Is SPL less at bass frequencies when the driver is having to cover a full range input

[Enstip]

Apologies if this is a stupid question as my knowledge is limited (or you may conclude, nil).

I can see from the frequency response chart for my bookshelf speakers that they start to roll off at around 80Hz. I guess these curves are measured using a frequency sweep input, meaning the SPL at each frequency is in reality, the level achieved at that instant, as if a single frequency tone was being input, despite the sweep being a continuum.

So my question is, if the input is real music, and hence a complex waveform of many frequencies, would the speaker still achieve the same SPL for a given frequency component of the complex input, or does the fact that it is producing such a varied output mean the SPL for a given frequency, maybe particularly bass frequencies, is harder to achieve?

 

[boxerfan88]

Multi-tone chart would give and approximation of how the driver or speaker will perform with music like signal.

Keep pushing multi-tone levels and see how much SPL is generated at the various bands will answer your question.

For general understanding, you can take a look at manufacturer who provide those max spl charts. Typically we will see the mid-woofer struggling with bass at high spl, yet able to push out high spl in the midrange.

If u play super loud and love bass, you cannot run away from having subwoofer. For bass, nothing beats cone surface area & displacement.

 

[NTK]

Loudspeakers, ideally, behave as linear devices. A defining characteristics of a linear systems is that, when:
Input = A ⇒ Output = X
and Input = B ⇒ Output = Y
then Input = A + B ⇒ Output = X + Y

This property is known as "superposition", which basically means that, when the loudspeaker is in its linear operating range, the total output of any combinations of inputs will be the sum of the individual output for each of the input. Therefore, the SPL of the each of the frequency component in the output should be the same as if it is reproducing only one single frequency component.

Without this property, EQ will not work, as the prescribed level of boost or cut at one frequency will be altered by the presence or absence and levels of contents in other frequencies.

As boxerfan88 wrote in post #2, we can get an idea on when the linear operating range ends using a multi-tone test. Below is the results of one of Erin's test. The top graph shows the distortion levels at a few input levels, and you can see that the speaker had a lot more trouble at the low frequencies where the required cone excursion is much higher. The bottom graph shows the SPL deviations and is probably more of the direct answer to your original question.

Yamaha NS-10M Studio Speaker Review

Yamaha NS-10M Studio Review

www.erinsaudiocorner.com

 

[alankila]

Apologies if this is a stupid question as my knowledge is limited (or you may conclude, nil).

I can see from the frequency response chart for my bookshelf speakers that they start to roll off at around 80Hz. I guess these curves are measured using a frequency sweep input, meaning the SPL at each frequency is in reality, the level achieved at that instant, as if a single frequency tone was being input, despite the sweep being a continuum.

So my question is, if the input is real music, and hence a complex waveform of many frequencies, would the speaker still achieve the same SPL for a given frequency component of the complex input, or does the fact that it is producing such a varied output mean the SPL for a given frequency, maybe particularly bass frequencies, is harder to achieve?

Yes, it will behave almost exactly like this. The frequency response given by the analysis is real. What they do with the log sweep is a mathematical trick where they produce the impulse response of the system by dividing the measurement with the original excitation signal that went into the system, and the impulse response derived this way is turned into the frequency response. The same analysis also yields the harmonic distortion or the nonlinear parts of the response.

I often use broadband pink noise moving microphone method to confirm the single point exponential sweep results, and they overlap like 100 % until the exact microphone location begins to matter somewhere around 300 Hz, and beyond that point single point and moving mic method measurements begin to diverge.

 

[Multicore]

So my question is, if the input is real music, and hence a complex waveform of many frequencies, would the speaker still achieve the same SPL for a given frequency component of the complex input, or does the fact that it is producing such a varied output mean the SPL for a given frequency, maybe particularly bass frequencies, is harder to achieve?

If the speaker can produce a certain frequency only in the absence of others then that would be an example of distortion, i.e. non-linear behavior. Many speakers do in fact distort at low frequencies and at SPL that is normal in real life use. Your question concerns the speaker at the limits of its low frequency linear behavior and it can't be answered in general. We need specifics.

For example, this speaker at low frequency and 86 to 90 dB SPL produces more 2nd and 3rd harmonic than fundamental at some low frequencies. A big old main monitor will distort much less.

