Jump to content
IGNORED

Speaker Design & Response Targets


Recommended Posts

Posted (edited)

Edit:¬† HeheūüėÄ the first few comments here are extracted out of a Beginner thread that had became too technical, so it now looks like I started this thread 'out of the blue'!¬† In fact, I was just trying to defend a bloke who liked the sound of a speaker that someone else implied had a crap frequ response curve. Back to the original commentary:

 

Frequency curve preferences are so highly individual that a sweep result doesn't say much on its own.  Not to mention how much the purchaser's room effect will also come into it. 

 

In any case, speakers can be DSP'd and/or given subwoofer support at any future time.

 

But what you cannot do is add better dynamics or resolution to a speaker once purchased!

Edited by tripitaka
  • Like 1
Link to comment
Share on other sites



3 hours ago, Satanica said:

 

Well it all depends on what your goals are, but a flat in room response as low as possible simply means you have a system that is relaying more information of the recording. So for example, the open E-string on a bass guitar (which is the top thickest string) has a fundamental tone of 41 hertz. So at this frequency the speaker should be flat, to reproduce what is on the recording. The speaker in question is over 10 decibels lower at this frequency and that means it’s less than half as loud as it should be at recreating E-string notes of a bass guitar. Of course room acoustics come into play, which could make it better at this particular frequency, but they could also make it worse.

 

There is also the missing fundamental effect which causes people to hear bass which is not there. The brain creates the missing bass from the harmonics. Of course it's not the same as a real 41Hz note. 

  • Like 1
  • Wow 1
Link to comment
Share on other sites

Posted (edited)
3 hours ago, tripitaka said:

Frequency curve preferences are so highly individual that a sweep result doesn't say much on its own.  Not to mention how much the purchaser's room effect will also come into it. 

 

In any case, speakers can be DSP'd and/or given subwoofer support at any future time.

 

But what you cannot do is add better dynamics or resolution to a speaker once purchased!

 

Every speaker should be designed as neutral as  possible, to match those neutral as possible electronics. From there one can adjust whatever one likes. But unless I'm mistaken we all actually are far more similar than different in what in room response we are going to prefer from a given speaker in a given room. 

 

I'm not sure you understand this but CTA-2034 measurements as in the case of a using a Klippel near field scanner (Erin's Audio Corner) or anechoic chamber means that multiple frequency response measurements are taken from multiple angles. From memory I think it may be as many as 70. There's nowhere for a speaker to hide. 

 

The room will definitely start to dominate the sound under about 400Hz, above that the speaker will dominate the sound and with the  anechoic measurements can derive an accurate depiction of what the in room response will be. 

 

Equalisation can can only do so much and cannot fix all speaker problems, especially in frequency ranges that the directivity is poor. Because equalising correctly for one speaker angle will be wrong for another. I think this is mentioned in the review. 

Edited by Satanica
  • Like 3
Link to comment
Share on other sites

6 hours ago, Satanica said:

 

Every speaker should be designed as neutral as  possible, to match those neutral as possible electronics.

 

Sorry Satanica, while I usually agree with what you say, I don't agree with this one. People who want objective neutrality are a subset of our hobby. There are many, many more who like a bit of coloration and non-Harman and non-neutral frequency response.  Different products for different taste and different approaches to the hobby. I have come across systems built to a completely different philosophy to mine, and the vast majority of them have been enjoyable, despite the presence of flaws. It's like Cindy Crawford's mole. 

  • Like 3
Link to comment
Share on other sites

Even if it were a global principle that all speakers should be designed to be as neutral as possible, in reality there are so many things to balance that a designer just comes up with whatever they think is the best compromise within the manufacturing cost  But others may take a wildly differently view about the choices made and the nature of the compromise achieved?

Link to comment
Share on other sites



2 hours ago, Keith_W said:

Sorry Satanica, while I usually agree with what you say, I don't agree with this one. People who want objective neutrality are a subset of our hobby. There are many, many more who like a bit of coloration and non-Harman and non-neutral frequency response.  Different products for different taste and different approaches to the hobby. I have come across systems built to a completely different philosophy to mine, and the vast majority of them have been enjoyable, despite the presence of flaws. It's like Cindy Crawford's mole. 

