Frequency and -3db questions
- Status
1 - 13 of 13 Posts
Joined
·
7,335 Posts
The -3dB @ 80Hz means relative to some reference frequency and level, 80Hz is reproduced 3dB lower for the same input power. THis is typically 1kHz.
So if a 1watt input produced 86dB out of the loudspeaker with a 1kHz Sine Wave, it would only produce 83dB with an 80 Hz sine wave.
Speaker Frequency response is usually expressed as:
xxHz -- xxkHZ +/- tolerance dB
80Hz - 22kHz +/- 3dB for example.
Assuming the same conditions above, this means that from 80Hz-22kHz, with no room resonances coming in to play, the output at any given frequency would be somewhere between 83dB and 89dB.
I have seen some manufacturers do dual specs, like say:
50Hz - 22kHz +/- 1dB
35Hz - 25 kHz +/- 3dB
Which gives you a better idea of how the speaker performs.
I haven't looked at Sound and Vision in ages to see their graphs, so I can't give you any insight there.
Regards,
So if a 1watt input produced 86dB out of the loudspeaker with a 1kHz Sine Wave, it would only produce 83dB with an 80 Hz sine wave.
Speaker Frequency response is usually expressed as:
xxHz -- xxkHZ +/- tolerance dB
80Hz - 22kHz +/- 3dB for example.
Assuming the same conditions above, this means that from 80Hz-22kHz, with no room resonances coming in to play, the output at any given frequency would be somewhere between 83dB and 89dB.
I have seen some manufacturers do dual specs, like say:
50Hz - 22kHz +/- 1dB
35Hz - 25 kHz +/- 3dB
Which gives you a better idea of how the speaker performs.
I haven't looked at Sound and Vision in ages to see their graphs, so I can't give you any insight there.
Regards,
Joined
·
7,335 Posts
Evan,
I think you meant 3dB gap, not 3Hz gap
Even so, in the case I gave you, which is a classic THX speaker, this doesn't necessarily mean there's an issue.
They just say that from the rated response you will be no more or less than 3dB higher or lower than 1kHz (or whatever their reference level is).
THe key is after the -3dB point how fast things drop, and depending on speaker design is the answer.
Sealed enclosures drop off at 12dB per octave below the f3 (-3dB response point) and ported enclosures drop off at 24dB below the f3 (-3dB response point).
Full range planars drop off at 12dB/octave as well.
I probably made a bad choice on showing you an example with the THX speaker, which is specifically designed around an 80Hz low pass crossover.
Regards,
I think you meant 3dB gap, not 3Hz gap
Even so, in the case I gave you, which is a classic THX speaker, this doesn't necessarily mean there's an issue.
They just say that from the rated response you will be no more or less than 3dB higher or lower than 1kHz (or whatever their reference level is).
THe key is after the -3dB point how fast things drop, and depending on speaker design is the answer.
Sealed enclosures drop off at 12dB per octave below the f3 (-3dB response point) and ported enclosures drop off at 24dB below the f3 (-3dB response point).
Full range planars drop off at 12dB/octave as well.
I probably made a bad choice on showing you an example with the THX speaker, which is specifically designed around an 80Hz low pass crossover.
Regards,
Joined
·
7,335 Posts
Nope, you really typed 83Hz and no.
Don't confuse dB, which in this context is referring to Sound Pressure level with Frequency Response.
There is no gap. There's a roll off.
Let's assume the speaker I listed above has its f3 point at 80Hz as the setup suggests. It's ported, so the dropoff will naturally be 24dB/octave below that f3 point.
86dB output @ 1k means:
83dB output @ 80Hz.
59dB output @ 40Hz.
35dB output @ 20Hz.
So there is still frequency response, but it drops off rapidly, it isn't a brck wall where the speaker just stops, but the output becomes inconsequential after a while.
I hope this helped clarify things for you some more.
Regards,
Don't confuse dB, which in this context is referring to Sound Pressure level with Frequency Response.
There is no gap. There's a roll off.
Let's assume the speaker I listed above has its f3 point at 80Hz as the setup suggests. It's ported, so the dropoff will naturally be 24dB/octave below that f3 point.
86dB output @ 1k means:
83dB output @ 80Hz.
59dB output @ 40Hz.
35dB output @ 20Hz.
So there is still frequency response, but it drops off rapidly, it isn't a brck wall where the speaker just stops, but the output becomes inconsequential after a while.
I hope this helped clarify things for you some more.
Regards,
If you want to learn more about this and everything you need to know about acoustics, like what sound is, how it gets fram a microphone to your ear, room treatments, and lots more get
Master Handbook of Acoustics
by F. Alton Everest
Paperback - 592 pages 4th edition (September 22, 2000)
McGraw-Hill Professional Publishing; ISBN: 0071360972
about USD $20 I think.
Master Handbook of Acoustics
by F. Alton Everest
Paperback - 592 pages 4th edition (September 22, 2000)
McGraw-Hill Professional Publishing; ISBN: 0071360972
about USD $20 I think.
