I just measured five different positions - the main listening position and four other, 50 cm and 1 meter to the left and right from the main position.
Listening position vs 30 cm in front of the tweeter ribbon, 1 meter above the floor. That 5-6 kHz thing is still there.....
As far as I remember it is 1/6 octave smoothing and no gating. You need quite some distance to let the drivers integrate. You cannot measure large diaphragm drivers in the nearfield. The bass section measures different from top to bottom in the nearfield. I do not have that peak around 5 kHz. My 3.6 are not using the passive external crossover, it is replaced by an active crossover. Those who has been listening at my place are surprised about the bass delivery.
Last edited:
Even though the frequency response is less than perfect, the distortion is very low. I guess that's one of the reasons I love how they sound. At any sensible listening level, the distortion in the mids and treble of my active 3.6 and 3.7's is lower than the noise level of my measurement mic....
I know measuring distortion without an anechoic chamber may not give any reliable results, but anyway - here's a distortion measurement of my active 3.7s. Apart from something very odd happening around 230 Hz (room related perhaps), it's really low and mostly below the noise floor of the measurement mic. I've measured this at 85 dB on 3 meters (in the main listening position). Below 200 Hz... you don't want to see that..... 🙂
I am amazed that such a flat measurement was possible (apparently without too much averaging).
My experience with the MG 1.6 was that the measurement was dominated by the bass panel resonance at 46Hz with lots of ripples throughout the audio band. I thought that these speakers could never measure flat.
Ed
That Stereoplay measurement looks too good to be true ….
I will take a closer look at 3.7 and distortion tomorrow. It’s not going to be pretty …..
I will take a closer look at 3.7 and distortion tomorrow. It’s not going to be pretty …..
Standard Stereophile practice is to spatially average the upper frequencies, so the wiggles aren't as position-dependent as a single point measurement.
The image caption that goes along with the frequency response plot in post #12 above: "Fig.3 Magnepan MG3.6/R, anechoic response on-axis at 50", averaged across 30 degrees horizontal window and corrected for microphone response, with the complex sum of the nearfield woofer and midrange responses plotted below 300Hz."
But point taken that the lateral dispersion is chaotic, so measuring well is more difficult than a typical speaker.
https://www.stereophile.com/content/magnepan-magneplanar-mg36r-loudspeaker-measurements-part-2
Fig.4 Magnepan MG3.6/R, lateral response family at 50", from back to front: response 90 degrees-5 degrees off-axis on woofer side, reference response, response 5 degrees-90 degrees off-axis on tweeter side.
Here are low frequency distortion measurements of a 3.6/R, taken at 30 cm distance from the bass panel, 1 meter above the floor, in my room. The midrange fuse was removed during this measurement for safety reasons.
First, I show the frequency responses we get. It's interesting to see that the bass peak from Stereophile's measurement is there, as well at the fact that the Stereoplay measurement shown in an earlier post simply isn't trustworthy. There's virtually no output below 40 Hz.
Measurement levels:
Blue - 2 V / 1 W @ 4 ohms
Yellow - 4 V / 4 W @ 4 ohms
Green - 8 V / 16 W @ 4 ohms
Red - 16 V / 64 W @ 4 ohms
Here's all four measurements put together at the same level (using the "Separate" slider in REW). It's quite obvious that the speaker compresses the bass peak, but not so much above 50 Hz. At 16 V, the bass panel started to slap against the magnet plate when my sweep started at 15 Hz....
And now the distortion measurements. 2V first, then 4, 8 and 16 V. Red curve is 2nd harmonic, yellow 3rd harmonic.
First, I show the frequency responses we get. It's interesting to see that the bass peak from Stereophile's measurement is there, as well at the fact that the Stereoplay measurement shown in an earlier post simply isn't trustworthy. There's virtually no output below 40 Hz.
Measurement levels:
Blue - 2 V / 1 W @ 4 ohms
Yellow - 4 V / 4 W @ 4 ohms
Green - 8 V / 16 W @ 4 ohms
Red - 16 V / 64 W @ 4 ohms
Here's all four measurements put together at the same level (using the "Separate" slider in REW). It's quite obvious that the speaker compresses the bass peak, but not so much above 50 Hz. At 16 V, the bass panel started to slap against the magnet plate when my sweep started at 15 Hz....
And now the distortion measurements. 2V first, then 4, 8 and 16 V. Red curve is 2nd harmonic, yellow 3rd harmonic.
Last edited:
The frequency response at 30 cm is similar to what you get from a directional microphone, the proximity effect. At a normal listening distance the peak in the bass diminishes. The lowest tuning of the bass is at about 42 Hz after running them in. That peak will also limit the maximum output as it calls for very large excursions. There will be a drop below that peak, it is even more than 18 dB/octave due to the high Q peak. In a room, the frequency response will vary a lot, as the speakers response will be heavily influenced by the surrounding surfaces. It will modify the roll-off of the bass and can destroy the imaging. It takes quite some time to find the right spot for the speakers and the listeners. In some room these will never be found! Stereoplay made an interesting test of the Quad ESL-989. Measuring a large diaphragm dipole in different ways. The second picture is a pair of 3.6 with the passive external crossover. The peak at 100 Hz is from a loose button at the top of the bass driver, seems to be a common fault.
Attachments
Last edited:
For comparison, I dug out my measurement of the MG 1.6 (and my room!)
My graph showed similar peaks and valleys as the Stereophile measurement.
https://www.stereophile.com/content/magnepan-magneplanar-mg16qr-loudspeaker-measurements
My speakers have the bass panel resonance where one would expect (Stereophile must have measured a brand new speaker).
The graph illustrates the difficulty of measuring planars.
The peaks at 60Hz and 120Hz are hum.
Ed
My graph showed similar peaks and valleys as the Stereophile measurement.
https://www.stereophile.com/content/magnepan-magneplanar-mg16qr-loudspeaker-measurements
My speakers have the bass panel resonance where one would expect (Stereophile must have measured a brand new speaker).
The graph illustrates the difficulty of measuring planars.
The peaks at 60Hz and 120Hz are hum.
Ed
Attachments
Stereophile measured the bass panel within a very short distance, that is not representative for the speaker at a listening distance.
Yes, they may have measured a section of the panel that was tuned to 75Hz. From their measurement, one would conclude the speaker had no bass.
Ed
Ed