Speakers distortion

[Krunok]

Obviously the situation with speakers distortion is far from ideal. Although speakers produce by far the most distortion in the reporoduction chain it seems there are no standards in how to measure their distortion. As far as I was able to see many manufacturers don't specify any distortion data (Martin Logan, Revel, ..) not even for their flagship products and those that do don't do it in a consistent manner.

On the reviewers/testing side, only few of them (SoundStage, ..?) measure distortion when reviewing speakers but that is also done in a so rudimentary way that doesn't really come close to paint the whole pictures (see @Blumlein 88 post for Performa F206 few pages before).

While DACs and amps are being put to some very detailed distortion measurements, although most of the modern ones don't really come close to being problematic, situation with speakers is totally opposite. And yet, market doesn't seem to care. Users on this forum will write hundreds of posts discussing the difference between 0.000X% and 0.000Y% of THD with DACs and amps but when it comes to speakers not even remotely so.

Strange, isn't it..?

But then, situation is similar with room EQ. Although speakers linearity (unlike distortion) can significantly be enhanced with room EQ it seems not many users on this forum have done it, in spite of the fact that it is the most signifacnt thing with which SQ at our homes can be improved.

Is it because most audiophiles are not aware how important speakers perfomance is in the reproduction chain or for some other reason it remains a mistery to me.

 

[DonH56]

Although speakers linearity (unlike distortion) can significantly be enhanced with room EQ it seems not many users on this forum have done it, in spite of the fact that it is the most signifacnt thing with which SQ at our homes can be improved.

Note that deviation from linearity, i.e. nonlinear behavior, is distortion. Room EQ can change the frequency response to match the desired response (which may or may not be flat -- usually is not) but cannot correct nonlinearity in the speaker. The peaks and valleys of in-room frequency response can be considered "distortion" of the frequency response but are not nonlinear "classical" distortion.

 

[Cosmik]

Is it because most audiophiles are not aware how important speakers perfomance is in the reproduction chain or for some other reason

I honestly don't think that's universally true. What I think is the case is that much of the physics of speakers and acoustics is driven by 'laws of nature' and so there are no huge surprises out there. There's no point in going for 0.001% distortion because it probably won't be possible without making some other compromise that damages the sound in some other way (e.g. you might get very low distortion by reducing maximum displacement of the cone, but to do this you might need to make the cone area bigger than it ideally should be for other acoustic reasons).

Pre-distortion using DSP might reduce driver distortion, but with the risk that manufacturing tolerances, materials ageing, etc. result in mis-matching artefacts that sound worse than the unmodified driver.

 

[Blumlein 88]

Obviously the situation with speakers distortion is far from ideal. Although speakers produce by far the most distortion in the reporoduction chain it seems there are no standards in how to measure their distortion. As far as I was able to see many manufacturers don't specify any distortion data (Martin Logan, Revel, ..) not even for their flagship products and those that do don't do it in a consistent manner.

On the reviewers/testing side, only few of them (SoundStage, ..?) measure distortion when reviewing speakers but that is also done in a so rudimentary way that doesn't really come close to paint the whole pictures (see @Blumlein 88 post for Performa F206 few pages before).

While DACs and amps are being put to some very detailed distortion measurements, although most of the modern ones don't really come close to being problematic, situation with speakers is totally opposite. And yet, market doesn't seem to care. Users on this forum will write hundreds of posts discussing the difference between 0.000X% and 0.000Y% of THD with DACs and amps but when it comes to speakers not even remotely so.

Strange, isn't it..?

But then, situation is similar with room EQ. Although speakers linearity (unlike distortion) can significantly be enhanced with room EQ it seems not many users on this forum have done it, in spite of the fact that it is the most signifacnt thing with which SQ at our homes can be improved.

Is it because most audiophiles are not aware how important speakers perfomance is in the reproduction chain or for some other reason it remains a mistery to me.

I've the same query. Speakers are both where the rubber meets the road and by an order of magnitude the biggest issue in sound quality. Testing speakers is hard, but it's where real difference lies.

