Is a Speaker with Higher DB Sensitivity Always Louder?

Discussion in 'Amps and Cabs' started by Schaaner, May 16, 2019.

  1. Schaaner

    Schaaner Member

    Messages:
    24
    Joined:
    Aug 5, 2015
    Location:
    Zürich, Switzerland
    This could turn out to be a stupid question, but please let my try. So many factors seem to be relative when it comes to speakers.

    Imagine a 12w amp into a 2x12 cab with two speakers:
    1. One with 25w power and 97db sensitivity
    2. Another one with 65w power and 100 db sensitivity
    Is the second speaker louder in this scenario? Is that always the case no matter how hard both are pushed (eg. amp at volume 2/10 vs 8/10), or: is the second speaker only louder than the first speaker when it is pushed hard enough?
     
    eigentone likes this.
  2. wyatt

    wyatt Member

    Messages:
    4,163
    Joined:
    Oct 11, 2005
    Yes. It's louder all the time.

    That's why we recommend a more efficient speaker to people who are looking for get more headroom out of their amp...because it makes everything louder at every Volume setting.

    Sensitivity: XX decibels @ 1 meter @ 1 watt
     
    ripple, FbIsNotE and Schaaner like this.
  3. Jerrod

    Jerrod Member

    Messages:
    11,826
    Joined:
    Jan 7, 2002
    Assuming equal impedances, right?
     
    FbIsNotE likes this.
  4. tech21nyc

    tech21nyc Member

    Messages:
    1,544
    Joined:
    Aug 17, 2010
    The sensitivity is not the same from manufacturer to manufacturer.
     
    lemonman, QRSS, FbIsNotE and 4 others like this.
  5. msquared

    msquared Member

    Messages:
    102
    Joined:
    Aug 15, 2010
    Location:
    Lawrence, KS
    This is an incredibly important point. There is no standard for sensitivity measurements, nor are they typically done by an impartial third party. It would be nice if everyone used the same calibrated test gear, the same anechoic chamber, and the same series of distances, but there is absolutely nothing scientific about how this stuff is done.
     
  6. teemuk

    teemuk Member

    Messages:
    2,920
    Joined:
    Apr 21, 2009
    And a single rating for efficiency or SPL is always just a “nominal” value, just like in the case of impedance. In practice, the sensitivity of loudspeaker is different at different frequencies and cannot be expressed by a single rating. We need a graph, i.e. the “frequency response” of the speaker.

    Which begs several questions: How does a manufacturer derive this single “nominal” rating? Is it the average or perhaps the “peak” rating occurring only at specific, very narrow band of frequencies? At what frequency? How do they even measure their ratings?
    And, which loudspeaker is louder: a 100dB/1W tweeter or a 83dB/1W mid-range driver when one wishes to produce the frequency of, say, 400 Hz? Which one of them is louder at 7 kHz? And, do we humans perhaps sense some frequencies “louder” than some other frequencies, etc. In case you didn’t realize it, those are all rhetorical questions.

    The power rating of a speaker has practically almost nothing to do with perceived loudness. The little it has just covers few rough rules of thumb one can expect in relation of amplifier's output power to sound pressure a speaker produces:
    e.g. a 3 dB increase in driving power (2x increase in terms of output power) increases perceived loudness by factor of about 1.2x, a 6dB increase (4x more output power) by about 1.5x, and a 10 dB increase (10x more output power) by about 2x. Of course these are generalisations and subject to (subjective) differences in how individuals even perceive loudness and how all that interacts with frequencies in question.
    So, we can use the "power rating" to determine if perhaps the speaker can withstand X amount of increase in driving power required for Y amount of increase in loudness.
    Do also note that humans largely perceive "loudness" from the average pressure of the sound waves, not from peak pressure. Signal type matters: Compressed signals generally sound louder because of that "average" thingy.

