DiyMobileAudio.com Car Stereo Forum banner
21 - 36 of 36 Posts

·
Registered
Joined
·
555 Posts
I should have been more clear, my question was specifically in relation to subwoofers and midbasses/midranges, not tweeters. Crest factor doesn't really help either, because we don't listen to the same song all the time.
Regarding the "general ratio", I don't think so. I also don't think there's much of a standard for what's peak.

Even "RMS" is kind of ridiculous because there really isn't anything called "RMS Power". RMS is a property of an AC voltage. So, RMS is a VOLTAGE that, when squared and divided by the load's impedance, provides some amount of power.

The ACTUAL rating should be "continuous average".

There are a lot of ways to measure this but standard procedure is to use a noise of some kind--pink noise, for example. So, if the speaker is to handle 100 watts, then the amplifier's output is calibrated using a voltage for the speaker's impedance that results in 100 watts. Then, pink noise is played through the speaker. Pink noise is white noise (equal energy at every frequency) passed through a pink filter, which is a 3dB/oct low pass filter. So, 20Hz is the only frequency that gets the full 100 watts. at 40Hz, it's 50 watts. At 80Hz it's 25 watts and so on and so forth until 20k is 0.1 watts.

So, for midbass and midrange, the rating depends a lot on whether it's tested with a crossover or without a crossover.

CTA has a standard for this. EIAJ apparently has a standard for this. Maybe DIN has a standard for this. And on and on and on and on.

But, whether the rating is RMS or continuous average, this is the amount of continuous power your speaker will handle if it's a sub. If it's a midbass, midrange or tweeter, then it has likely been tested using a high pass filter--because this is how speakers work.

And no matter what people THINK they're sending to their mids and tweeters, they're likely sending a lot less simply because of how power is distributed in most music.

It isn't straightforward at all and none of my attempts to explain it over that past 20 years has been sufficient. The notion that speakers should be "overpowered" should only apply to listeners who can hear distortion in their speakers and will manage the levels appropriately. Overpowering doesn't PREVENT damage.
 

·
Registered
Joined
·
539 Posts
It isn't straightforward at all and none of my attempts to explain it over that past 20 years has been sufficient.
No, it is not straightforward. Nothing wrong with your explanations, this stuff is just that hard. A speaker seems very simple to the casual observer, but in reality it is a miracle of modern science and engineering. Josef Anton Hofmann (of Hofmann's iron law) had a PHD in Physics from Harvard and worked on the Manhattan project. That white and pink noise you are talking about is a Fourier transformation, that is something that most people have never heard of, and even fewer understand. I know from your interview on the SQOlogy podcast that you have worked with some of the greatest minds in the business, so you know all of this.

For the casual consumer, they need an easy heuristic. "The specs say this is a 50 watt tweeter so I need a 50 watt amp" will get most people close enough to make them happy.
 

·
Registered
Joined
·
555 Posts
No, it is not straightforward. Nothing wrong with your explanations, this stuff is just that hard. A speaker seems very simple to the casual observer, but in reality it is a miracle of modern science and engineering. Josef Anton Hofmann (of Hofmann's iron law) had a PHD in Physics from Harvard and worked on the Manhattan project. That white and pink noise you are talking about is a Fourier transformation, that is something that most people have never heard of, and even fewer understand. I know from your interview on the SQOlogy podcast that you have worked with some of the greatest minds in the business, so you know all of this.

For the casual consumer, they need an easy heuristic. "The specs say this is a 50 watt tweeter so I need a 50 watt amp" will get most people close enough to make them happy.
An easy heuristic would be great. We currently have one--RMS Power Handling--but they won't accept that one. LOL.
 

·
Registered
Joined
·
1 Posts
Peak numbers are inherently scams. How long is a peak? It can't be a measurement of thermal capacity if it's a peak. So it could be the limit of what your mechanical suspension can handle in an unspecified box at an unspecified frequency?
Peak numbers are scams? It's a measure of how much power can be delivered or handled by the equipment over a limited amount of time. Amplitude varies based on the signal being presented by the music with soft passages and loud passages over a specific period of time. An amp/speaker can deliver/handle more, peak, power over a very short period of time, and stay within spec, to handle the increased signal level as opposed to moderate levels over a more prolonged, continuous, period of time.
 

·
Registered
Joined
·
3,481 Posts
Regarding the "general ratio", I don't think so. I also don't think there's much of a standard for what's peak.

Even "RMS" is kind of ridiculous because there really isn't anything called "RMS Power". RMS is a property of an AC voltage. So, RMS is a VOLTAGE that, when squared and divided by the load's impedance, provides some amount of power.

The ACTUAL rating should be "continuous average".

There are a lot of ways to measure this but standard procedure is to use a noise of some kind--pink noise, for example. So, if the speaker is to handle 100 watts, then the amplifier's output is calibrated using a voltage for the speaker's impedance that results in 100 watts. Then, pink noise is played through the speaker. Pink noise is white noise (equal energy at every frequency) passed through a pink filter, which is a 3dB/oct low pass filter. So, 20Hz is the only frequency that gets the full 100 watts. at 40Hz, it's 50 watts. At 80Hz it's 25 watts and so on and so forth until 20k is 0.1 watts.

So, for midbass and midrange, the rating depends a lot on whether it's tested with a crossover or without a crossover.

