Studio monitors response curves measured today with REW (input please)...

[rob aylestone]

I learned long ago that buying cheap products that were a little less than perfect, and tweaking them, was always a bit of an illusion. In my office, I run a pair of RCF 5" monitors. I have 5 pairs of them, in flight cases for stage edge fills. For music serious listening, they're just OK, but I know what they sound like, and I know that if I can hear the bass guitar on the E string, then that E will be way, way too loud in the mix. I can deal with it. Tweaking cheaper speakers is just pointless. They are what they are. Same with the folk who spend hours turning a cheap Chinese mic into a U86 and a half. It will always be nearly an 87. For years I toured with an EAW all in one PA cab - big 15" for bass, and a mid and hi unit. Loved them - always sounded good, but when I bought a new, better and mega expensive system, I wished I'd carried on with the old one! The new one was, in every way measurable, better. The trouble was, I liked the sound of the old one. Two boxes, lovely sound, but folk at the front were deafened and people at the back struggled to hear in many venues. Twelve boxes dangling gave the same volume everywhere, but just wasn't as nice. Luckily, people only heard it for two hours, and thought it great!

 

[Papanate]

I learned long ago that buying cheap products that were a little less than perfect, and tweaking them, was always a bit of an illusion. In my office, I run a pair of RCF 5" monitors. I have 5 pairs of them, in flight cases for stage edge fills. For music serious listening, they're just OK, but I know what they sound like, and I know that if I can hear the bass guitar on the E string, then that E will be way, way too loud in the mix. I can deal with it. Tweaking cheaper speakers is just pointless. They are what they are. Same with the folk who spend hours turning a cheap Chinese mic into a U86 and a half. It will always be nearly an 87. For years I toured with an EAW all in one PA cab - big 15" for bass, and a mid and hi unit. Loved them - always sounded good, but when I bought a new, better and mega expensive system, I wished I'd carried on with the old one! The new one was, in every way measurable, better. The trouble was, I liked the sound of the old one. Two boxes, lovely sound, but folk at the front were deafened and people at the back struggled to hear in many venues. Twelve boxes dangling gave the same volume everywhere, but just wasn't as nice. Luckily, people only heard it for two hours, and thought it great!

EAW 650s? I always thought they sounded great.

 

[Ally-007]

Today, I measured the two Behringer Truth B2031A speakers, outdoors, plus a third speaker which is the older variant called B2031. Here are the readings. The mic was in the same position for all three speakers (26" away from speaker, pointing at the spot between the woofer and tweeter). Admittedly, there was a little bit of extraneous background noise, such as distant cars, occasional distant pidgeon crooning, etc, but the difference between the readings looks bizzare to me. Speaker No2 looks wild! Only the older variant (B2031, the bottom graph), looks nice and flat (ish). Unfortunately I only have one of those! It's possible that I was doing something wrong during the measurement process. The mic was on the verge of feedback, in order to get the levels acceptable, according to REW.

 

[rob aylestone]

EAW 650s? I always thought they sounded great.

No - LA325's Very heavy, and probably at least 5 coats of paint through their life, but never let us down.

 

[bouldersoundguy]

What does "mic was on the verge of feedback" mean? The mic's input should not be routed to the speakers.

 

[Papanate]

No - LA325's Very heavy, and probably at least 5 coats of paint through their life, but never let us down.

Those had 2 x 15” Speakers and were 3 way systems - very serious - lots of volume.

 

[Ally-007]

What does "mic was on the verge of feedback" mean? The mic's input should not be routed to the speakers.

Thanks for that, BSG. I had the mic plugged into channel 2 of the audio interface and used the same channel for output, going to whichever single speaker was being tested. Perhaps that was a mistake, was it?

 

[bouldersoundguy]

Thanks for that, BSG. I had the mic plugged into channel 2 of the audio interface and used the same channel for output, going to whichever single speaker was being tested. Perhaps that was a mistake, was it?

Input monitoring should be off.

 

[rob aylestone]

That could indeed have wrecked it because you get weird comb filtering effects that add in spikes to the response curve where they overlap which combine or reject to a degree.

 

[Ally-007]

Input monitoring should be off.

Thank you, B! I actually thought of that, before today's repeated tests (see further below). Big difference! But still troubled by the wildly un-flat resposes recorded.

 

[Ally-007]

That could indeed have wrecked it because you get weird comb filtering effects that add in spikes to the response curve where they overlap which combine or reject to a degree.

Thank you, Rob. Yes, could well be, as I got vastly more comparable curves from the three speakers, today (see post below).

 

[Ally-007]

Today, I tested the three speakers again, and got much more comparable curves from the three Truth speakers (probably because the REW setup was more correct and input monitoring was switched off). All three speakers were measured outdoors, with the measurement mic 26" away from each speaker, pointing to the spot midway between woofer and tweeter. All speakers had their room EQ adjusters set to 0, i.e., no EQ applied. Extraneous noise was mostly just a bit of wind rustling the leaves of a nearby tree. While the three curves from today's test are much more comparable, I am very perplexed at the drastic lack of flatness (approx 30dB of variance, which is obviously way out of line for so-called studio monitors). I'm wondering why all three appear to be so insanely un-flat compared to other peoples' measurements of this speaker model. Can anyone suggest why? Thank you!
Image below:
RED = Behringer Truth B2031A (speaker A).
GREEN = Behringer Truth B2031A (Speaker B).
BLUE= Behringer B2031 (Speaker C - an older variant, acquired for spare parts, if needed).