    Basically, speaker’s “power rating” is just a rating that indicates how much driving power the driver can safely handle. It’s also derived very differently than output power ratings of amplifiers so those two figures aren’t specifically comparable with each other.
    Also - and this is a very important point - speaker's power rating also pretty much just covers the thermal effects, not effects related to excursion. For example, in regards to excursion limits a 100W rated speaker may withstand that 100W driving power at 1kHz but not at 100 Hz because of excursion being greater at lower frequencies.
    So, my advise for everyone is to ignore the speaker's power rating and use your ears. If the speaker sounds like its about to die then it probably is, regardless of whether power driving it is "within the rating" or not. Also, one can't safely expect that speaker rated for, say, 10W can safely handle output from a 10W amplifier because those two ratings aren't really related at all.
     
    Last edited: May 16, 2019
  7. eigentone

    eigentone Supporting Member

    Messages:
    6,011
    Joined:
    Jun 26, 2015
    Location:
    Massachusetts
    Is a Speaker with Higher DB Sensitivity Always Louder?

    No. Sensitivity is not synonymous with loudness or volume.

    The main reasons:
    1. Different manufacturers use different approaches to measure Sensitivity.
    2. People hear and process sound differently. See Yanny/Laurel. They also hear sound differently at different amplitudes. See Fletcher/Munson. You can take two speakers with the same Sensitivity and you may be sure one is louder in a specific test scenario and your friends may hear the other speaker as louder or the same. Quadruple the power and the results may change some. Quadruple the power again and the results may change again.
    3. Some methods used to measure Sensitivity can be very off-mark, if you consider Sensitivity synonymous with Volume. Celestion is a popular speaker maker. Their method for measuring Sensitivity is a 1 Watt 1 kHz tone, measured at 1 meter on axis. If you have looked at guitar speaker frequency response charts, they are far from flat. Guitar speakers add a lot of character. Measuring a single frequency and equating that to volume is like writing a book report based on a few pages.
    4. No Sensitivity measurement for guitar speakers accounts for dispersion, to my knowledge. A 10" has different dispersion than a 15". A driver with much wider dispersion may have a lower Sensitivity, even though it's pumping out more total sound pressure.
    5. Whether it's louder also depends on the actual signal it sees (ie your guitar, performance, and amp.) The speaker accentuates some frequencies more than others. If you have a very midrangey amp with a very mid-scooped speaker, then the two may complement each other in a way pleasing to the ear but it may simply not be the loudest combination. And if you are running the amp at full power, then a speaker/cab that reproduces lows well could help you get more volume because the amp would not need to put out as much lows.
    6. Whether it's louder also depends on the cabinet and room it's in.

    Is that always the case no matter how hard both are pushed (eg. amp at volume 2/10 vs 8/10), or: is the second speaker only louder than the first speaker when it is pushed hard enough?

    Presumably, 2 is the amp clean and 8 is the amp with some tube drive. A driven tube amp changes the strengths of various harmonics. Again, the answer is that the speaker with the higher sensitivity will not necessarily sound louder. See #2, #3, and #5.

    I see way too many people oversimplify what Sensitivity is. They look to it for different speakers and conclude something along the lines of "Speaker A is 3 dB louder." Don't do that. It's an oversimplification and the conclusion is often inaccurate. In reality, ears, speakers, and acoustics are all very complex and nonlinear.
     
    Last edited: May 16, 2019
  8. easyed

    easyed Supporting Member

    Messages:
    1,918
    Joined:
    May 14, 2017
    Location:
    Seattle
    Everybody knows that speaker efficiency is measured 1 Watt, 1 Meter. But few recognize that the measurement is at 1000 Hz. So manufacturers design speakers to be slightly peaked at 1000 Hz. A speaker with a fairly flat response may have better overall loudness at the higher and lower frequencies.

    Guitar speakers are not designed to be hi-fi, and so the sensitivity rating for guitar speakers is a pretty good indicator of relative loudness,
     
    Schaaner and theRagman like this.
  9. Jon Silberman

    Jon Silberman 10Q Jerry & Dickey Gold Supporting Member

    Messages:
    39,003
    Joined:
    Jan 3, 2002
    Location:
    Northern VA
    Exactly. So the answer is NO, not ALWAYS. +, as has been noted, the typical dB measurement is limited to a specific frequency.
     