Joined
·
5,900 Posts
Quote:
Right at the crossover frequency, the main speakers and the sub are each producing information at the same amplitude. If the sub is 3 dB down and the mains are 3 dB down, the sum of the two equals 0 (zero) and the frequency response of the system is flat at the crossover frequency. If your mains and your sub were both 0 dB down at the crossover frequency, you'd have a +3 dB bump.
| That means that there would be a 3Hz gap if you set up a THX system with a 80Hz crossover because the speaker is actually limited to 83Hz? |
Joined
·
6,980 Posts
Quote:
Evan,
I've looked at this thread a few times now, and the main problem I keep coming across is where to start. Obviously one tangent on the problem has been taken up, but it is easy to get overly confused with what is going on if some basics are not clearly understood. I'll pick out a couple issues to take note of and then hopefully you can ask some further questions based on those answers and how they relate to what you are looking over in the magazines.
First and foremost, be cautious when you are reading measurements in magazines and especially those published by manufacturers (yes, I work for a loudspeaker manufacturer ). While many different graphs may all have similar horizontal and vertical markings with different "curves" plotted, rarely are two measurements directly comparable.
I'm sure someone else here can dig out a link showing where different instruments' frequency ranges fall which will give you a basis for what type of sounds we are talking about when we look at different frequency ranges on a response curve.
These response curves typically have a vertical scale in dB or decibels. In this context, a decibel is a unit measure of sound pressure level, often abrieviated as SPL. SPL is measured using a microphone and associated electronics. Realize that the decibel is a logrithmic measurement, which means that unlike the power from our amplifiers, 2x the power to a speaker does not mean twice the decibel reading. In terms of power into a loudspeaker, every doubling of power equates to approximately a 3dB increase in sound pressure level, while every halving equates to approximately a 3dB decrease in sound pressure level. The -3dB roll-off points for the low end response of a speaker comes from this as it is considered the "half power point."
Of course what we need to remember is that a loudspeaker's response doesn't just turn off below this half power point, but rather it rolls off at some rate, with some speakers rolling off faster than others. Likewise output from a speaker can most certainly be useable even when it is more than 3dB below the arbitrary reference level. The above said, another interesting slight of spec shall we say is the casual use of a +/-3dB specification. The "+" in the spec allows for a total of a 6dB swing within the frequency range noted. This allows a manufacturer to specify the low end cut-off or response as where ever the response drops 6dB below the peak of the response curve while still giving entirely accurate specifications.
Beyond the notation games, without understanding the conditions of the tests, and the smoothing (or averaging) applied, it becomes much more difficult to compare curves from different sources. Most manufacturer's response curves have 1/3rd octave smoothing, which hides many blips in the response, where 1/6th to 1/12th octave resolution is more appropriate to identify problems in a speaker, but of course don't look as pretty or smooth. Also be sure to check the vertical scale of the curve, as a 10dB/ division curve will look a lot smoother than the same curve plotted at a 3dB/ division vertical scale. Referring specifically to the Sound & Vision curves, it is important to understand what their different graphs are really depicting. The average response over 30deg is more useful to show the power response of the speaker yet may hide dips or peaks in the response as you move from on to off axis.
Realize that no one curve, measurement or specification will give you a complete picture of what a speaker will sound like or how it behaves. In fact it is precisely the combination of many measurements and the interpretation of their relative significance which is key.
Hope this helps more than confuses,
| Originally posted by eaadams on a speaker what does it mean "-3dB point = 80hz?" In adition to that could someone point me to a site that explains frequency responce as well as those speaker charts like in Sound and Vission. Thanks |
I've looked at this thread a few times now, and the main problem I keep coming across is where to start. Obviously one tangent on the problem has been taken up, but it is easy to get overly confused with what is going on if some basics are not clearly understood. I'll pick out a couple issues to take note of and then hopefully you can ask some further questions based on those answers and how they relate to what you are looking over in the magazines.
First and foremost, be cautious when you are reading measurements in magazines and especially those published by manufacturers (yes, I work for a loudspeaker manufacturer
). While many different graphs may all have similar horizontal and vertical markings with different "curves" plotted, rarely are two measurements directly comparable.I'm sure someone else here can dig out a link showing where different instruments' frequency ranges fall which will give you a basis for what type of sounds we are talking about when we look at different frequency ranges on a response curve.
These response curves typically have a vertical scale in dB or decibels. In this context, a decibel is a unit measure of sound pressure level, often abrieviated as SPL. SPL is measured using a microphone and associated electronics. Realize that the decibel is a logrithmic measurement, which means that unlike the power from our amplifiers, 2x the power to a speaker does not mean twice the decibel reading. In terms of power into a loudspeaker, every doubling of power equates to approximately a 3dB increase in sound pressure level, while every halving equates to approximately a 3dB decrease in sound pressure level. The -3dB roll-off points for the low end response of a speaker comes from this as it is considered the "half power point."
