Speakers distortion

[Blumlein 88]

What would be the point fo that as I'm moving the mic during measurement? Wavelength of 2kHz tone is app 17cm so I would encounter tops and downs many time during my measurement.

Besides, you own the mic too so why don't you try to make some meaurements and see what you will get?

Because all my hours are currently spent at the hospital with an ailing relative. So most of my recent posting is from a phone.

I did take a couple hours the other night trying things and decided on room distortion isn't straight forward to seeing what actual distortion is. Sorry.

If my proposal can be made to work and it gets nearly the same result as mmm then at least you'll have corroborated it. When I get time I'll try it.

 

[Krunok]

Because all my hours are currently spent at the hospital with an ailing relative. So most of my recent posting is from a phone.

I did take a couple hours the other night trying things and decided on room distortion isn't straight forward to seeing what actual distortion is. Sorry.

If my proposal can be made to work and it gets nearly the same result as mmm then at least you'll have corroborated it. When I get time I'll try it.

Oh, I see - sorry to hear that.

Frankly, I'm probably missing something but I don't understand what's the big difference between measuring frequency response and distortion. Both of them are actually frequency response except in the distortion case those additional tones are not played but are induced by speaker and they are quite silent. But I really don't understand if RTA MMM works best for frequency response why it wouldn't work best for distortion as well.

 

[Krunok]

I made 5 measurements of left speaker at 80dB at 120Hz, RTA with MMM. They all gave distortion in the range (0,20 - 0,23).

This is the snapshot of 6th measurement which I didn't finish as phone rang, but as you can see it's pretty much made of 2nd and 3rd harmonics while others are near to be burried in the noise floor.



If this is correct and my speakers really have THD in the range of 0,2% than my tube amp is dangerously messing with the sound they produce as it also has THD of 0,2%.

P.S. I picked up 120Hz instead of 100 to avoid noise of home appliances

 

[sergeauckland]

Oh, I see - sorry to hear that.

Frankly, I'm probably missing something but I don't understand what's the big difference between measuring frequency response and distortion. Both of them are actually frequency response except in the distortion case those additional tones are not played but are induced by speaker and they are quite silent. But I really don't understand if RTA MMM works best for frequency response why it wouldn't work best for distortion as well.

Because when measuring an in-room response you want to measure the effect of the room overlaid on the loudspeaker's anechoic response including the effect of reflections. When measuring distortion you (presumably!) want to measure the distortion of the loudspeaker only, not the distortions caused by the room's response. As mentioned above, if the microphone is in a null or peak, the measured distortion will have nothing to do with distortion of the loudspeaker and it will change with microphone position.

S.

 

[Krunok]

Because when measuring an in-room response you want to measure the effect of the room overlaid on the loudspeaker's anechoic response including the effect of reflections. When measuring distortion you (presumably!) want to measure the distortion of the loudspeaker only, not the distortions caused by the room's response. As mentioned above, if the microphone is in a null or peak, the measured distortion will have nothing to do with distortion of the loudspeaker and it will change with microphone position.

S.

I do want to measure the distortion caused by the room's response as well as that is what I'm also hearing, isn't it?

Btw, as I already pointed out, as I'm moving mic over a relatively large area and taking more than 70 samples I believe nulls/peaks don't play any role, at least not at higher frequencies.

 

[DonH56]

I do want to measure the distortion caused by the room's response as well as that is what I'm also hearing, isn't it?

Btw, as I already pointed out, as I'm moving mic over a relatively large area and taking more than 70 samples I believe nulls/peaks don't play any role, at least not at higher frequencies.

I already suggested running a cal routine with each measurement... Sweep and measure at a low level, then repeat at higher level, using the scaled lower-level response to compensate for the influence of the room on the frequencies. The cal would be automatically performed as part of the sweep so as long as you do not move the mic during the sweep things should be OK. It will take longer, natch, and you will have the usual problem of separating signal from noise when the signal is small. And the low-level sweep will include whatever distortion is present through the signal chain (from SW to speakers to mic to SW), so if the distortion is not low enough (negligible), the measurement enough will include additional (uncompensated) distortion terms.

The room's response is not classical distortion, i.e. it is not a nonlinear effect that adds distortion tones not present in the original signal. The room will emphasize and de-emphasize signal frequencies that are already there (from the source). That is not the distortion you want to measure, you want to know how much distortion (nonlinearity) the speakers (etc.) add to the signal.

