There was a bit of discussion here: https://www.diyaudio.com/forums/mul...e-driver-output-capability-6.html#post6277664
On what happens when speakers are used way below their maximum capability.
Now, the test I did looks fairly conclusive - I was reliably collecting data from speakers that I couldn't hear acoustically, but a sensitive microphone was able to pick up the sweeps just fine.
That test speaker was a Kef KTS3001SE, a 4.5" coaxial, driven by an Akai AM2250 amplifier - nothing state-of-the-art, but they sound pretty good.
I'd like to do some more testing at very low SPLs, to see if anything does show up with different drivers etc. So far, we have one datapoint which suggests that speakers are linear down into inaudibility. It might be an outlier, or it might be typical - I don't know.
What I do have, though, is a set of very different drivers, a selection of amplifiers, and a willingness to test things.
Here's the list of drivers. I don't want to test all of them, but am happy to test a few.
- Seas H1252 - 8" aluminium cones
- B&C DE250 - 1" compression driver
- Some no-name 4" poly coned drivers
- Bose 2.5" full-range driver
- Fostex FE126E w/P10 treatments
- Tang Band W6-1139 mini-sub
- No-name 8" PA coaxials
- Faital 10HX230
Amplifiers
- Cambridge CXA80 (my daily driver HiFi amp)
- Crown MA5002VZ
- Crown MA12000i
- Powersoft T-series
My instinct says the CXA80 is the obvious choice for this testing. The Powersoft amps have DSP baked-in, although I could use a digital feed from my laptop. With the kilowatts of power available, though, I'd be concerned they'd be operating "down in the noise". Thoughts on that?
The microphone will remain a Beyer MC930, and the USB interface will be a Behringer UMC202HD. Despite it's budget pricing, the UMC202HD has a very low input noise, close to the theoretical lower limit based on thermal effects etc.
If anyone would like to send drivers in for testing, we can talk about that in the future. For now, I think it'd be best to see if there are any differences between drivers, so we should take extreme examples, such as Seas H1252 vs 8" PA coaxial, and Fostex FE126E vs no-name 4".
I'm hoping to get around to doing some measurements in a few days time, and will report back when I do. In the mean time, any suggestions or encouragement would be welcome.
Cheers, all.
Chris
On what happens when speakers are used way below their maximum capability.
Now, the test I did looks fairly conclusive - I was reliably collecting data from speakers that I couldn't hear acoustically, but a sensitive microphone was able to pick up the sweeps just fine.
That test speaker was a Kef KTS3001SE, a 4.5" coaxial, driven by an Akai AM2250 amplifier - nothing state-of-the-art, but they sound pretty good.
I'd like to do some more testing at very low SPLs, to see if anything does show up with different drivers etc. So far, we have one datapoint which suggests that speakers are linear down into inaudibility. It might be an outlier, or it might be typical - I don't know.
What I do have, though, is a set of very different drivers, a selection of amplifiers, and a willingness to test things.
Here's the list of drivers. I don't want to test all of them, but am happy to test a few.
- Seas H1252 - 8" aluminium cones
- B&C DE250 - 1" compression driver
- Some no-name 4" poly coned drivers
- Bose 2.5" full-range driver
- Fostex FE126E w/P10 treatments
- Tang Band W6-1139 mini-sub
- No-name 8" PA coaxials
- Faital 10HX230
Amplifiers
- Cambridge CXA80 (my daily driver HiFi amp)
- Crown MA5002VZ
- Crown MA12000i
- Powersoft T-series
My instinct says the CXA80 is the obvious choice for this testing. The Powersoft amps have DSP baked-in, although I could use a digital feed from my laptop. With the kilowatts of power available, though, I'd be concerned they'd be operating "down in the noise". Thoughts on that?
The microphone will remain a Beyer MC930, and the USB interface will be a Behringer UMC202HD. Despite it's budget pricing, the UMC202HD has a very low input noise, close to the theoretical lower limit based on thermal effects etc.
If anyone would like to send drivers in for testing, we can talk about that in the future. For now, I think it'd be best to see if there are any differences between drivers, so we should take extreme examples, such as Seas H1252 vs 8" PA coaxial, and Fostex FE126E vs no-name 4".
I'm hoping to get around to doing some measurements in a few days time, and will report back when I do. In the mean time, any suggestions or encouragement would be welcome.
Cheers, all.
Chris
Hi Chris,
I think I've read a few posts by Planet10 about 'downwards dynamic range' (or something like that) being important for sound quality, so think your plan to test is a good one.
A lot of people say pro drivers 'need a bit of juice' to get going which would suggest they think that the drivers don't follow a signal properly at low volumes.