Of course what we need to remember is that a loudspeaker's response doesn't just turn off below this half power point, but rather it rolls off at some rate, with some speakers rolling off faster than others. Likewise output from a speaker can most certainly be useable even when it is more than 3dB below the arbitrary reference level. The above said, another interesting slight of spec shall we say is the casual use of a +/-3dB specification. The "+" in the spec allows for a total of a 6dB swing within the frequency range noted. This allows a manufacturer to specify the low end cut-off or response as where ever the response drops 6dB below the peak of the response curve while still giving entirely accurate specifications.
Beyond the notation games, without understanding the conditions of the tests, and the smoothing (or averaging) applied, it becomes much more difficult to compare curves from different sources. Most manufacturer's response curves have 1/3rd octave smoothing, which hides many blips in the response, where 1/6th to 1/12th octave resolution is more appropriate to identify problems in a speaker, but of course don't look as pretty or smooth. Also be sure to check the vertical scale of the curve, as a 10dB/ division curve will look a lot smoother than the same curve plotted at a 3dB/ division vertical scale. Referring specifically to the Sound & Vision curves, it is important to understand what their different graphs are really depicting. The average response over 30deg is more useful to show the power response of the speaker yet may hide dips or peaks in the response as you move from on to off axis.
Realize that no one curve, measurement or specification will give you a complete picture of what a speaker will sound like or how it behaves. In fact it is precisely the combination of many measurements and the interpretation of their relative significance which is key.
Hope this helps more than confuses,
Quote:
This is usually the case (Butterworth alignment) but depends on the system crossover alignment. If you are using a Linkwitz-Riley alignment, the sub and mains will add correctly if the respective responses are -6 dB at the crossover frequency.
Quote:
Neglecting effects from room influence and assuming equal distances to mains and sub (also assuming approximately the same acoustic phase for all transducers), the bump will be close to +6 dB.
| Right at the crossover frequency, the main speakers and the sub are each producing information at the same amplitude. If the sub is 3 dB down and the mains are 3 dB down, the sum of the two equals 0 (zero) and the frequency response of the system is flat at the crossover frequency. |
Quote:
| If your mains and your sub were both 0 dB down at the crossover frequency, you'd have a +3 dB bump. |
Joined
·
6,980 Posts
Iceman,
The point which you are contending isn't exactly right, but isn't exactly wrong. More importantly, the term and concepts you are reffering to are most certainly well beyond the knowledge level of the original question. There is no problem getting into such discussions, and I welcome it, but there is also nothing to gain by drowning someone asking a basic question in technical jargon. While I know your intent is not to mislead, but it is this state of glazed over eyes which allow marketers get away with such wild claims and statements.
Regards, and no offense intended.
The point which you are contending isn't exactly right, but isn't exactly wrong. More importantly, the term and concepts you are reffering to are most certainly well beyond the knowledge level of the original question. There is no problem getting into such discussions, and I welcome it, but there is also nothing to gain by drowning someone asking a basic question in technical jargon. While I know your intent is not to mislead, but it is this state of glazed over eyes which allow marketers get away with such wild claims and statements.
Regards, and no offense intended.
Mark,
Admittedly slightly off-topic, but so is lots of other info in this thread. I do not think any of the stated facts in my last post are the least bit wrong however, even if there is certainly even more OT acoustics to discuss. Let's save this for later.
Admittedly slightly off-topic, but so is lots of other info in this thread. I do not think any of the stated facts in my last post are the least bit wrong however, even if there is certainly even more OT acoustics to discuss
. Let's save this for later.I almost forgot something really important:
I do not think that all of the posts in a thread have to be directly related to the original question. There are always interested readers who want to know more. I know I surely do, especially in some of the other AVS forums, and I always welcome information that is at least reasonably in line with the original topic. The readers may always choose what info they need and are capable of digesting. Even when the technical expertise is way above my head, I do believe there may be something to learn and believe me, I have. The absence of censorship or self-censorship of the kind Mark Seaton is promoting is one of the reasons for myself being a member of this great forum.
When I find the time, I will also try to write a thorough explanation of the concept of sound power/intensity in relation to sound pressure and the related dB issues as the confusion seems very deep on this point.
I do not think that all of the posts in a thread have to be directly related to the original question. There are always interested readers who want to know more. I know I surely do, especially in some of the other AVS forums, and I always welcome information that is at least reasonably in line with the original topic. The readers may always choose what info they need and are capable of digesting. Even when the technical expertise is way above my head, I do believe there may be something to learn and believe me, I have. The absence of censorship or self-censorship of the kind Mark Seaton is promoting is one of the reasons for myself being a member of this great forum.
When I find the time, I will also try to write a thorough explanation of the concept of sound power/intensity in relation to sound pressure and the related dB issues as the confusion seems very deep on this point.
1 - 13 of 13 Posts
- Status
Join the discussion
Join now to ask and comment!