If you are averaging the results across readings then room response is being averaged as well so you are not getting the true response from the speakers themselves. That is actually going the opposite way -- you are making speaker distortion more difficult to isolate. I thought this thread was about speaker distortion; room response is a different subject.

 

[Krunok]

I already suggested running a cal routine with each measurement... Sweep and measure at a low level, then repeat at higher level, using the scaled lower-level response to compensate for the influence of the room on the frequencies. The cal would be automatically performed as part of the sweep so as long as you do not move the mic during the sweep things should be OK. It will take longer, natch, and you will have the usual problem of separating signal from noise when the signal is small. And the low-level sweep will include whatever distortion is present through the signal chain (from SW to speakers to mic to SW), so if the distortion is not low enough (negligible), the measurement enough will include additional (uncompensated) distortion terms.

The room's response is not classical distortion, i.e. it is not a nonlinear effect that adds distortion tones not present in the original signal. The room will emphasize and de-emphasize signal frequencies that are already there (from the source). That is not the distortion you want to measure, you want to know how much distortion (nonlinearity) the speakers (etc.) add to the signal.

If you are averaging the results across readings then room response is being averaged as well so you are not getting the true response from the speakers themselves. That is actually going the opposite way -- you are making speaker distortion more difficult to isolate. I thought this thread was about speaker distortion; room response is a different subject.

I don't see how distortion can possibly be measured in a room without taking into account rooms influence at it. It would be similar to measuring speakers in-room response while speakers are playing in the room but hoping you will somehow measure response without rooms influence.

I am trying to measure distortion I hear in my room with my speakers. I'm aware such distortion figure measured that way couldn't be compared to that same speakers measured in some other room. I'm also aware that if we want to compare distortion figures speakers need to be measured in anechoic chamber, but it seems manucfaturers don't do that.

But I really dont' see how can I measure distortion of my speakers playing in room without taking room into account, unless of course I redecorate my room to an anechoic chamber.

 

[sergeauckland]

I don't see how distortion can possibly be measured in a room without taking into account rooms influence at it. It would be similar to measuring speakers in-room response while speakers are playing in the room but hoping you will somehow measure response without rooms influence.

I am trying to measure distortion I hear in my room with my speakers. I'm aware such distortion figure measured that way couldn't be compared to that same speakers measured in some other room. I'm also aware that if we want to compare distortion figures speakers need to be measured in anechoic chamber, but it seems manucfaturers don't do that.

But I really dont' see how can I measure distortion of my speakers playing in room without taking room into account, unless of course I redecorate my room to an anechoic chamber.

What I don't understand is why you care about the in-room distortions, because as Don pointed out, what you're then measuring is NOT distortion. Distortion is what your loudspeakers do, regardless of location. That's what matters. That's also why you need to know the anechoic (or pseudo-anechoic) response of the loudspeakers before you measure the in-room response, whether at one place or averaged over many. If you don't know what the loudspeakers are doing intrinsically, what point is there in knowing the pretty random changes caused by a room? Bear in mind that the in-room response will change, sometimes radically, with small changes in loudspeaker or listening position, and even moving a piece of furniture will change the in-room response. The in-room response is a useful test of known good loudspeakers to optimise positioning of both loudspeakers and furnishings, but pretty pointless unless the loudspeakers are known to be good first.

S.

 

[DonH56]

I don't see how distortion can possibly be measured in a room without taking into account rooms influence at it. It would be similar to measuring speakers in-room response while speakers are playing in the room but hoping you will somehow measure response without rooms influence.

I am trying to measure distortion I hear in my room with my speakers. I'm aware such distortion figure measured that way couldn't be compared to that same speakers measured in some other room. I'm also aware that if we want to compare distortion figures speakers need to be measured in anechoic chamber, but it seems manucfaturers don't do that.

But I really dont' see how can I measure distortion of my speakers playing in room without taking room into account, unless of course I redecorate my room to an anechoic chamber.

You are missing my point. I agree you need to know the in-room frequency response to accurately measure the relative distortion levels, and as @sergeauckland said you need to know the response independent of the room or as I said must compensate for the room's response. The problem is you cannot measure the room without using your speakers to do it, so their distortion (plus whatever else is added by the signal chain, including the mic and preamp back into the PC) is added to (embedded in) the measured result. You need a scheme that separates speaker distortion from the room's response.