 

[Krunok]

Note that deviation from linearity, i.e. nonlinear behavior, is distortion. Room EQ can change the frequency response to match the desired response (which may or may not be flat -- usually is not) but cannot correct nonlinearity in the speaker. The peaks and valleys of in-room frequency response can be considered "distortion" of the frequency response but are not nonlinear "classical" distortion.

I wouldn't really consider those peaks and valleys distortion as they would affect the base tone as well. Distortion is characteristic of the speaker and those peaks and valleys are "conditions" in which speaker is playing.

 

[andreasmaaan]

Obviously the situation with speakers distortion is far from ideal. Although speakers produce by far the most distortion in the reporoduction chain it seems there are no standards in how to measure their distortion. As far as I was able to see many manufacturers don't specify any distortion data (Martin Logan, Revel, ..) not even for their flagship products and those that do don't do it in a consistent manner.

On the reviewers/testing side, only few of them (SoundStage, ..?) measure distortion when reviewing speakers but that is also done in a so rudimentary way that doesn't really come close to paint the whole pictures (see @Blumlein 88 post for Performa F206 few pages before).

While DACs and amps are being put to some very detailed distortion measurements, although most of the modern ones don't really come close to being problematic, situation with speakers is totally opposite. And yet, market doesn't seem to care. Users on this forum will write hundreds of posts discussing the difference between 0.000X% and 0.000Y% of THD with DACs and amps but when it comes to speakers not even remotely so.

Strange, isn't it..?

But then, situation is similar with room EQ. Although speakers linearity (unlike distortion) can significantly be enhanced with room EQ it seems not many users on this forum have done it, in spite of the fact that it is the most signifacnt thing with which SQ at our homes can be improved.

Is it because most audiophiles are not aware how important speakers perfomance is in the reproduction chain or for some other reason it remains a mistery to me.

One reason is that, although driver nonlinear distortion can be measured relatively accurately with a fairly basic setup, accurately measuring nonlinear distortion of a whole loudspeaker requires an anechoic chamber. To my knowledge, of the main review sites that post measurements, Soundstage is the only one with access to one (the NRC's anechoic chamber).

Other mags/sites could try a number of techniques to measure loudspeaker distortion, for example by wiring up only one driver at a time (but leaving the crossover in place), then measuring at a sensible distance with the speaker as far from room boundaries as possible, then splicing together the measurements from all the different drivers. But this would still only give an approximation which might be considered unfair to the manufacturer, despite being useful to the consumer.

 

[Krunok]

[QUOTE="Cosmik, post: 133042, member: 101"There's no point in going for 0.001% distortion because it probably won't be possible without making some other compromise that damages the sound in some other way (e.g. you might get very low distortion by reducing maximum displacement of the cone, but to do this you might need to make the cone area bigger than it ideally should be for other acoustic reasons).[/QUOTE]

It is true with todays technology. However, if we radically change the technology so the sound is produced by "energising" the air within the space so it starts to resonate (say like molecules of water in microwave owen, but in broafer spectrum) then it would be very improtant to reduce distortion in order to achieve better SQ than with todays speakers which shift the air with mechanical vibrations.

 

[Krunok]

One reason is that, although driver nonlinear distortion can be measured relatively accurately with a fairly basic setup, accurately measuring nonlinear distortion of a whole loudspeaker requires an anechoic chamber. To my knowledge, of the main review sites that post measurements, Soundstage is the only one with access to one (the NRC's anechoic chamber).

Other mags/sites could try a number of techniques to measure loudspeaker distortion, for example by wiring up only one driver at a time (but leaving the crossover in place), then measuring at a sensible difference with the speaker as far from room boundaries as possible, then splicing together the measurements from all the different drivers. But this would still only give an approximation which might be considered unfair to the manufacturer, despite being useful to the consumer.

Yet, the THD diagram of the Revel F206 produced by Soundstage that @Blumlein 88 posted is pathetic as it doesn't show anything.

I still believe that manufacturers, at least ones that are making expensive speakers, should provide distortion data for their products.

