Is attack speed being measured by anyone?

[KEW]

The "tightness" of a sub is often discussed, but does it ever get measured?

It seems like it would be easy enough to send a rim tap (or for the sake of standardization a specific signal such as a square wave) and see what the response at a microphone looks like. I'm sure the best would look rather poor (compared to an amp's performance), but relative differences would be good to see.

Would this work?
Is it already being done, and I just don't recognize it (Ilkka has a few graphs I don't follow)?

TIA
Kurt

 

[jostenmeat]

I'm utterly unqualified to even open my mouth here, but I think that what many people subjectively hear as "tight" is in fact better decay, or less reverberation. Less overhang. As far as speed, if it was actually faster, then the frequency would be higher.

 

[highfigh]

How fast it responds to acceleration and deceleration is partly based on the mass of the motor assembly and on the ability of the magnet/voice coil to cause it to move. No speaker will reproduce a square wave perfectly but that would be a good test if you wanted to check into this. If you do, you need to compare what goes in with what the speaker produces. Impulse tests, shown on an oscilloscope will tell you how much it rings, too.

 

[annunaki]

Most of what many call "speed", or "attack" in this case, is generated by the mid-bass units in a speaker system rather than the sub itself. The sub just helps to reinforce it.

There is more to it all than this but it has been covered many times. Jostenmeat is on the right path here.

 

[KEW]

I'm utterly unqualified to even open my mouth here, but I think that what many people subjectively hear as "tight" is in fact better decay, or less reverberation. Less overhang.

If you are correct on the overhang, it seems like that would show on a mic as well.

As far as speed, if it was actually faster, then the frequency would be higher.

I am also over my head, but can muse; and what better way to learn

There is the speed as in the frequency of the vibrations which would change the pitch.
But there is also speed as in how quickly does the cone jump to the peak of the first oscillation. But you make a good point, much of my curiosity is how close to the full stroke (full signal ampitude) does the cone get before it reverses direction? Does it overshoot? Mechanical systems aren't very pretty compared to electronics - having mass and inertia. I imagine that the perimeter of the cone is still going out as the interior is reversing. In any case, it seems like mic'ing a signal with a sharp rise and fall would give you an idea of how well the sub is mimicing the input signal.

 

[KEW]

Thanks for the additional replies.

annunaki,
So would it not be beneficial to measure the degree to which the attack reinforces the midbass unit? What is it that makes a sloppy sub sloppy?
Is it all about FR or is there another measure which could show "tightness".

I know on standard speakers, some offer more detail than others. I've always assumed that these are more precisely controlled drivers, but that would not show in a FR curve (maybe it shows i the distortion?)

I have read much about this, but I guess I am not finding the right places.

 

[KEW]

How fast it responds to acceleration and deceleration is partly based on the mass of the motor assembly and on the ability of the magnet/voice coil to cause it to move. No speaker will reproduce a square wave perfectly but that would be a good test if you wanted to check into this. If you do, you need to compare what goes in with what the speaker produces. Impulse tests, shown on an oscilloscope will tell you how much it rings, too.

Yes, I think an impulse test would probably cover what I am after.

But, why isn't this part of testing. Is it simply of no consequence to the realized SQ?

 

[mtrycrafts]

.. As far as speed, if it was actually faster, then the frequency would be higher.

But then, the crossovers would pass it up the the higher frequency drivers.

 

[mike c]

sorry could not resist ... but when i think of attack speed ... i think:

 

[highfigh]

Yes, I think an impulse test would probably cover what I am after.

But, why isn't this part of testing. Is it simply of no consequence to the realized SQ?

Since a speaker isn't a particularly efficient way of transforming electrical impulses into sound, a little more precise won't make much difference unless it's a high power situation. It may be a noble quest to make a speaker that can follow a square wave accurately but when the room has multiple reflections and standing waves, I doubt it would be of much benefit.

BTW- speed and frequency are two different issues- the cone just moves from Point A to Point B faster, not more often. It also means that once it gets to one of those points, it stops faster. It's changing electricity into mechanical motion and this method (heavy cone and motor assembly) isn't the best way but it works pretty well, in many cases.

It is part of testing for some companies. Look at Peerless and Vifa spec sheets- they usually have an impulse plot.

 

[TLS Guy]

The "tightness" of a sub is often discussed, but does it ever get measured?