Hi, no need to apologise for not agreeing with me. I think that not all but most speakers are designed with neutrality as the goal. Neutrality is simply a white canvas for which you can add colour if you so choose. When I refer to neutrality I mean anachoic neutrality which does not mean flat in room response. I think you may have linked to this not that long ago but I link it again for any others that are intetrested.

I just try to keep things as simple as possible. Dr T (I presume you know who I mean) says neutrality is the name of the game and who am I to argue?

Link to comment
Share on other sites

Posted (edited)
2 hours ago, tripitaka said:

Even if it were a global principle that all speakers should be designed to be as neutral as possible, in reality there are so many things to balance that a designer just comes up with whatever they think is the best compromise within the manufacturing cost  But others may take a wildly differently view about the choices made and the nature of the compromise achieved?

I think you're right and an obvious example is to sacrifice a bit of neutrality for more output. For a particular budget it probably mostly comes down to what compromises a designer wants to make. But, I think one of the biggest audiophile mistakes is to think that speakers (and other components) primarily come in different colours or flavours, when really they come in different versions of non perfect neutrality.

Edited by Satanica
Link to comment
Share on other sites

1 minute ago, Satanica said:

I just try to keep things as simple as possible. Dr T (I presume you know who I mean) says neutrality is the name of the game and who am I to argue?

 

Yes, I did link to that video some time ago :) And with full respect to Dr. T, I presume you are talking about his papers with Dr. O about preferred target curves. Unfortunately I have not read the original papers, but I did read his description of his papers in his book. What struck me was the rather small sample size which seemed to consist mostly of Harman employees. I suspect that if he recruited a larger sample size, like say 200 random people from SNA, the standard deviations would be much wider and he may not even be able to draw a preferred target curve ;) 
 

  • Like 2
Link to comment
Share on other sites

Posted (edited)
40 minutes ago, Keith_W said:

 

Yes, I did link to that video some time ago :) And with full respect to Dr. T, I presume you are talking about his papers with Dr. O about preferred target curves. Unfortunately I have not read the original papers, but I did read his description of his papers in his book. What struck me was the rather small sample size which seemed to consist mostly of Harman employees. I suspect that if he recruited a larger sample size, like say 200 random people from SNA, the standard deviations would be much wider and he may not even be able to draw a preferred target curve ;) 
 

I'm not really sure if we are talking the same thing or not. I'm not referring to in room target curves, I'm referring to the anechoic measurements that should measure flat (neutral). But if you're referring to in room Harman Curves, they are based on the assumption that a speaker measures flat (anechoic).

Edited by Satanica
Link to comment
Share on other sites

On 08/07/2023 at 12:25 PM, tripitaka said:

But what you cannot do is add better dynamics or resolution to a speaker once purchased!

Those two terms are bandied about a lot on audiophile forums. What do they really mean in relation to a speaker enclosure, its drivers, and its crossover (if present and if used)?

 

Dynamics

At very high driving levels a speaker cone can reach the limit of its excursion forwards or backwards resulting in a flattening of the waveform, not dissimilar to amplifier clipping,. Obviously we don't want to drive a driver so hard that occurs. So the speaker system has to have an adequate power rating (and sensitivity).

 

Theoretically, at sustained high driving levels a driver's voice coil can heat up sufficiently that its resistance increases significantly leading to a temporary drop in the driver's sensitivity.

 

 

Resolution

This isn't so readily explained or defined!  Perhaps it just means clarity.

 

It may relate to low harmonic and intermodulation distortion.  It may relate to less colouration in the frequency response.  It may relate to an evenness in the off-axis frequency response that helps with the integrity of the stereo image.

 

As regards unevenness in the frequency response, this could be addressed to some extent with Digital Signal Processing. 