CTA has a standard for this. EIAJ apparently has a standard for this. Maybe DIN has a standard for this. And on and on and on and on.

But, whether the rating is RMS or continuous average, this is the amount of continuous power your speaker will handle if it's a sub. If it's a midbass, midrange or tweeter, then it has likely been tested using a high pass filter--because this is how speakers work.

And no matter what people THINK they're sending to their mids and tweeters, they're likely sending a lot less simply because of how power is distributed in most music.

It isn't straightforward at all and none of my attempts to explain it over that past 20 years has been sufficient. The notion that speakers should be "overpowered" should only apply to listeners who can hear distortion in their speakers and will manage the levels appropriately. Overpowering doesn't PREVENT damage.
Regarding how much power a speaker is actually using, I tested my old 15" ported sub and it was getting 36 watts at my normal listening level.
 

·
Registered
Joined
·
3,481 Posts
An easy heuristic would be great. We currently have one--RMS Power Handling--but they won't accept that one. LOL.
We like to test the limits! But I would never try to return a driver that got more than its RMS rating, my testing is on my dime, not yours.
 

·
Registered
Joined
·
2,300 Posts
Many people simply dont understand some specs or how they are derived, and a lot of manufacturers put out numbers in a way that looks better than they are (if you dont look careful at how they are defined) and count on that people just dont read them correctly. some of them are sens.-SPL, power handling, X-max. some similar practice is present also at rating amplifiers.
 

·
Registered
Joined
·
663 Posts
Peak numbers are scams? It's a measure of how much power can be delivered or handled by the equipment over a limited amount of time. Amplitude varies based on the signal being presented by the music with soft passages and loud passages over a specific period of time. An amp/speaker can deliver/handle more, peak, power over a very short period of time, and stay within spec, to handle the increased signal level as opposed to moderate levels over a more prolonged, continuous, period of time.
The post you quoted literally explains why. How long is the peak? Is it 1 cycle at 20hz? Is it DC current for 20ms? Is it 1 cycle at 20k? How about 1 cycle at 100k? Is 20k sustained for 50ms?
There is no standard to it.
So a "peak rating" tested by feeding a burst of 1 cycle of 20khz is going to be much higher than if it was tested by feeding it 1khz for 50ms.
Hence why many see it as a scam. No one knows what it means cause it can mean anything the manufacturer wants.
 

·
Premium Member
Joined
·
7,316 Posts

… Pink noise is white noise (equal energy at every frequency) passed through a pink filter, which is a 3dB/oct low pass filter. So, 20Hz is the only frequency that gets the full 100 watts. at 40Hz, it's 50 watts. At 80Hz it's 25 watts and so on and so forth until 20k is 0.1 watts.
That actually adds up to 200W, but I think that we all get it.
  • 160=12
  • 320=6
  • 640=3
  • 1280=1.5
  • 2500= 3/4
And if we start here at 2500, then with the 200W pink noise, we have about a watt going to the tweeters.


The post you quoted literally explains why. How long is the peak? Is it 1 cycle at 20hz? Is it DC current for 20ms? Is it 1 cycle at 20k? How about 1 cycle at 100k? Is 20k sustained for 50ms?
There is no standard to it.
So a "peak rating" tested by feeding a burst of 1 cycle of 20khz is going to be much higher than if it was tested by feeding it 1khz for 50ms.
Hence why many see it as a scam. No one knows what it means cause it can mean anything the manufacturer wants.
It started in the context of 100W (so I assume tweeter or MR).
If it is in the context of music then that 1000/100 ratio is 10dB, which is not an unreasonable crest factor value.

So it would be as long a typical musical transients last, but I am not sure how long they are..
 

·
Registered
Joined
·
207 Posts
I know my speakers are only getting a few watts because I can only take about 90dba. So why do aftermarket speakers always sound better with a dedicated amp versus the one in a stock or aftermarket receiver?
 

·
Registered
Joined
·
108 Posts
I know my speakers are only getting a few watts because I can only take about 90dba. So why do aftermarket speakers always sound better with a dedicated amp versus the one in a stock or aftermarket receiver?
Lots of possible reasons.
Dynamic headroom, total harmonic distortion at the volume level being used, overall linearity, channel separation, damping factor, bad preamp design, poor isolation of noise on the power line, no isolation from digital circuits within the head unit....

All devices will introduce some degree of distortion, but the timbre of the distortion can be different and change whether that distortion sounds pleasant.
 

·
Premium Member
Joined
·
7,316 Posts
I know my speakers are only getting a few watts because I can only take about 90dba. So why do aftermarket speakers always sound better with a dedicated amp versus the one in a stock or aftermarket receiver?
Which factory HUs?
 

·
Premium Member
Joined
·
7,316 Posts
I agree with that in the big picture, but there is no standard on that being what is tested now.
I am not sure ^that^ is true?
Isn’t here some DIN or ISO type of pink noise standard?

One could take that digital stream and do a histogram, or better yet a manifold… of level versus time.
It is probably something like a gaussian distribution, as noise is usually modelled and written in the literature as “pure white Gaussian noise”. But in this case it is “pretty in pink”.

Is there a file somewhere of what is used for these speaker tests?

(If I was testing my own speakers I would probably choose a file with compression as it gives a higher RMS loudness and the maximum peaks are therefore a bit lower.)
 
21 - 36 of 36 Posts
Top