 

[rob aylestone]

The three traces are so similar with just subtle differences that makes me think the measurement system is the culprit here. I think you said the mic was an 8000 measurement mic, so we can be pretty sure it is pretty flat. Outside means no resonances, so 30dB of reduction at some points indicates something very odd. Have you tried spot frequency tones and a sweep? The dip is centred on three octaves in the middle, and with such a reduction in level, a 20 - 20kHz sweep should reveal this to your ears - and if you record the mic on a track, you should be able to also see the level drop as the sweep goes up. If it's the speakers, then the sweep tone should also drop. If the sweep tone remains nearly the same level with a few dB sway - then your pink noise source has been filtered so is not equal energy at audio frequencies. I suppose the mic could be faulty, so swapping that out should look a little different - again, just a few dB sway around where it should be. The three almost identical traces makes me suspect an issue with the test conditions.

 

[TalismanRich]

Manufacturers measure their speakers in a anechoic chamber, so that any reflection nodes will be absent. That way they only record the output of the speakers themselves, not the room.
You might want to read this article about a means to measure a speaker in a "real room".

How to Measure Free-Field Speaker Response Without an Anechoic Chamber

One of the biggest challenges with working from home during the pandemic was not having access to an anechoic chamber. This short tutorial, explains a simple solution for this: a “splice” measurement. The capability for such measurements has been included in SoundCheck since 2001. Despite its...

audioxpress.com

Obviously the problem is that the actual performance in your room in a real situation may not be anywhere close to the true output of the speaker, since room nodes will always be in play to some degree.

 

[Ally-007]

The three traces are so similar with just subtle differences that makes me think the measurement system is the culprit here. I think you said the mic was an 8000 measurement mic, so we can be pretty sure it is pretty flat. Outside means no resonances, so 30dB of reduction at some points indicates something very odd. Have you tried spot frequency tones and a sweep? The dip is centred on three octaves in the middle, and with such a reduction in level, a 20 - 20kHz sweep should reveal this to your ears - and if you record the mic on a track, you should be able to also see the level drop as the sweep goes up. If it's the speakers, then the sweep tone should also drop. If the sweep tone remains nearly the same level with a few dB sway - then your pink noise source has been filtered so is not equal energy at audio frequencies. I suppose the mic could be faulty, so swapping that out should look a little different - again, just a few dB sway around where it should be. The three almost identical traces makes me suspect an issue with the test conditions.

You may be right, Bob. I hope so! I only have the one measurement mic, unfortunately, and those things are expensive. To buy an alternative one would likely cost half of what I paid for all three speakers. One thing that could be a culprit, is that I had each speaker sitting on the edge of a plastic garden table. It's possible that the table had some resonance, accentuating the bass frequencies. I will run the test again, tomorrow, with the speakers sitting a a Workmate collapsable DIY saw bench, which should be practically resonance-free, as it has very little in the way of flat surfaces. I'll also think about recording a sine wave sweep. Thank you.

 

[Ally-007]

Manufacturers measure their speakers in a anechoic chamber, so that any reflection nodes will be absent. That way they only record the output of the speakers themselves, not the room.
You might want to read this article about a means to measure a speaker in a "real room".

How to Measure Free-Field Speaker Response Without an Anechoic Chamber

One of the biggest challenges with working from home during the pandemic was not having access to an anechoic chamber. This short tutorial, explains a simple solution for this: a “splice” measurement. The capability for such measurements has been included in SoundCheck since 2001. Despite its...

audioxpress.com

Obviously the problem is that the actual performance in your room in a real situation may not be anywhere close to the true output of the speaker, since room nodes will always be in play to some degree.

Thanks for the input, Rich. I will think about that article. I should stress that my latest measurements were done outdoors, in the middle of a lawn, to eliminate room resonances though.

 

[rob aylestone]

I meant use any old mic - as their frequency response ups and down are rarely much different where you have issues

 

[Ally-007]

I meant use any old mic - as their frequency response ups and down are rarely much different where you have issues

Oh I see. Yes, I can do that, for sure. Thanks!

 

[bouldersoundguy]

I would put a folded up towel under the speaker to isolate it from whatever it's on. You could also try putting the mic in line with the tweeter.

 

[Ally-007]

Wow, this is more like it!! After reading your helpful comments, this time I placed each speaker atop my 5ft concrete-block garden wall, facing away from it at a 45-degree angle, and such that the nearest other garden walls were 18 yards away, behind the mic, and at 45-degree angles, to minimise any reflections coming back towards the mic. There was nothing behind the speaker position but a large open field. The nearest building was also ten yards further away than before. Background noise was also reduced, as the breeze and distant noises had subsided considerably. I never thought these changes would make such a huge difference! This time, the mic was placed one metre from the speaker, in each case. The three new response curves look vastly more desirable than before! All three curves have 1/6 smoothing applied, as before.

I'm still puzzled by the speaker represented by the green line, which is so different from the other two, above 2kHz. Perhaps the tweeter is sub-par. (2kHz is the croossover frequency.) I guess I could use the pink-line speaker and the blue-line speakers together as a pair, even though pink is a B2031A, and blue is the older B2031 variant. Or I could try swapping two of the tweeters over. I may be able to flatten the 70Hz-150Hz hump with EQ, and maybe boost 40kHz to 65kHz. I'm not bothered by the incline from 3kHz to 10kHz, as this should compensate for my age-related high-frequency hearing loss. Thanks in advance for any further insights you can share. It'll be much appreciated.