Is there anyway you could work out a simple to copy measurement setup so that others could do similar tests and post the results ? I could measure my cheap JBL pro boxes and post the sweeps for example...(I use a Umik1 with REW or trueRTA)
Would be a bit easier than posting drivers back and forth.
Cheers,
Rob.
I think I've read a few posts by Planet10 about 'downwards dynamic range' (or something like that) being important for sound quality, so think your plan to test is a good one.
A lot of people say pro drivers 'need a bit of juice' to get going which would suggest they think that the drivers don't follow a signal properly at low volumes.
Is there anyway you could work out a simple to copy measurement setup so that others could do similar tests and post the results ? I could measure my cheap JBL pro boxes and post the sweeps for example...(I use a Umik1 with REW or trueRTA)
Would be a bit easier than posting drivers back and forth.
Cheers,
Rob.
My advice for the equipment you use is simply to make sure that you've picked the stuff with the lowest inherent noise. Your signal to noise ratio is already going to be suffering a lot because of the signal being so low, to increase measurement accuracy as much as possible you'll need to keep the noise down.
Do you have a box for those Faital subs you acquired recently? Or maybe just test your current Beyma subs.
My thinking is that heavy prosound drivers get the most accusations of sounding a bit flat at low volumes. I think it would be telling if they tracked linearly.
I'll give a shot as well. I have silicone double spider B&C 18's, BMS 15N850, Eminence NSW6021.
My thinking is that heavy prosound drivers get the most accusations of sounding a bit flat at low volumes. I think it would be telling if they tracked linearly.
I'll give a shot as well. I have silicone double spider B&C 18's, BMS 15N850, Eminence NSW6021.
Last edited:
This looks to be an interesting topic.
Two kinds of noise. If it's amplifier noise, at least you can average it down. The other kind..
Unless I'm mistaken this would also involve testing loud, and watching the tails. The concept may be a little unclear but surely it's not only about what can be done at low level, but also what can be heard at low level under the acoustic noise floor.
This is also what gets said about responses that aren't smooth. It's as if they drop out unevenly. I don't have any Fletcher Munson issues with a properly smooth response right down to the threshhold of hearing. As it turns out, cheap efficient pro drivers sometimes have an unsmooth response.RobWells said:
Hey Rob,
The reason I'm using the mic and soundcard I've chosen is because they're both low-noise. The Beyer MC930 often gets used for large-scale classical recording, where background noise during the quiet bits would be unacceptable. It has a noise figure of 16dBA, and 30mV/Pa output. ie, low-noise and high sensitivity (in terms of SPL in to voltage out). It also has the useful option of a 6dB/octave low-cut at 250Hz, which would even out the proximity effect.
By comparison, my Beyer MM1 measurement mic has 26dBA of background noise, and 15mV/Pa output. ie, 10dB more noise and 6dB less signal.
Your measurement setup may or may not work. You're welcome to give it a try.
Hi Ernie,
I haven't built the box yet - life makes other plans when you're not looking.
Do the drivers actually need to be in a box?
The mic will be less than an inch away from the cone, so I suspect the cabinet won't contribute much to the measurement. I think free-air measurements would be just fine.
That said, I do have a P.Audio SD21-1800N in a sealed box that I could dig out. That thing has enough moving mass to make the cabinet walk around.
FWIW, my bet is that accusations of "flat" or "lifeless" sound at low volumes has more to do with Fletcher-Munson curves (and/or harmonic distortion, where people might "want" 1% or whatever) than the drivers themselves. As I've pointed out, though, we only have one datapoint so far. The only way to find out will be to test.
Cheers, all.
Chris
Hi Chris,
I'd be happy to 'have a go' at this. If unboxed I can do a JBL M115 8A , a Lab12 and an EVM15L.
I've got a couple of CD/horn combos if you think it's worth trying them?
So driver unboxed, inside / outside ? Driver on the floor/table/clamped by magnet up in air?
Mic 1" from dustcap ?
Any gating on measurement?
Cheers,
Rob.
I'd be happy to 'have a go' at this. If unboxed I can do a JBL M115 8A , a Lab12 and an EVM15L.
I've got a couple of CD/horn combos if you think it's worth trying them?
So driver unboxed, inside / outside ? Driver on the floor/table/clamped by magnet up in air?
Mic 1" from dustcap ?
Any gating on measurement?
Cheers,
Rob.
Hey Chris. I have the Behringer UMC204HD and I suppose it's quiet. Distortion seems a bit high on a loop-back, but that might be the DAC.
The test results will be interesting. But as Allen mentions, maybe only half of the equation. But since we have so far not been able to get a clear picture of low level signals below loud ones, at least it's a start.
The test results will be interesting. But as Allen mentions, maybe only half of the equation. But since we have so far not been able to get a clear picture of low level signals below loud ones, at least it's a start.