 

[Krunok]

I've the same query. Speakers are both where the rubber meets the road and by an order of magnitude the biggest issue in sound quality. Testing speakers is hard, but it's where real difference lies.

Cannot agree more.

Btw, I'm trully impressed with your car analogy!

 

[Cosmik]

It is true with todays technology. However, if we radically change the technology so the sound is produced by "energising" the air within the space so it starts to resonate (say like molecules of water in microwave owen, but in broafer spectrum) then it would be very improtant to reduce distortion in order to achieve better SQ than with todays speakers which shift the air with mechanical vibrations.

It reminds me of the plasma speaker - great in theory maybe, but with the disadvantage of producing toxic gas and/or needing special tanks of expensive gas to run.

The primitive mechanical version of a speaker still sounds damn good!

 

[Krunok]

It reminds me of the plasma speaker - great in theory maybe, but with the disadvantage of producing toxic gas and/or needing special tanks of expensive gas to run.

Ok, let's forget about shifting the air as it really complicates things. And even if you solve it somehow you still have problem with ears transforming the acoustic signal back to electric impulses in a not so perfect way. So why not skip that part and try to imagine DAC that would connect wirelessly to a converter under your skin which will be directly connected to your auditory vestibular nerve. Solve that and you immediately have perfect SQ!

The primitive mechanical version of a speaker still sounds damn good!

IMHO even the best of them sound barely acceptible.

 

[Cosmik]

So why not skip that part and try to imagine DAC that would connect wirelessly to a converter under your skin which will be directly connected to your auditory vestibular nerve. Solve that and you immediately have perfect SQ!

I agree, but it would be like wearing stereo headphones - unless you had a head tracker and software that could at least simulate the way speakers give you a stable source direction with head turning. Speakers do several things brilliantly:

• They allow you to turn your head and the image remains in the same place, just like a real performance.
• They allow you to move around to a more or less limited extent.
• They allow you to physically feel the acoustic power, just like a real performance
• They blend your voice and other sounds with the performance in that you and the recording both share the same acoustics (of the listening room). It's very natural.
• They allow more than one person to hear the performance and talk to each other at the same time, like a real performance

 

[Krunok]

I agree, but it would be like wearing stereo headphones - unless you had a head tracker and software that could at least simulate the way speakers give you a stable source direction with head turning. Speakers do several things brilliantly:

• They allow you to turn your head and the image remains in the same place, just like a real performance.
• They allow you to move around to a more or less limited extent.
• They allow you to physically feel the acoustic power, just like a real performance
• They blend your voice and other sounds with the performance in that you and the recording both share the same acoustics (of the listening room). It's very natural.
• They allow more than one person to hear the performance and talk to each other at the same time, like a real performance

Once we master the technology to interact directly with our sensory system we will be able to do everything that todays speakers do, and much more.

But I would rather that we establish here at least the distortion limits todays speakers have, but not only the ones in the price range of $5000-$25000, but also the more affordable ones. Unfortunately, in order to do that we would need to have measurements, but not many folks seems to be interested in spending $100 for the mic and spend few days to learn how to do measurements in REW, which would also come very handy for doing room EQ.

 

[sergeauckland]

Once we master the technology to interact directly with our sensory system we will be able to do everything that todays speakers do, and much more.

But I would rather that we establish here at least the distortion limits todays speakers have, but not only the ones in the price range of $5000-$25000, but also the more affordable ones. Unfortunately, in order to do that we would need to have measurements, but not many folks seems to be interested in spending $100 for the mic and spend few days to learn how to do measurements in REW, which would also come very handy for doing room EQ.

I've stopped contributing to other audio forums for just the reason you mention. People seem happy to spend many hundreds on cables, fuses even!, but won't spend a little on any test equipment that would allow them to understand exactly what their kit was doing. I know it takes a bit of effort to understand measurements, but no more than trying to understand cables or all the other nonsenses this hobby is cursed with.

S

 

[Dialectic]

Once we master the technology to interact directly with our sensory system we will be able to do everything that todays speakers do, and much more.