You can use manufacturer's or reviewer's measurements to provide a baseline for your speakers, which are often taken in an anechoic chamber or use a time-gated measurement to "take out" the room's effects.

Let me try again, then I have to get back to work: Send pink noise or a swept tone at a low level through the system and measure the response. Make it low enough that speaker distortion is negligible and record the response. Assume the ripples are due to the room and not the speakers, or at least not from speaker distortion (ripples due to drivers and such will be included but those are not nonlinear distortions). Now increase the level to the desired measurement level and repeat. SCale the room's response to the new power level and back out its effects.

Using numbers: Play a low-level set of tones at 1 kHz, 2 kHz, and 3 kHz. Say the 1 kHz is at 60 dB, 2 kHz at 66 dB (6 dB high), 3 kHz at 55 dB (5 dB low). Now play a 1 kHz tone at 90 dB and again check the 2 and 3 kHz terms (second and third harmonics). You must reduce the 2 kHz measurement by 6 dB and increase the 3 kHz level by 5 dB to determine (roughly) the actual second and third harmonic distortion terms independent of (compensating for) the in-room response. This assumes nothing changed except signal amplitude between readings. Repeat for all the other frequencies and positions desired.

HTH - Don

 

[Krunok]

What I don't understand is why you care about the in-room distortions, because as Don pointed out, what you're then measuring is NOT distortion. Distortion is what your loudspeakers do, regardless of location. That's what matters. That's also why you need to know the anechoic (or pseudo-anechoic) response of the loudspeakers before you measure the in-room response, whether at one place or averaged over many. If you don't know what the loudspeakers are doing intrinsically, what point is there in knowing the pretty random changes caused by a room? Bear in mind that the in-room response will change, sometimes radically, with small changes in loudspeaker or listening position, and even moving a piece of furniture will change the in-room response. The in-room response is a useful test of known good loudspeakers to optimise positioning of both loudspeakers and furnishings, but pretty pointless unless the loudspeakers are known to be good first.

S.

Here is my logic..

Obviously room alters the "distortion response", and sure, nobody argues with that. It also alters frequency response as well. I measured it so I can compensate for it via room EQ. Frequency response that I measured is certainly not how my speakers would respond in anechoic environment, and yet you didn't have complaints on that. But now when I'm doing the same thing with "distortion response" you are saying only anechoic distortion response should be considered. Why so, when the only thing that is really different between the two is that I have no means to compensate for "distortion response"?

 

[DonH56]

Here is my logic..

Obviously room alters the "distortion response", and sure, nobody argues with that. It also alters frequency response as well. I measured it so I can compensate for it via room EQ. Frequency response that I measured is certainly not how my speakers would respond in anechoic environment, and yet you didn't have complaints on that. But now when I'm doing the same thing with "distortion response" you are saying only anechoic distortion response should be considered. Why so, when the only thing that is really different between the two is that I have no means to compensate for "distortion response"?

Because the room is not adding distortion, just changing the amplitude of frequencies that are already there. Distortion adds new frequencies, and the room affects how loud those frequencies are relative to the fundamental. If you want to figure out your room's response, then sweep and measure. If you want to find the distortion from your speakers, you must do that independently of the room or it will change the results, making your speakers seem like they have more or less distortion than they actually create.

If you have a buzzer and want to know how loudly it plays, you measure in a quiet room. If you measure it in a construction zone, or in a bar with conversations and loud music, then you must subtract out the noise from the venue to determine how loudly the buzzer can play. If you want to know how much distortion the speaker generates, you must compensate for other frequency variations to isolate the ones you want to measure. If I know A and B and C = A+B then I can solve for C. If A changes or I add D then the equation is no longer valid. Too many variables changing at once means you cannot solve the equation.

This would be way easier in person, with paper and pencil or a whiteboard...

 

[sergeauckland]

Here is my logic..

Obviously room alters the "distortion response", and sure, nobody argues with that. It also alters frequency response as well. I measured it so I can compensate for it via room EQ. Frequency response that I measured is certainly not how my speakers would respond in anechoic environment, and yet you didn't have complaints on that. But now when I'm doing the same thing with "distortion response" you are saying only anechoic distortion response should be considered. Why so, when the only thing that is really different between the two is that I have no means to compensate for "distortion response"?