  • Like 2
Link to comment
Share on other sites



6 minutes ago, MLXXX said:

Those two terms are bandied about a lot on audiophile forums. What do they really mean in relation to a speaker enclosure, its drivers, and its crossover (if present and if used)?

 

Dynamics

At very high driving levels a speaker cone can reach the limit of its excursion forwards or backwards resulting in a flattening of the waveform, not dissimilar to amplifier clipping,. Obviously we don't want to drive a driver so hard that occurs. So the speaker system has to have an adequate power rating (and sensitivity).

 

Theoretically, at sustained high driving levels a driver's voice coil can heat up sufficiently that its resistance increases significantly leading to a temporary drop in the driver's sensitivity.

 

 

Resolution

This isn't so readily explained or defined!  Perhaps it just means clarity.

 

It may relate to low harmonic and intermodulation distortion.  It may relate to less colouration in the frequency response.  It may relate to an evenness in the off-axis frequency response that helps with the integrity of the stereo image.

 

As regards unevenness in the frequency response, this could be addressed to some extent with Digital Signal Processing. 

Fair enough, I stand corrected.

 

I am enthusiastic supporter for calling out BS Audiophile Lore, so I don't want to part of the problem! ūüôā

  • Like 1
Link to comment
Share on other sites

3 hours ago, MLXXX said:

Those two terms are bandied about a lot on audiophile forums. What do they really mean in relation to a speaker enclosure, its drivers, and its crossover (if present and if used)?

 

Dynamics

At very high driving levels a speaker cone can reach the limit of its excursion forwards or backwards resulting in a flattening of the waveform, not dissimilar to amplifier clipping,. Obviously we don't want to drive a driver so hard that occurs. So the speaker system has to have an adequate power rating (and sensitivity).

 

Theoretically, at sustained high driving levels a driver's voice coil can heat up sufficiently that its resistance increases significantly leading to a temporary drop in the driver's sensitivity.

 

 

Resolution

This isn't so readily explained or defined!  Perhaps it just means clarity.

 

It may relate to low harmonic and intermodulation distortion.  It may relate to less colouration in the frequency response.  It may relate to an evenness in the off-axis frequency response that helps with the integrity of the stereo image.

 

As regards unevenness in the frequency response, this could be addressed to some extent with Digital Signal Processing. 

From JG Holt:

definition (also resolution) That quality of sound reproduction which enables the listener to distinguish between, and follow the melodic lines of, the individual voices or instruments comprising a large performing group.

 

'dynamics' is a very wide term which avoids definition, but JGH had this to say about a couple of items:

dynamic Giving an impression of wide dynamic range; punchy. This is related to system speed as well as to volume contrast.

dynamic range 1) Pertaining to a signal: the ratio between the loudest and the quietest passages. 2) Pertaining to a component: the ratio between its no-signal noise and the loudest peak it will pass without distortion.

Link to comment
Share on other sites

1 hour ago, Cloth Ears said:

Giving an impression of wide dynamic range; punchy. This is related to system speed ...

I have seen references to system speed, and perhaps we think of amplifier slew rate as an aspect of it; though modern power amplifiers typically have no trouble with slew rate.

 

With speakers if there is a high Q as might be the case for a bass driver around the resonant frequency of the enclosure, it can take a little while for a signal from the amplifier at that resonant frequency to result in a build up to maximum intensity of SPL in the speaker enclosure, and a while for that sound pressure level in the enclosure to decay once the driving signal from the amplifier at around the resonant frequency has ceased.

 

DSP can tame a really prominent peak around a resonant bass frequency (a boomy bass) by reducing the drive level around that frequency but that would not overcome the technical delay in achieving the maximum speaker enclosure intensity  of a sustained low frequency note in the music coming from the power amplifier.  (Whether such delays are audible in themselves as delays would be another question.   Is a boomy bass merely a prominent bass, or is it a bass that audibly lingers?)

Edited by MLXXX
Link to comment
Share on other sites

On 8/7/2023 at 12:25 PM, tripitaka said:

 

In any case, speakers can be DSP'd at any future time.!