So your question then becomes: how much distortion and how bad is it in real-life listening? The best we can usually do to predicatively answer those questions is to look at the frequency response and distortion graphs for your bookshelf speaker together with measurements or estimates of the SPL and spectrum of the signals you listen to. Alternatively you could actually measure it yourself in situ.

 

[Enstip]

Thanks for the replies so far. Perhaps I need to take this a step at the time, if you could comment as we go. Below is a low frequency sweep of each speaker (Focal Kanta No1 by the way). To my novice mind, the left channel is getting bass boost through the corner of the room and the right channel bass loss due to the distance from the corner and openings. Until about 60Hz that is, when the effects seem to swap. How is this explained? (Mic is at listening position btw, with no EQ, for both channels distortion is significant until 42Hz).

 

[Heinrich]

Apologies if this is a stupid question as my knowledge is limited (or you may conclude, nil).

There is no frequency in nature. Frequency is a mathematical model. In parts the hearing follows a similar approach, aka "bands", but it is not exactly the same.

When it comes to measure a speaker, the signal is a voltage transfered to a power at the speaker's input. That power is converted to a force, hence to an acceleration (Newton's mechanics). This acceleration makes the sound. It doesn't depend in any way of frequency!

It is only so, that for convenience the input and the output are described in terms of frequency. As said, that's just and only a mathematically consistent way to describe forces.

Of course there is more into it, but that needs - as you may have suspected, maths.

It is absolutely irritating, maybe even wrong to explain that stuff without the Fourier series.

 

[boxerfan88]

From the chart, I see a room mode approximately 62Hz, and a null approximately 123Hz. The L & R are shifted a little from each other probably because the room isn’t symmetrical.

 

[RayDunzl]

So my question is, if the input is real music, and hence a complex waveform of many frequencies, would the speaker still achieve the same SPL for a given frequency component of the complex input, or does the fact that it is producing such a varied output mean the SPL for a given frequency, maybe particularly bass frequencies, is harder to achieve?

Excursion is related to frequency for a single speaker.

According to this estimation, it quickly becomes moot as the frequency rises.

For 95dB at 1m with a 6.5" speaker:

Add the excursion for the different frequencies for a possible peak value.

Most of the time some would cancel others to some degree.

So, a speaker producing multiple tones is unlikely to have a problem with excursion limits.

It will wiggle with the signal.

 

[a4eaudio]

Thanks for the replies so far. Perhaps I need to take this a step at the time, if you could comment as we go. Below is a low frequency sweep of each speaker (Focal Kanta No1 by the way). To my novice mind, the left channel is getting bass boost through the corner of the room and the right channel bass loss due to the distance from the corner and openings. Until about 60Hz that is, when the effects seem to swap. How is this explained? (Mic is at listening position btw, with no EQ, for both channels distortion is significant until 42Hz).

View attachment 440502

The two speakers are not playing any differently. Take measurements of each speaker at about 1m, far from any boundaries and gate the measurements and you will see that. However, the sound pressure levels ARE different at the listening position due to room reflections. Move both speakers to different spots and measure from a new listening position and everything will change. But the speakers haven't changed, their output is still exactly the same (as each other, and as they were at the previous locations in the room).

 

[gnarly]

Excursion is related to frequency for a single speaker.

According to this estimation, it quickly becomes moot as the frequency rises.

For 95dB at 1m with a 6.5" speaker:

View attachment 440520

Hi, Something appears bad wrong with that chart.

A 6.5" speaker can't make 20Hz at 95dB at 1m.........can't come remotely close .

Using piston calc, which has proven accurate for me many times, with various sized drivers in DIY multi-ways,
looks like about 78dB. https://www.baudline.com/erik/bass/xmaxer.html

 

[thewas]

One could add 3-6 dB for a ported design (like most nowadays) but ports for such typical sized home hifi loudspeakers are usually not tuned as deep so actually the opposite happens, the output is actually even lower then without one at 20 Hz.

 

[sigbergaudio]

A lot of very complicated answers here, given the fact that you are saying that your knowledge is very limited.