It seems like it would be easy enough to send a rim tap (or for the sake of standardization a specific signal such as a square wave) and see what the response at a microphone looks like. I'm sure the best would look rather poor (compared to an amp's performance), but relative differences would be good to see.

Would this work?
Is it already being done, and I just don't recognize it (Ilkka has a few graphs I don't follow)?

TIA
Kurt

You are clinging to the myth of "fast bass". There is no such thing. To produce a spike you need a wide band frequency response. A sub only covers very low frequencies and does not have the frequency response to produce a fast upright spike.

You can measure what you want and it is frequently done. The number is Qts.

This reflects the quality of the resonance, or broadness of the resonant peak.

Below a Qts of 0.5 the bass will seem to stop before it gets going and seem unnatural. Above 1.0 it will be definitely boomy. Popular wisdom says 0.7 for Qts is about right, personally I think it is too high. I like it around 0.5.

Now you won't get that from a ported or PR sub, unless you use a second order coupled cavity sub, where Qts can be tightly controlled. That is why I favor TL speakers, even though they take up a lot of real estate. You can do it with a sealed sub, but you have to use brute force to get bass extension, which means an expensive driver with huge xmax, so you can equalize to get bass extension, as F3 is so high without it. To drive it you need a mega amp.

The TL on the other hand can easily achieve low bass extension, at good efficiency without pushing things to the limit. The driver gets support from the pipe over about 1.5 octaves to support the direct radiation from the speaker. At the same time you get a natural deep bass that is light on its feet.

Part of the reason for this is that a loudspeaker cone is a very poor coupler to the room. A ported Qb4 solves this, as the port couples well, but the reproduction is prone to be resonant with highish Qts. A pipe couples very well to the space and a horn best of all, with the highest efficiency.

 

[3db]

The "tightness" of a sub is often discussed, but does it ever get measured?

It seems like it would be easy enough to send a rim tap (or for the sake of standardization a specific signal such as a square wave) and see what the response at a microphone looks like. I'm sure the best would look rather poor (compared to an amp's performance), but relative differences would be good to see.

Would this work?
Is it already being done, and I just don't recognize it (Ilkka has a few graphs I don't follow)?

TIA
Kurt

My quess would be using an impulse to meausre the response of the sub. The closer the output resembles the input espcially with the same decay rate as the input, then I suppose you have a "fast tight " sub. The longer the decay (measured against the input signal , the more sloppy the sub would sound.

 

[annunaki]

So long as a woofer is properly designed with a motor designed to effectively handle its mass, there should not be anything to worry about with the driver itself. The rest comes down to enclosure, loading, and in a lot of cases, the listener's preference and reference points.

There are many variables in any system which can dictate how the sub sounds in relation to accuracy. TLS Guy was along the right path here. The enclosure has a lot to do with the final result.

 

[TLS Guy]

My quess would be using an impulse to meausre the response of the sub. The closer the output resembles the input espcially with the same decay rate as the input, then I suppose you have a "fast tight " sub. The longer the decay (measured against the input signal , the more sloppy the sub would sound.

No sub would reproduce the impulse. An impulse has frequencies out to 20 kHz at least.

Fournier analysis will demonstrate that.

So a sub will horribly round off the spike. That in no way means it is a bad sub. You are just looking at the wrong parameter, and trying to get your data the wrong way. In fact in a totally meaningless way.

 

[lsiberian]

The key measurement in bass performance is distortion. The lower the distortion the better the driver. This is why an Ultra will sound better than Dayton Titanic. It has lower distortion.

The box tuning is also important. A .7 tuning is desired though it usually goes lower with proper stuffing. This is where people say it makes the box appear larger because technically it lowers the tuning of the box.

 

[mtrycrafts]

sorry could not resist ... but when i think of attack speed ... i think:...]

Good one
By the way, time to upgrade the little ones pictures

 

[mtrycrafts]

The key measurement in bass performance is distortion. The lower the distortion the better the driver. This is why an Ultra will sound better than Dayton Titanic. It has lower distortion.

The box tuning is also important. A .7 tuning is desired though it usually goes lower with proper stuffing. This is where people say it makes the box appear larger because technically it lowers the tuning of the box.

Why are sub performance measure with a 10% distortion? Because you cannot hear it at the sub's band.

 

[KEW]

Thanks for the responses.
The "Transient response" parameter mentioned in the "Ultimate sub shoot-out" thread sounds a lot like what I had in mind and I'm curious to read about how that is done and see the result.

Cheers,
Kurt