The problem is that the speaker response varies as you change the angle ‚Ķ. And EQ doesn‚Äôt change this‚Ķ. So EQ cannot ‚Äúfix‚ÄĚ a bad speaker design ‚Ķ.. only ‚Äúmodify the problem‚ÄĚ.¬†

  • Like 1
Link to comment
Share on other sites

7 hours ago, MLXXX said:

.   Is a boomy bass merely a prominent bass, or is it a bass that audibly lingers?)

They are the same thing…. From an objective perspective.  

 

if you take a speaker with a flat response, and the apply some ‚ÄúEQ‚ÄĚ to make a peak in the response (doesn‚Äôt have to be ‚Äúelectronic‚ÄĚ EQ ‚Ķ. It could be anything which makes th peak, moving seats, adding a port, changing the cabinet, a capacitor, anything)

 

‚Ķ. Then when you look at the time response it also ‚Äúlingers‚ÄĚ (lasts longer). ¬†
 

when you alter the frequency response, you also alter the phase response.  
 

so for example a peak in the bass response, is also a ringing in the time response.  

Link to comment
Share on other sites



11 hours ago, davewantsmoore said:
18 hours ago, MLXXX said:

.   Is a boomy bass merely a prominent bass, or is it a bass that audibly lingers?)

They are the same thing…. From an objective perspective.  

Yes. However, I was interested in whether the bass would be likely to seem to be delayed, subjectively. Perhaps pursuing that query of mine would be an unnecessarily specific  diversion for this thread. 

 

I do find the audiophile term "dynamic" a bit hard to pin down. My thoughts about the subjective nature of the delay of a boomy bass were prompted by this particular quoted definition and its reference to speed:

19 hours ago, Cloth Ears said:

dynamic Giving an impression of wide dynamic range; punchy. This is related to system speed as well as to volume contrast.

 

Link to comment
Share on other sites

1 hour ago, MLXXX said:

Yes. However, I was interested in whether the bass would be likely to seem to be delayed, subjectively.

In my own head, I answered that (at least indirectly) .... let me elaborate.

 

In a speaker.....  If the frequency response is flat.... then the time response is also "flat" (not delayed or advanced).

 

So... no matter what the cause of the peak.... or the solution to the peak  (whether it be the Q of the bass driver.... the output loading of the amplifier.... electronic EQ .... the design of the port)  if the resulting response is flat, then the time response is correct.

 

So, if you have a "high-Q" bass driver, and it results in a peak in the SPL.... yes, you will have a "lingering" time response.    The distortion extends both in amplitude and time...... but if you have the same high-Q bass driver, it is conspired with all the other things (cabinet, filters, ports, whatever) to produce a flat amplitude response.... then there is no time distortion.

 

So... depending on the overall setup, will dictate whether there is any time distortion or not.

 

Then, the remaining issue (assuming, "yes it exists") ... is obviously can you hear it, or not.

 

 

20 hours ago, MLXXX said:

DSP can tame a really prominent peak around a resonant bass frequency (a boomy bass) by reducing the drive level around that frequency but that would not overcome the technical delay in achieving the maximum speaker enclosure intensity  of a sustained low frequency note in the music coming from the power amplifier. 

No.  This is wrong.    (ie.  "yes, it does/can overcome it")

 

If the amplitude error is corrected (by whatever means, mechanical, acoustical, electronic, whatever) .... then the delay (in the output, which is what you listen to) is corrected.

Said another way.... when you correct the peak, you are correcting the delay.    Because they are (literally) the same thing.

 

1 hour ago, MLXXX said:

Perhaps pursuing that query of mine would be an unnecessarily specific  diversion for this thread. 

Perhaps <shrug> ... I hope I'm not being unnecessarily verbose (these topics are hard to answer well in 25 words)

 

An important caveat to all the above is "in a speaker".    In a speaker, the amplitude and phase are 2-sides to the same coin (to attributes of the same event) ..... but in a room, the reflections / diffraction do not necessarily have the amplitude/phase behaviour.