Sorry, but what is 'distortion response'?

You can compensate for frequency response variations using EQ. You can't compensate for distortion as it's non-linear. It will vary with frequency and level, and whether measured in-room or pseudo anechoically. Knowing how much distortion your loudspeakers produce is useful, although there's not a lot you can do about it. Knowing how much distortion you're measuring in room is of no value whatsoever as it tells you nothing about what the loudspeakers are doing, nor does it give you any possibility of improving it.

S.

 

[Krunok]

You are missing my point. I agree you need to know the in-room frequency response to accurately measure the relative distortion levels, and as @sergeauckland said you need to know the response independent of the room or as I said must compensate for the room's response. The problem is you cannot measure the room without using your speakers to do it, so their distortion (plus whatever else is added by the signal chain, including the mic and preamp back into the PC) is added to (embedded in) the measured result. You need a scheme that separates speaker distortion from the room's response.

You can use manufacturer's or reviewer's measurements to provide a baseline for your speakers, which are often taken in an anechoic chamber or use a time-gated measurement to "take out" the room's effects.

Let me try again, then I have to get back to work: Send pink noise or a swept tone at a low level through the system and measure the response. Make it low enough that speaker distortion is negligible and record the response. Assume the ripples are due to the room and not the speakers, or at least not from speaker distortion (ripples due to drivers and such will be included but those are not nonlinear distortions). Now increase the level to the desired measurement level and repeat. SCale the room's response to the new power level and back out its effects.

Using numbers: Play a low-level set of tones at 1 kHz, 2 kHz, and 3 kHz. Say the 1 kHz is at 60 dB, 2 kHz at 66 dB (6 dB high), 3 kHz at 55 dB (5 dB low). Now play a 1 kHz tone at 90 dB and again check the 2 and 3 kHz terms (second and third harmonics). You must reduce the 2 kHz measurement by 6 dB and increase the 3 kHz level by 5 dB to determine (roughly) the actual second and third harmonic distortion terms independent of (compensating for) the in-room response. This assumes nothing changed except signal amplitude between readings. Repeat for all the other frequencies and positions desired.

HTH - Don

Oh, I don't think I'm missing your point, I believe your point and method you're proposing is quite clear to me. But I am simply trying to point out that in a same way room is affecting frequency response it is also affecting distortion response. So the following question can be asked: doesn't the anechoic distortion response have the same meaning as anechoic frequency response? Isn't it the true room response thing that I actually hear in my room and not the anechoic variants?

 

[Krunok]

Because the room is not adding distortion, just changing the amplitude of frequencies that are already there.

Heh... isn't that a "tomayto, tomahto" kind of thing? If I alter amplitudes of the distortion harmonic components I am certainly altering the distortion response, am I not?

 

[DonH56]

Oh, I don't think I'm missing your point, I believe your point and method you're proposing is quite clear to me. But I am simply trying to point out that in a same way room is affecting frequency response it is also affecting distortion response. So the following question can be asked: doesn't the anechoic distortion response have the same meaning as anechoic frequency response? Isn't it the true room response thing that I actually hear in my room and not the anechoic variants?

See previous post. The topic was not what you hear; it was what distortion the speakers produce. At least that's what I thought the topic was... Yes, of course the room affects the distortion, but that is the problem. If you want to know the speakers distortion you must take the room out of the equation. In-room frequency response and the distortion your speakers produce are separate things. Yes, the room will change the level of distortion over frequency, but the speakers cannot change the room.

Heh... isn't that a "tomayto, tomahto" kind of thing? If I alter amplitudes of the distortion harmonic components I am certainly altering the distortion response, am I not?

No it is two different things. One is the amount of distortion the speaker generates, the other is how the room affects those distortion. The room does not create the distortion, only modifies the levels that are produced by the speaker, hiding the "real" speaker response. You are not measuring the distortion from the speaker but rather that distortion convoluted with the room's response. You can do that if you wish, and it is useful to know what terms will be emphasized and de-emphasized by the room, by the speakers will have a different response at different places in your room let alone in somebody else's room. If you want to compare the distortion your speakers generate to measurements from somebody else or the manufacturer, then you must remove the room form the distortion plots.

