Quote:
I would like to ask, can you describe how you measure the individual driver frequency response of speakers?

You can find the 3 articles I wrote on how and why I measure loudspeakers at http://www.stereophile.com/features/99/, http://www.stereophile.com/features/100/ and http://www.stereophile.com/features/103/. You should find the answer to your question, but please post again if you need clarification.

John Atkinson
Editor, Stereophile

Quote:
I'll hijack this thread and ask for a description/explanation how to read all these measurements.

Would you please be more specific about what it is that you don't understand? If you could give specific questions, I think it'll allow John to give you a good, clear answer.

I read Stereophile for years before understanding the technical measurements of components; so I'm sure you are correct in your assumption that many, if not most, do not fully understand the measurements and how they might impact system matching. I still don't understand many of the measurements, but Robert Harley's book on audio helped me get a better handle on some of the more significant measurements.

On many reviews, JA will mention how a particular measurement is bothersome or that may require "careful system matching." In those instances, an extra paragraph explaining this abnormality would be very useful for the non techies and novice audiophiles.

However, there is much to be learned by reading the review and the measurements...even if it seems a bit overwhelming. Eventually, even a thick skull like mine can be penetrated given enough repetition.

Indeed, you are not alone in this regard.

I try to follow along with the graphs, but, must admit I'm often only left with an inkling of what might be going on... what I would find very useful is a small key somewhere in the magazine consisting of what an ideal graph of each particular "Figure" (i.e., Waterfall plot) would be so that I would have a baseline or context to put the current review info in. Maybe an ideal set of graphs for speakers and amps could be made available as a download so that readers could print out a little chart to compare the measurements of products in each issue?

When a friend in college gave me a copy of Stereophile, he said, "I don't get the math, but at least they listen to the [censored] before they measure it."

I too heartily agree with the need for a guide on reading measurements. While I'm not an engineer, I typically read the measurements before I read the review and try to understand the relationship between the measurements and the listening impressions. Thus JA's comments in that regard are very helpful. But I'd like ask for more. Here are some suggestions, from least to most ambitious:

1) Annotate individual graphs with markers (e.g., "(1)") so that when JA refers to specifics in the figures such as details about frequency response, dispersion, impedance, phase, etc., its much easier to pick out exactly what is being referenced. This would be particularly helpful in the more complex 3D plots, and the impulse response plots. I seem to recall such markers were used in the measurements of the Apple Ipod, and other DACs.

2) A bit more explanation of why some performance effects are expected based on measurements. For example, there are frequent discussions of how difficult a speaker will be to drive based on combination of impedance, phase angle, and frequency. Why is the Penaudio Senranade's "combination of 5.9 Ohms and -40degs electrical phase angle at 55Hz" "awkward"? (See, also the discussion of the Genesis 5.2). Similarly, there are frequent references to how changes in phase or impedance will "tilt" up the frequency response. A bit more 'splaining would help.

3) Add a little color (red) to the graphs to shade areas of concern such as excessive bumps or dips in frequency response, resonances, etc. Thus, even if these problem are not discussed explicitly in the review, readers learn to "see" problems in measurements. This would be great training to folks like me who may never fully get the underlying electrical theory.

4) A true Dummies Guide to reading the measurements. In particular, it would be great to have examples of "bad", "typical" and "excellent" examples of the different measurements. I realize that this could be a bit controversial, but I bet many would accept that its JA's considered view, and necessarily gospel.
Thanks,

AMEN!

Great to see that I'm not the only one not quite following the graphs here (was my post that hi-jacked this thread but it turned out anonymous up there).

> when JA refers to specifics in the figures such as details about frequency response, dispersion, impedance, phase, etc. <

One measurement I never see - which is arguably as important as frequency response - is distortion, both THD and IM. Especially IM, because that's the most damaging musically. It would be great if John would include distortion because it's much more important than impedance or phase.

--Ethan