JA is very thorough and gets lots of stuff correct, but he still mesures his speakers in his little listening room. I am much more interested in his listening comments than his measurements.

Which, BTW, are fine for frequencies above say, 250 Hz. Graphing response below that is likely to be, let's call it, unreliable. In this regard, I believe only Doug Schneider at Soundstage! provides true test measurements as he uses the NRC facility's anechoic chamber to measure speakers. I don't believe any of the other review magazines have this level of measurement sophistication at their disposal. And even the NRC facility has to interpolate bass repsonse because the chamber is not big enough to support frequencies below about 80 Hz. So if you ignore low bass frequency response in the measurements, most of the mags do a pretty good job.

So how to measure bass then? IMO, it's easy, just correct for the room and measure the in-room response to get the actual extension of the speaker system in the room. This won't tell you anything about max output levels, distortion, dynamics or musicality of course, same as basic frequency response curves don't ... you (or the reviewer) still need to LISTEN.

Hope this was insightful, or at least fun.

Moderator wrote: JA is very thorough and gets lots of stuff correct, but he still mesures his speakers in his little listening room. I am much more interested in his listening comments than his measurements.

Which, BTW, are fine for frequencies above say, 250 Hz. Graphing response below that is likely to be, let's call it, unreliable. In this regard, I believe only Doug Schneider at Soundstage! provides true test measurements as he uses the NRC facility's anechoic chamber to measure speakers. I don't believe any of the other review magazines have this level of measurement sophistication at their disposal. And even the NRC facility has to interpolate bass repsonse because the chamber is not big enough to support frequencies below about 80 Hz. So if you ignore low bass frequency response in the measurements, most of the mags do a pretty good job.

So how to measure bass then? IMO, it's easy, just correct for the room and measure the in-room response to get the actual extension of the speaker system in the room. This won't tell you anything about max output levels, distortion, dynamics or musicality of course, same as basic frequency response curves don't ... you (or the reviewer) still need to LISTEN.

Hope this was insightful, or at least fun.

Both insightful and fun. I agree that the low frequency measurements are the hardest to do. But I do like the step response measurement that JA does. I think that is where the T1s and (I'm sure) TRefs excel. I think time alignment is very important.

I can't find any measurements at Soundstage! Maybe I'm in the wrong place. Do you have a link?

Thanks,

Charlie

charliehatch wrote: I can't find any measurements at Soundstage! Maybe I'm in the wrong place. Do you have a link?

www.soundstagenetwork.com/index.php?opti...ference-loudspeakers

Moderator wrote:

charliehatch wrote: I can't find any measurements at Soundstage! Maybe I'm in the wrong place. Do you have a link?

www.soundstagenetwork.com/index.php?opti...ference-loudspeakers

And @Moderator, would you be so kind as to offer a layman's term "appraisal" of the graphs as presented in the article. Basic explanation will suffice. Thanks.

GDHAL wrote:

Moderator wrote:

charliehatch wrote: I can't find any measurements at Soundstage! Maybe I'm in the wrong place. Do you have a link?

www.soundstagenetwork.com/index.php?opti...ference-loudspeakers

And @Moderator, would you be so kind as to offer a layman's term "appraisal" of the graphs as presented in the article. Basic explanation will suffice. Thanks.

GDHAL,

Since I'm the one who asked for it, I'll take some of the load off our Moderator and give my explanation. Anybody can feel free to chime in and expand or correct.

The first two plots show the response amplitude over frequency as you change from on-axis to 15, 30, 45, 60, and 75 degrees off axis. That's to the side, I'm pretty sure. It is normal for treble to fall off as you move around to the side of the speaker. You can hear this if you listen to some music and move around to the side.

The third plot is an average of five responses measured at on-axis, +/- 15 degrees to the side and +/-15 degrees above and below. That smooths things out and gives you an idea of what you'd hear in that listening "window". Keeping in mind the limitations of trying do do these measurements, this is really great flat response with a little hump in the treble above 10 kHz. The T1 has this hump also, if I recall from the Stereophile measurements. You can find the equivalent measurements of the T1 on this website if you drop down the Measurements menu.

Next two plots have total harmonic distortion plus noise. I think they use an average response (not exactly the same as the plots above), then, for each test frequency, add up the total harmonic distortion amplitudes and plot that below the main response line. The highest is around 50 Hz and is far below the main signal. Upper plot was done at 90 dB, lower plot at 95 dB. The 95 dB plot has more distortion, as you would expect at higher power, but it remains at very low level relative to the signal.

The next two plots show the deviation from linearity. Ideally, speakers (and amps) are linear devices in that, if you double the input signal level, the output will go up by exactly a factor of 2. If you triple the input, output goes up by a factor of 3. This is not an ideal world, so there is some non-linearity. I think they made the upper plot by inputting two signals, one that produced a 70 dB output, and another that produced a 90 dB output. Then they use a computer to reduce the 90 dB to a 70 dB equivalent and subtract the two signals. If the system is absolutely linear, the result will be zero. You can see there's a small difference, especially above 3 kHz or so, but even so, this deviation is small, about 3 dB at worst. For the most part it's less than 1 dB off. The lower plot of the pair does the same exercise with 95 dB and 70 dB signals, and this plot has a little more nonlinearity. But both of these constitute excellent, clean performance in my mind.

The last two plots show the impedance amplitude and phase over frequency. Stereophile plots these two things on the same plot when they do their measurements. It looks like the minimum impedance from these plots is a little over 3 ohms. I have B&W 683 S2 speakers that have a much wider range of 3 ohm behavior, even though they are also nominally 8 ohm speakers. So this kind of variation is pretty normal. I am still trying to understand what impedance amplitude and phase mean in terms of amplifier load, not just for TRefs, but for speakers in general.

You can compare this set of measurements with the GE T1 by using the Measurements menu.

Hope this helps.

Charlie

Very helpful and appreciated Charlie. Thank you.