Measuring Some Vintage Speakers

[bconline]

Here's a comparison of the Sonus Faber Concertino with the larger and more modern Focal Chorus 806V.

 

[bconline]

I also had some time to improve my measuring technique:

• Measured outside, on concrete driveway. Unfortunately, it was breezy. So the noise floor was high.
• Calibrated the soundcard.
• Soundcard direct to power amp. No preamp.
• 1 meter distance, 1 meter height, tweeter axis.
• Solid sand-filled stand.
• Lower level (~75db) full range
• Higher level (~85db) starting at 50hz, to avoid overdriving the small woofer
• 25 millisecond gate
• 1/12 octave smoothing

 

[bconline]

Horizontal polar response at 0, 10, 20, 30, and 40 degrees. Graphed with 6 millisecond gate. Shown from 200hz up.

 

[TheBatsEar]

Very cool lines, my dude.
Also very cool Sonus Fabers.

 

[dagfinn]

I also had some time to improve my measuring technique:

• Measured outside, on concrete driveway. Unfortunately, it was breezy. So the noise floor was high.
• Calibrated the soundcard.
• Soundcard direct to power amp. No preamp.
• 1 meter distance, 1 meter height, tweeter axis.
• Solid sand-filled stand.
• Lower level (~75db) full range
• Higher level (~85db) starting at 50hz, to avoid overdriving the small woofer
• 25 millisecond gate
• 1/12 octave smoothing

View attachment 366580 View attachment 366579
View attachment 366578

Looks like you're making progress . Did you remeasure any of the initial speakers to compare vs indoors results?

 

[bconline]

Looks like you're making progress . Did you remeasure any of the initial speakers to compare vs indoors results?

Yes, I remeasured the Focal 806V outside at the same level that I measured it inside. Frankly, I was disappointed that the outside measurement seems just as lumpy as the inside one. Here's inside (purple) v.s. outside (green) with no gating:

Here they are with a 25 millisecond gate:

Here they are with a 6 millisecond gate:

 

[DanTheMan]

Raising them off the ground further will give you much nicer graphs, but in truth they look pretty darn good outdoors. I don't remember if you have much room in your garage, but if your garage is cleaned out in the middle and you put the speakers on a 5ft stand, you'd be in good shape. Nonetheless those outdoor measurements will really teach you about SBIR problems.

 

[DanTheMan]

I decided to fix these Bose 201 series IV up a little bit further. I bought an Infinity car dome midrange for a very good deal. sort of assuming Infinity would build a great crossover to start with and mocked it up in a cardboard box with the car tweeters I had placed in the Bose. After fiddling with the crossover a bit, I was able to go from "complete disaster", to a very nice summing. However, the handoff is still audible and I think it's because of the discontinuity in radiation pattern:

I honestly can't imagine a situation where the stock crossover would work well.

In any case, I didn't have the parts on hand to make a 3-way work so I went back to the 2-way and eventually got to this (with on hand parts from the daze of my youth:

Those 4 measurements are covering about 60 degrees across the front on one side. The room is left in and the grill is in place. This is easily the best I've heard this speaker and even though its measurements aren't great, it manages to sound pretty reasonable. I may build a little notch filter in this later to knock down the 1.5kHzregion, but beyond that, this is a vast improvement over the original for not a lot of outlay. I'm sure it could be easily bettered for that matter. When Amir EQs these great speakers he gets to measure, he's being ultra picky. Make no mistake about that. Bose might still be in the home speaker making business had they just spent a little extra (money) on these (and several of their designs). If I can do this with a multimeter, measuring microphone and REW, imagine how easy it would be for a company using circuit modeling. I'd have to buy a $30 audio interface to make that happen and a couple resistors, but I'm content with these for the time being. I do have more perfect speakers, but these will be nice for the second system in the man cave.

 

[JimmyBuckets]

Any idea what the step response looks like on the home brew time aligned jobby?

 

[DanTheMan]

I’ll have to check. I don’t have access to my computer right now and I’m not sure I saved the measurements

 

[JimmyBuckets]

No worries was just interested to see if it were actually time coherent. Awesome post!

 

[bconline]

I have the original measurement file. But don't know how to generate the step response chart. Here's what I get from REW when I click on Impulse response and then the Step Response Overlay. Is this right??

 

[RickS]

This is comparable to looking at the IR so you need to choose % from the drop-down box in the upper left of the chart.

You need to do this same step as part of determining the correct measurement gate setting to eliminate first reflections. Explanation of IR windows for the gate is in the REW help...

How to do this is discussed in detail in this thread:

How to make quasi-anechoic speaker measurements/spinoramas with REW and VituixCAD

Quasi-anechoic measurements are a way to capture the 'real' frequency response of a speaker without the influence of a room. They are called 'quasi-anechoic' because they have the potential to get you really close to the results one might achieve with a non-reflective environment like an...

www.audiosciencereview.com

If after reading this you still do not understand what the gate setting should be, please post your REW project,

 

[bconline]

@Rick Sykora Thank you for the guidance. I'll read-up and learn something!

 

[Vladimir Filevski]

Mission 753 Freedom, from 1996. (Image from internet)

Measured frequency response, original crossover and stuffing/damping (blue=woofers, green=midranges, violet=tweeter, red=total, all shaped by the crossover). Smoothing 1/12 octave. Resonance in the vertical box channel behind midrange enclosure at 200 Hz. Dip at 3 kHz. Midrange cone breakup at 5 and 6 kHz, not suppressed by the original crossover.

I modified just the midrange filter and put 8 cm foam behind midrange enclosure - big difference:

Interestingly, price of the modified midrange filter is lower than than the original one. Why Mission opted for inferior crossover is a mystery...

 

[Vladimir Filevski]

B&W 603 S2 (not Anniversary!) from 1999. (Image from internet)

Measured frequency response (smoothing 1/12 octave). Infamous dip at 2.5 kHz and elevated tweeter output (so called "detailed sound"). Woofer sucks big time (weak magnet, Qts = 1,2).

I changed the topology of the woofer filter, upgraded midbass and woofer filter from first to second order (just two capacitors and one resistor added) and switched polarity of the tweeter. Much better! Off axis better too. Sensitivity improved by 1 - 2 dB.

 

[bconline]

Sorry for the long delay in posting. I finally had some time to learn more about how REW works. @JimmyBuckets and @Rick Sykora Does this step response look right for the time-aligned home brew speaker? I don't understand why it drifts into negative % territory after 2.5milliseconds.

 

[RickS]

Sorry for the long delay in posting. I finally had some time to learn more about how REW works. @JimmyBuckets and @Rick Sykora Does this step response look right for the time-aligned home brew speaker? I don't understand why it drifts into negative % territory after 2.5milliseconds.

View attachment 401445

No, should ideally on have only one spike. What is your timing reference?

In Analysis preferences, how is Align IR peak set?

 

[bconline]

No, should ideally on have only one spike. What is your timing reference?

In Analysis preferences, how is Align IR peak set?

These?

 

[bconline]

No, should ideally on have only one spike.

Could that be caused by the drivers not really being time-aligned relative to the measuring microphone?