ESL measurements

Getting ready to part with my ESL speakers. Others might be curious about the performance of my panels made around 1977 from Dayton-Wright cells.

These are FR (1/6 smoothing) and distortion (dB below fundamental). No EQ. Mic at listener's position, 10-12 feet. Rather small room with reflections. Middle loudness. Path includes all kinds of vintage boxes and DSP (set flat). Lots of noise around my house when measurements were taken so some false peaks are present. "-60dB" is 1% distortion, which is not bad for this kind of measurement with speakers.
 

Attachments

  • Speaker no cloth.jpeg
    Speaker no cloth.jpeg
    169.7 KB · Views: 114
  • L and R panels FR full range with no EQ.jpg
    L and R panels FR full range with no EQ.jpg
    62.2 KB · Views: 110
  • L and R panels distortion full range with no EQ.jpg
    L and R panels distortion full range with no EQ.jpg
    68.9 KB · Views: 117
Last edited:
  • Like
Reactions: 1 user
And here are traces for the system as I normally listen to it: with sub-woofer, ribbon tweeter, and a bit of EQ.

Just for fun, here's both channels together. (With large curved panels, the interactions of the channels aren't as dramatic except in low bass.)
 

Attachments

  • Normal system FR with woofer and tweeter and EQ.jpg
    Normal system FR with woofer and tweeter and EQ.jpg
    66.7 KB · Views: 46
  • Normal system distortion with woofer and tweeter and EQ.jpg
    Normal system distortion with woofer and tweeter and EQ.jpg
    63.8 KB · Views: 45
  • Normal system FR both channels with woofer and tweeter and EQ.jpg
    Normal system FR both channels with woofer and tweeter and EQ.jpg
    54.9 KB · Views: 47
Last edited:
Distortion measurements can be tricky (at least when done by amateurs). I think it is fair to say there is no way to cheat to make them look better (for a given signal level) than they are, but there are ways of making them look worse. Having said that, below are the traces for the units same as above, except the REW signal went directly into the amp (from the headphone output of my lap, believe it or not) with my fabulous Sony hall-acoustic processor, DSP, and all kinds more stuff and also, BTW, no construction noises outside my house. Of course, still in the measurement chain (and contributing distortion at each step are the mic, the mic mixer panel with phantom voltage supply, ADC into the laptop.....

Not sure (because the set-up was quite different), but I think the improvement may be because the output was a bit louder (but better, eh). Since the curves are so close, the ESL panels apparently have identical distortion at different loudness levels.

The image shows the L and R full range and about 15 dB above them a copy of a trace previously posted. Now that's really better distortion, eh. Actually, in the range under 300, quite impressive, if I may boast.

If anybody is curious, also attached below are the FR curves for this distortion run. The output levels shown in REW mean nothing absolute since just changing the setting on the mic mixer changes the "speaker output" that REW plots.

For the compulsive curious people out there, my 1977 panels have 6 DW cells each and play, wonderfully, into free air. DW speakers have 8 cells each and play into heavy welding gas inside a sealed bag. DW speakers are quite nice full range, going down to around 50 Hz (box resonance around 55 Hz) but maybe need a bit of tweeter hacking. No kidding.
 

Attachments

  • L and R direct better distortion.jpg
    L and R direct better distortion.jpg
    73 KB · Views: 40
  • L and R direct FR.jpg
    L and R direct FR.jpg
    64.3 KB · Views: 37
Last edited: