Ajax,
Inevitably this is the problem with posting graphs and why I really do not like doing it. Understanding how a speaker will perform and establishing its specifications requires interpreting a family of curves (amongst many other measurements and tests). In the case of the graphs in question here there are two issues that require further information than these graphs provide and one problem with this posters interpretation of the decibel spread. These are the standard on axis amplitude response graphs measured in the anechoic chamber at the NRC in Ottawa and represent what would arguably be the most important and most common measurements of a speaker’s performance. Below are the three issues:
1) Since the anechoic chamber at the NRC is rated to be accurate to 85 Hz we also look at measurements done in free space to figure out the true response of the speaker below this frequency. There is a correction curve written for the chamber at the NRC below 85 Hz and it does work fairly well (it is incorporated into these graphs) but not to the point of being able to rely on it within a few db.
2) Multiple microphone positions within a narrow range, which we refer to as the listening window, will also create small changes in the response including some fairly high Q changes above 14 kHz. These sorts of changes need to be interpolated to eliminate anomalies that are not part of the actual listening experience and to decide on the specifications.
3) +/- 3 db means 6 db from the highest point on the curve to the lowest point on the curve. This would mean that since the highest point on both the M22 and M60 curves is 90 db the lowest allowable point would be 84 db. This would be 70 Hz on the M22 and 40 Hz on the M60.
In the case of the M80 I would come up with 55 Hz also within a spread of 86 db to 92 db. And here we have an excellent example of why the family of curves is so important when figuring out how a speaker will actually perform and establishing its specifications. As I am sure many on this board have experienced first hand, the M80 has more and deeper bass than the M60 (a quick listen test will show that clearly), but if we only look at these graphs then our specifications would have the M60 going down to 40 Hz and the M80 only going down to 55 Hz. Specifications that most closely represent the performance of the speaker cannot be built on a single graph.
