 500 to 250 x dilutions, but two steps in the middle (dilution factors approx. 2 000 and 1 000) were undetected throughout three panel sequences with same four panel members. The noses of trained panel were successfully calibrated with n-butanol shortly before the panel in question. Ascending concentration sequence and yes/no method was used. Enclosed is an example of a panel sequence result.

The sample contained for example bitumen and other (unknown) compounds related to asphalt manufacturing. However, samples from the same (and other corresponding) process containing similar compounds, or actually any other samples I can remember, have not behaved this way ever before, that being said with adequate control database.

I am aware that there are such compounds that behave this way, for example methylmerkaptan and thioacetone. My question is what compounds are "unlinearly" odorous in a described manner; odorous only in certain concentrations, but cannot be detected with human nose in certain others?