 

This is why "phase correction" (aka. linear phase, FIR, convolves, etc.) in a speaker is not very useful, and can also be misused to correct for "illusions" in our measurements ...... but in a room, ie. "room correction" then it can help the overall response if done carefully.

 

This is the difference between some "room" corrections systems and others.... some assume that all phase and amplitude are UN-related (ie. they can be fixed independently) .... and others do not necessarily assume this (they try to determine, if that is the case).

 

1 hour ago, MLXXX said:

I do find the audiophile term "dynamic" a bit hard to pin down. My thoughts about the subjective nature of the delay of a boomy bass were prompted by this particular quoted definition and its reference to speed:

High-q (steep, sharp) deviations in the amplitude response... depend a lot on the content (ie. do notes or sounds happen at those frequencies).

 

If you audition a much lower Q peak, or shelf, to give more vs less bass. (eg. +4dB vs -4dB to everything below 150Hz, or whatever) ..... the subjective effects are quite interesting.     "faster, tighter, clearer, more dynamic" ..... vs "muddy, thick, slow, etc".     It does depend a lot on the content too, and what people are used to, or enjoy.

Link to comment
Share on other sites

1 hour ago, davewantsmoore said:
22 hours ago, MLXXX said:

DSP can tame a really prominent peak around a resonant bass frequency (a boomy bass) by reducing the drive level around that frequency but that would not overcome the technical delay in achieving the maximum speaker enclosure intensity  of a sustained low frequency note in the music coming from the power amplifier. 

No.  This is wrong.    (ie.  "yes, it does/can overcome it")

Yes it could be achieved, at the expense of pre-ringing.   With DSP you could delay the rest of the music signal delivered to the speaker system, to give the SPL at the bass driver resonant frequency (as modified by the enclosure) an opportunity to build up. 

 

So for example if a piece of music began with a sustained chord from a pipe organ, and if a low note in the chord coincided with the bass resonant frequency of the bass driver in the speaker enclosure, the beginning of low note as delivered by the audio power amplifier could be brought forward in time to compensate for the sluggish response of the driver.  In the extreme case of an enclosure with an unusually high Q, it might even be possible for a listener to hear that the beginning of that particular note was sounding early; though I think that the onset of a pipe organ low note is gradual and so the effect of bringing the beginning of that note forward in time could go unnoticed.

 

High amplitude low frequency percussion, with a quick rise time, would be more demanding.

Edited by MLXXX
Link to comment
Share on other sites

1 hour ago, MLXXX said:

Yes it could be achieved, at the expense of pre-ringing. 

No... this isn't necessarily so.

A correction filter that corrects a peak in the response (a minimum phase filter) does not inherently have any pre-ringing.   So if you got any, then it would be a mistake by the filter creator.

 

1 hour ago, MLXXX said:

With DSP you could delay the rest of the music signal delivered to the speaker system, to give the SPL at the bass driver resonant frequency (as modified by the enclosure) an opportunity to build up. 

You either haven't been understanding what I have been typing, or you just doin't believe me.

 

Imagine that the peak was 6dB.   You take away that peak, with a -6dB filter (the amplitude response is now flat).    Now, there is no time distortion remaining.... there is no "building up", as you put it.    You do not "delay the rest of the music signal delivered to the speaker system" .... you just correct the peak so it is flat, and then all the time distortion is gone too.

 

1 hour ago, MLXXX said:

So for example if a piece of music began with a sustained chord from a pipe organ, and if a low note in the chord coincided with the bass resonant frequency of the bass driver in the speaker enclosure, the beginning of low note as delivered by the audio power amplifier could be brought forward in time to compensate for the sluggish response of the driver.

The drive response is only sluggish, because it does not have a flat frequency response.    If you flatten its frequency response, then it has a "perfect" time response.

 

1 hour ago, MLXXX said:

In the extreme case of an enclosure with an unusually high Q

An enclosure with an unusually high Q.... has an unusually large peak in the bass response.    So instead of a (for example) 6dB filter.... you might need 9dB.... whatever "EQ" is was to get the response flat.

 

Note... the "EQ to flatten the response" doesn't have to be "DSP".   It could be capacitor, a different sized port or box, sticking bits of felt to the cone.... anything.   One the peak has beeb flattened, the time response is also corrected.

Link to comment
Share on other sites

2 minutes ago, davewantsmoore said:

No... this isn't necessarily so.

A correction filter that corrects a peak in the response (a minimum phase filter) does not inherently have any pre-ringing.   So if you got any, then it would be a mistake by the filter creator.

 

In the graph below is an example of the effect I had in mind. Assume an electronic percussion instrument has provided 5 cycles of a 100 Hz sine wave.  (If you listen to 5 cycles of sine wave at 100Hz, it does sound percussive.)

 

Assume for the sake of argument that the bass driver in the enclosure has a bass resonant frequency of 100Hz and requires 6.1dB of attenuation to smooth out the overall lower frequency response of the speaker system.

 

It will be seen that the default graphic equalizer function of the wave editor, Audacity, introduces low level pre-ringing and post-ringing in the processed signal.

 

 

What filter would you recommend to avoid the pre-ringing?

 

100HzgraphicequalWithText.jpg.ce62d20d3ad1b243e6096bfbe0386c2d.jpg

 

Link to comment
Share on other sites



2 hours ago, MLXXX said:

Example

Your example is kinda "backwards" from the real world.

 

In the real world, you will start with a signal that has the unwanted thing.    Say a peak in the amplitude (and 'ringing' in time)

 

You apply the filter which removes the amplitude error (which also solves the time error) .... and the after chart, shows the correct signal (there will be no pre/post ringing or any other error).

Link to comment
Share on other sites

2 hours ago, MLXXX said:

What filter would you recommend to avoid the pre-ringing?

 

Start with your before signal.

Add 6dB boost (this is your "bass resonant frequency" error)

Apply the opposite 6db cut.

The result will be the "before" signal.

Link to comment
Share on other sites

Hey Dave, where do cabinet resonances fit into this scenario? ¬†You know, the waterfall plot shows a ‚Äúmountain range‚ÄĚ delay at a certain frequency due to poor cabinet design or construction. ¬† Can this be ameliorated by manipulating the FR or is it similar to a room effect?

Link to comment
Share on other sites

2 hours ago, davewantsmoore said:

Your example is kinda "backwards" from the real world.

 

In the real world, you will start with a signal that has the unwanted thing.    Say a peak in the amplitude (and 'ringing' in time)

 

I am certainly no expert in the realm of DSP design for speaker system correction. However, it seems to me that the example I have presented is quite ordinary and "real world": a speaker system with boomy bass attributable to resonance.

 

To reiterate details: in my example, the speaker system when fed a sine wave test signal slowly varying in frequency produces a sound pressure level peak of 6.1 dB centred on 100Hz [because of resonance].   This would give it a subjectively boomy bass when reproducing music, if left untreated.

 

If we are to eliminate the peak at  100Hz in the speaker system frequency response only using DSP, we will need to attenuate the steady state signal amplitude sent to the speaker system by 6.1 dB at 100Hz (and by lesser amounts near 100Hz).

 

Using a standard graphic equalizer (such as offered by the freeware Audacity), we can achieve (1) a 6.1 dB reduction centred on 100Hz, and (2) have the centre of the processed waveform align in is timing with the centre of the original waveform.  However as illustrated in my graph, doing so has led to pre-ringing and post-ringing for the signal to be sent to the speaker system. 

 

Even though the speaker system is sluggish around 100Hz it would nevertheless emit as sound  some of that pre-ringing signal.  (I note that the pre-ringing artefact  sent to the speaker system may well not be audible of itself for the human ear.)

 

I am not aware of a DSP fiter design that will avoid pre-ringing artefacts for the very specific example I have described here.  I chose that example at random. Other resonant frequencies could be used (e.g. 80Hz).  However as this is not a subject area in which I'm knowledgeable, I'd welcome suggestions as to how DSP could possibly do this job of correcting the boomy bass, without introducing ringing artefacts in the signal sent to the loudspeaker system. 

 

 

Link to comment
Share on other sites

8 hours ago, MLXXX said:

I am certainly no expert in the realm of DSP design

What we have been discussing is not a "DSP thing".

 

The "correction filters" I am talking about could be done using anything.   The could be done by connecting resistors, capacitors and inductors .... or by placing a blanket over the front of the speaker.... or by sitting in a different location... or by anything.    Anything which corrects the response.

 

8 hours ago, MLXXX said:

for speaker system correction. However, it seems to me that the example I have presented is quite ordinary and "real world": a speaker system with boomy bass attributable to resonance.

Yes, that is a real world example (it is the example we have been discussing) .... you just drew the picture of it wrong (in your 100Hz wave chart).

 

You drew a chart where the "after" chart has "pre ringing".     In reality (in your example), it is the before chart which has the ringing..... you apply a filter to correct the amplitude error (which also corrects the ringing) and then the "after" chart looks like the "perfect 100Hz wave".

 

8 hours ago, MLXXX said:

To reiterate details: in my example, the speaker system when fed a sine wave test signal slowly varying in frequency produces a sound pressure level peak of 6.1 dB centred on 100Hz [because of resonance].   This would give it a subjectively boomy bass when reproducing music, if left untreated.

Yes.... so the picture of this, looks like your "after" chart that you posted.

 

8 hours ago, MLXXX said:

If we are to eliminate the peak at  100Hz in the speaker system frequency response only using DSP, we will need to attenuate the steady state signal amplitude sent to the speaker system by 6.1 dB at 100Hz (and by lesser amounts near 100Hz).

Yes.

 

8 hours ago, MLXXX said:

Using a standard graphic equalizer (such as offered by the freeware Audacity), we can achieve (1) a 6.1 dB reduction centred on 100Hz, and (2) have the centre of the processed waveform align in is timing with the centre of the original waveform. 

Yes... if you do that, you will get a chart that looks like your "before" chart.

 

8 hours ago, MLXXX said:

However as illustrated in my graph, doing so has led to pre-ringing and post-ringing for the signal to be sent to the speaker system. 

No, see above.  You did it "backwards".

 

8 hours ago, MLXXX said:

only using DSP

All "filters" are the same.    Thinking about it in terms of "only DSP" will, if anything, hinder understanding the topic.

 

8 hours ago, MLXXX said:

Even though the speaker system is sluggish around 100Hz it would nevertheless emit as sound  some of that pre-ringing signal.  (I note that the pre-ringing artefact  sent to the speaker system may well not be audible of itself for the human ear.)

After it is corrected .... there will be no "ringing" ..... both the amplitude and phase will be corrected (see above, you did it wrong).

 

Prior to the correction.... yes, there is the time distortion.   Any systems with a frequency response that is not flat has time distortion like this.

 

 

8 hours ago, MLXXX said:

I am not aware of a DSP fiter design that will avoid pre-ringing artefacts for the very specific example I have described here.

We need to clear, that you worked your example wrong (but you don't appear to be listening to me when I tell you that).

 

BUT.... if you hypothetically had a result (I'm talking about a different hypothetical example now, not yours) which did have such ringing in it .... then you can correct this using DSP.    You just need a filter which does not alter the amplitude, but does alter the time (only DSP filters can do this).

 

 

8 hours ago, MLXXX said:

However as this is not a subject area in which I'm knowledgeable, I'd welcome suggestions as to how DSP could possibly do this job of correcting the boomy bass, without introducing ringing artefacts in the signal sent to the loudspeaker system. 

I just told you, 5 times.

Edited by davewantsmoore
Link to comment
Share on other sites



  • Recently Browsing ¬†¬†0 members

    • No registered users viewing this page.
√ó
√ó
  • Create New...
To Top