"if the input is real music, and hence a complex waveform of many frequencies, would the speaker still achieve the same SPL for a given frequency component of the complex input(..)"

Short answer is Yes for all practical purposes they will behave like you'd expect from the frequency response, despite the fact that it is playing a multi-frequency signal (aka music).

The exception is if you play louder than your loudspeakers are designed to do. Then the response will become unlinear depending on which driver(s) reaches their maximum capacity. Often that happens in the bass first, so then you may get even less bass compared to the rest of the frequency range, compared to what the response chart for the speakers imply.

 

[Heinrich]

"if the input is real music, and hence a complex waveform of many frequencies, would the speaker still achieve the same SPL for a given frequency component of the complex input(..)"

What if not, listening experience? In, case would it depend on the absolute volume? And so forth.

 

[sigbergaudio]

What if not, listening experience? In, case would it depend on the absolute volume? And so forth.

I don't understand the question. Can you please rephrase and/or elaborate?

 

[Heinrich]

I don't understand the question. Can you please rephrase and/or elaborate?

Oh, it wasn't a real question, but a rhetorical one. If the linear superposition of signals (there is no frequency in nature) would not hold, what are the consequences?
I think it was in order to give a hint, how the original poster's question could be answered by just taking advantage of simple dialectics.

I put more and more "frequencies" in, in case, what would I expect as a sound, how would that sound?
Would it be dependent on the absolute sound pressure level.
Is that effect a genaral one, or only applicable to speakers?
And so forth.

Only that few questions/answers may resolve the issue the OP had.

 

[gnarly]

Excursion is related to frequency for a single speaker.

According to this estimation, it quickly becomes moot as the frequency rises.

For 95dB at 1m with a 6.5" speaker:

View attachment 440520

Add the excursion for the different frequencies for a possible peak value.

Most of the time some would cancel others to some degree.

So, a speaker producing multiple tones is unlikely to have a problem with excursion limits.

It will wiggle with the signal.

Hi again, adding a second reply to this post, because despite the 20Hz data being off somehow (and presumably rest of chart off as well),
I do think looking at excursion hits the nail on the head in attempting to answer OP's question about potential bass limitation with music volume increase.

A fundamental concept is that for equal SPL across the frequency spectrum, displacement of a driver need to increase 4X per octave lowering.
Since a driver's area is fixed, it means for whatever frequency span it is covering, it needs to increase excursion 4X for each octave reduction....to maintain equal SPL across the range.
4X excursion requires 4X the power (wattage).

An amp's wattage capacity gets drained by low frequency content. This limits/compresses bass capability. (There's also the issue of can the driver handle the needed excursion to begin with.)
An amp's power sag due to low frequency excursion, also limits the voltage available for higher frequency content/transients, where excursion and current draw are comparatively much less and would not be effected were it not for the lower frequency power draw.
This is an aspect of compression (amplifier induced compression) that a sine sweep won't show, and can be particularly acute for passive designs.

 

[Heinrich]

So my question is, if the input is real music, and hence a complex waveform of many frequencies, would the speaker still achieve the same SPL for a given frequency component of the complex input,...

Yes, for the most part. If it does not, then we call this phenomenon "distortion".

 

[sigbergaudio]

Yes, for the most part. If it does not, then we call this phenomenon "distortion".

Or perhaps more accurate compression. We can see distortion and still reproduce a linear fundamental response at the expected SPL (or even higher if the distortion adds to the original signal).

While if the amplifier and/or drivers in the loudspeaker are unable to maintain the requested SPL, we see compression, aka reduction in some of or all of the frequency range.

 

[RayDunzl]

4X excursion requires 4X the power (wattage).

At a specific frequency, maybe.

I should think not when comparing octaves.

If so then across 10 octaves the difference in power would be huge.

4x4x4x4x4x4x4x4x4 = 262144 times the power from high to low frequency, same SPL. It doesn't happen.

You put 1 watt across speaker across the frequency range and the SPL is the same at all frequencies.

Which is how the frequency response of a speaker might be measured - pick an appropriate fixed voltage for the test.

The power delivered by a steady voltage amplitude across a fixed impedance doesn't vary with frequency.

(The SPL of a real speaker with varying impedance will vary a couple of dB, I get that)