    Schaaner and eigentone like this.
  10. C-4

    C-4 Member

    Messages:
    11,510
    Joined:
    Mar 23, 2009
    Location:
    I love Red Bank, NJ! Florida now Europe later
    Taking everything these players above have wisely posted, my final thought would be to try to get SPL ratings from multiple speakers to be the same or within 1 db. It doesn't guarantee that all the speakers will have exactly the same loudness, but if they were made to tight tolerances, a speaker with a 3 db difference to the next speaker "may" be heard to be louder. If this does occur, you will not necessarily get the desired blended effect of two different speakers equally.

    This could be a good thing or not so good thing, depending on what the player hears at volume from the cab. One speaker being slightly more dominant then the other may turn out to be a good thing. That will depend on who is listening to the output from that cab.
     
    eigentone and Schaaner like this.
  11. De Batz

    De Batz Member

    Messages:
    1,974
    Joined:
    Nov 13, 2015
    And 1 kHz is some way higher than the fundamental frequency of anything on the guitar. That's not to say that the bulk of the sound is necessarily at the fundamental frequency (from limited experiments 20 years ago, it is, though). But the measured point here isn't even in the range of fundamental frequencies.
     
    eigentone likes this.
  12. zenas

    zenas Member

    Messages:
    7,110
    Joined:
    Jul 1, 2011
    Interesting question, the "standard testing" is done at one watt, maybe at say ten watts things change?
     
    eigentone likes this.
  13. sharpshooter

    sharpshooter Member

    Messages:
    3,906
    Joined:
    Jun 29, 2008
    Companies can also use "tricks" to boost the sensitivity numbers.
    The easiest way is to make the cone very thin/lightweight, and a "loose" surround,, this gets the sensitivity rating higher, but when you start pushing some watts the speaker starts breaking-up.
    The measurement is done at 1 watt,, do you play your amp at 1 watt??,, probably not.
    And has been noted aforehand, speakers are not linear,, depending upon construction details, a lessor rated speaker might easily be louder under "real use conditions" than one that is advertised as having a higher sensitivity rating.
     
    zenas and eigentone like this.
  14. De Batz

    De Batz Member

    Messages:
    1,974
    Joined:
    Nov 13, 2015
    So, working on the assumption that you're referring to Eminence speakers, how is it that they manage to give such high power handling figures? Presumably the power handling is a combination figure, so they can't just put a ludicrously high temperature voice coil in then say it'll handle 2000W, even though the cone itself is shredded at 40W. Or do they use a cone material that is thin and flexible but strong in tension so it doesn't give up in that way?
    This is a serious question, BTW, I'm not challenging your point in itself.

    I struggle to interpret Eminence on-paper ratings, because whilst their numbers seem to be reasonably internally coherent, they don't correspond to a predictable difference to (for example) Celestion, Fane or Tayden in terms of perceived loudness. I'd use the sensitivity figure for Jensen, Celestion and so on to give me a sense of how loud the speakers might be, because I'm used to them.
     
  15. teemuk

    teemuk Member

    Messages:
    2,920
    Joined:
    Apr 21, 2009
    I wouldn’t get too caught up on concept of ”fundamental frequencies”. Except for synthesizers, I’m not aware of any musical instrument that would produce mere sine waves. In fact, “timbre” of an instrument is largely defined by harmonics additional to fundamental frequency. A guitar signal, for example, features third, fourth and even fifth harmonic in about same magnitude as the fundamental. Even second harmonic is very strong.
    When deliberate distortion (harmonics) is introduced by overdrive and clipping it boosts upper order harmonics even more. Simultaneously to producing distortion it is common to attenuate the lower order harmonics, including the fundamental. A generic "high gain" distortion design may drastically filter frequencies below about 1 kHz in order to reduce that "farty" low frequency intermodulation distortion.

    Speaking of 1 kHz tests signals… I think 1 kHz is quite “interesting” frequency for several reasons related to human hearing. For example, human hearing is very sensitive and most “linear” at mid-range frequencies, to which 1 kHz “co-incidentally” happens to fall. In fact, when considering “weighting” to compensate non-linear effects of our hearing (in regard to loudness, e.g. Fletcher-Munson) it just happens that at around 1kHz we need NONE.
    1 kHz is also close to “center” if we divide average human hearing range into octaves. Many tone controls are designed with “crossover” at around 1 kHz. Coincidence?
    The loudspeaker’s response at mid-range is also probably the most vital part in defining the “tone” of that loudspeaker.
    Average distance of human ears is also closely related to wavelength of a 1 kHz signal. We employ this phenomenon, for example, in detecting direction of a sound source. For example, frequencies above about 1 kHz are shadowed by the head so ear closest to sound source hears a stronger signal. On the other hand, below 1 kHz we sense the delay between signals arriving to each ear. Our brain uses all this information to process direction and distance of sound.