But I would rather that we establish here at least the distortion limits todays speakers have, but not only the ones in the price range of $5000-$25000, but also the more affordable ones. Unfortunately, in order to do that we would need to have measurements, but not many folks seems to be interested in spending $100 for the mic and spend few days to learn how to do measurements in REW, which would also come very handy for doing room EQ.

I'd be interested to see a tutorial on how to place a mic in and around the listening position to take measurements in REW. Mic placement is glossed over in the tutorials I've seen, so if any ASR member knows of one that covers mic placement thoroughly, I'd be grateful for a link.

I'm using a Behringer measurement mic but may splurge for an Earthworks in the future. No audiophile cable purchases for me...

 

[SIY]

I'd be interested to see a tutorial on how to place a mic in and around the listening position to take measurements in REW.

That's one of the strengths of Sonarworks- it's all automated. And IME, just about foolproof.

 

[Krunok]

I'd be interested to see a tutorial on how to place a mic in and around the listening position to take measurements in REW. Mic placement is glossed over in the tutorials I've seen, so if any ASR member knows of one that covers mic placement thoroughly, I'd be grateful for a link.

I'm using a Behringer measurement mic but may splurge for an Earthworks in the future. No audiophile cable purchases for me...

I suggest you read this article, it will give you good idea what it is all about.

Regarding the moving patterns around your LP, maybe you can start with the ones DiracLive is using:



Use RTA in REW with pink noise file. Move your mic from point to point in a slow steady speed and try to make at least 60-70 averages. Use it for corrections up to 600Hz. Good luck!

Edit: you can use the same points for log sine sweep. Average them after you made them. The results you'll get with both methods should be very similar.

 

[Blumlein 88]

I don't have a single link. But the home theater shack forum that has REW has covered almost any topic you can think of.

 

[Cosmik]

Once we master the technology to interact directly with our sensory system we will be able to do everything that todays speakers do, and much more.

Well you can get a fair impression of this new world by using in-ear monitors, head tracking systems, computer game audio synthesis etc. The main limitation isn't a bit of distortion in the transducers, nor how the audio gets to your brain, but the recording/synthesis of the material you're listening to.

You can't record real acoustics such that you can reproduce them perfectly later, except from one position in space. But in the computer games industry they synthesise the acoustics you would be hearing if certain 'dry' sources were making sounds in the intended virtual environment with its layout, dimensions and materials. This allows you to move around, turn your head, etc. and, because of head tracking and on-the-fly DSP, to hear what you would be hearing in this synthetic space.

But so what? If it's absolutely amazing, then that's how music might be created in future - but I don't see much demand for it at the moment. And that's because I think the alternative that we have now is not only good enough, but actually optimal. It's sociable, natural, and - to my ears - sounds great when done properly. It isn't even as if you're aware of the speakers with mere stereo.

In the computer games industry, they can go even further and give you the visuals as well, using VR headsets. And is it popular? No. It's a novelty that people find boring and unpleasant pretty quickly even though it is genuine, futuristic science fiction from the future! It's about as popular as the videophone that I was so desperate to get in years gone by, but now that I have it in my pocket I never use...

 

[RayDunzl]

Why isn't the midrange driver similar in design to high freq driver, like the ones Dynaudio used long time ago? These days all of manufacturers for midrange driver seem to use similar design to bass driver, but smaller. Why is that?

I don't think this was answered...

Assuming HF driver = a little dome, and the mid and low are cones...

The little dome is driven on its edges - it's small enough to do so.

Scale that up a few inches and it might get pretty big.

Mid-cones are driven (typically) closer to their center, 1"~2" or so for the size of the voice coil.

Edge driving a larger driver - I suppose you'd run into various difficulties with weight/magnet size/winding density, etc.

I assume it has been done, also assume others didn't see the results equal the effort.

Of course it has been done -- https://www.google.com/search?q=dom...cbfAhULh1QKHZgQCf8Q_AUIDygC&biw=1601&bih=1205

I don't see any domed woofers though...