 

[Krunok]

Sorry, but what is 'distortion response'?

It's a 3D table, frequency, level and "induced harmonic response" for that frequency at that level (produced by speakers and altered by the room).

Sorry, but what is 'distortion response'?

You can compensate for frequency response variations using EQ. You can't compensate for distortion as it's non-linear. It will vary with frequency and level, and whether measured in-room or pseudo anechoically.

In theory I can't see why it cannot be done. I just need to know the whole table and have a real time DSP engine that will drive the inverse function of that response in a very same way it is done with frequency response.

 

[Krunok]

See previous post. The topic was not what you hear; it was what distortion the speakers produce. At least that's what I thought the topic was... Yes, of course the room affects the distortion, but that is the problem. If you want to know the speakers distortion you must take the room out of the equation. In-room frequency response and the distortion your speakers produce are separate things. Yes, the room will change the level of distortion over frequency, but the speakers cannot change the room.

Please see my reply to @sergeauckland
Ultimately I would like to have "distortion response" compensating feature as we have with "frequency response". And for that I need to take into account how room affects distortion in the same way as when doing room EQ with frequency response.

I know it sounds strange, but please try to think out of the box here..

 

[DonH56]

In theory I can't see why it cannot be done. I just need to know the whole table and have a real time DSP engine that will drive the inverse function of that response in a very same way it is done with frequency response.

Think about that a little more. How does the DSP know what is distortion from your speaker (or anything else)? If there is a 2 kHz peak the DSP can reduce that peak, but it does not know if the extra energy is the second harmonic distortion from a 1 kHz tone or the fundamental of a 2 kHz tone.

Please see my reply to @sergeauckland
Ultimately I would like to have "distortion response" compensating feature as we have with "frequency response". And for that I need to take into account how room affects distortion in the same way as when doing room EQ with frequency response.

I know it sounds strange, but please try to think out of the box here..

Funny, I have often received kudos for thinking outside the box. Maybe I am too old fashioned and stupid for ASR... We are telling you how to compensate the response, see the previous posts. I think we have a fundamental disconnect in understanding and knowledge base. Shades of that other thread.... I have to get back to work, will check back later...

 

[Krunok]

Think about that a little more. How does the DSP know what is distortion from your speaker (or anything else)? If there is a 2 kHz peak the DSP can reduce that peak, but it does not know if the extra energy is the second harmonic distortion from a 1 kHz tone or the fundamental of a 2 kHz tone.

I don't have all the details yet. DSP would probably have to analyse the presence of fundamental 2kHz tone and then compensate for the effects of 2nd harmonic of 1kHz tone and 3rd harmonic of 500Hz tone. Let's here assume we would correct only for 2nd and 3rd harmonic as it would probably get too hard for todays computers to do it.

Funny, I have often received kudos for thinking outside the box. Maybe I am too old fashioned and stupid for ASR... We are telling you how to compensate the response, see the previous posts. I think we have a fundamental disconnect in understanding and knowledge base. Shades of that other thread.... I have to get back to work, will check back later...

I sincerely apologise, I never meant to have it sounded like that. The sheer fact you're willing to participate in this discussion shows you're making an effort not to think conventionally. It was just a poor choice of words from my side.

 

[RayDunzl]

Frankly, I'm probably missing something but I don't understand what's the big difference between measuring frequency response and distortion. Both of them are actually frequency response except in the distortion case those additional tones are not played but are induced by speaker and they are quite silent.

The distortions (harmonics) rise faster than the fundamental as the speaker exceeds its comfort zone and the fundamental level increases.

They are not silent.

They may or may not be recognized with music - I think it disturbs the "clarity" of the presentation. I don't perceive my little JBLs to be as "clean" as the electrostats at moderately high levels with music.

Harmonic distortion does not visually affect the frequency response (RTA) measurement (much), because, the harmonic level is (typically) lower than the measured fundamental across the frequency band. Even if the harmonic "adds" to the level reported the sum is small.

Assume 80dB fundamental with a -10dB second harmonic (severe, easily audible on a tone), with additional tones at -20, -30, -40, etc:

Just calculating SPL here, an 80dB main tone, and these severe harmonics contribute only another .45dB:

JBL vs ML at 82dB at the listening position.



I listen to the JBLs all the time - TV, radio, casual CD - except when I sit down and really want to hear some music.