    Basically, in acoustics the 1 kHz is a “magic number” one seems to encounter in many places.

    But I agree, in regards to speaker’s perceived “loudness” the measurement at 1 kHz is just part of the whole story. But it wasn’t probably picked “by accident”.
     
    Daytona57, eigentone and Jerrod like this.
  16. zenas

    zenas Member

    Messages:
    7,110
    Joined:
    Jul 1, 2011
    When I finally get moved into my permanent residence, and get my crap set up, I'm going to do some wattage measurements. Just curious how many watts I'm actually using at the speakers.
    Pretty much get all the volume I ever need and then some with a Princeton Reverb or 6v6 Reverberockets. Those amps aren't even 20 watts, I think 15 for the PR and 12 for the RRs.(gotta scope those out too)
    On the flip side about my favorite amp is a 76ish Twin Reverb, says 100 watts but I'm playing it in the same volume range as the small amps. Anyway I'm just curious how many watts I'm actually using.
    Should note that I play loud, often louder at home than anywhere else, a perk of country living. Someone in an apartment would use less than me for sure.
     
  17. teemuk

    teemuk Member

    Messages:
    2,920
    Joined:
    Apr 21, 2009
    BTW, given logarithmic relation of perceived loudness vs. power, and crest factors of typical signals, I'd say a 100W amp in average drives the speaker closer to 1W of power than to 100W.

    1W of driving power has potential to be very, very loud. (e.g. My 0.5W amp is way too loud for an apartment use). A sensitity of mere 85 dB / 1W means you risk damaging your hearing by driving the amp at mere 1W of power for continuous period of time.

    What happens when you push the speaker past that 1W... Well, it simply gets louder - and if you're lucky and the speaker doesn't fail due to excessive dissipation of heat - continues doing so rather "linearly" to the point where it just starts to distinctly distort by several different mechanisms. Just like an amp operates when you push it to produce more than its rated output power. AFAIK, the relation of power and perceived loudness is quite "linear" up to that distorting point. There could be side effects like "thermal compression" (which affects operation of the amplifier driving the speaker). Thermal compression basically means that the impedance of the speaker increases as the voice coil heats up (similarly to incadescent light bulbs).
     
    Schaaner, Daytona57 and eigentone like this.
  18. De Batz

    De Batz Member

    Messages:
    1,974
    Joined:
    Nov 13, 2015
    You're absolutely right about the timbre being determined by all the harmonic content. Somewhere in my parents' loft is a paper I wrote during Physics A level that basically shows the harmonic series of a set of guitar sounds, together with oscilloscope traces. My kit was basic - a Yamaha Pacifica and a Peavey Studio Pro, along with a DOD Grunge - but I used a computer scope and FFT to get a sense of what there was in those waves. From memory, the Grunge had this extraordinary wave where at what should have been the top of a peak it shot to the opposite side of the line and back again, like a big notch or wedge in the wave (not in the EQ curve!). Whether with that, or with the clean sound of the guitar through the amp, or with the guitar unamplified, the fundamental frequency was where most of the energy was, at a guess by a factor of at least 2 (being proportional to amplitude squared). Our hearing, as you go on to argue, is focused on the mid-range (of our hearing) frequencies between about 1000Hz and 5000Hz, and therefore much less energy at those frequencies could still lead to us reporting it being louder. I'm not sure whether I would get the same results if I re-did that experiment with my current kit and with newer scopes and FFTs, but it's a place to start.

    A speaker characteristic that's supposed to tell us about the relative volume of a (guitar) speaker needs to allow us to compare across as well as within manufacturers, so for Eminence to set themselves up as an outlier is just a bit annoying.
     
    eigentone likes this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice