Jump to content

Bass punch threshold


TTS56A

Recommended Posts

So the steady-state (long window) FR got worse but phase is improved?  If that is so, and mmore regions are min phase, simple EQ will likely fix the FR excursions.  

 

Will be interesting to see how it measures (FR/IR/PVL) vs what you think when you are done.

 

I improved my ETC/IR with my plinth bass traps (30 cuft pink-fluffy insulation in a semi-WAF container):

 

post-20-0-56084400-1461815817_thumb.png

 

post-20-0-21960700-1461815817_thumb.png

 

 

 

 

 

JSS

Link to comment
Share on other sites

  • Replies 106
  • Created
  • Last Reply

I disagree that 110 dB is only "a bit over threshold" for tactile sensation, but this may be a matter of personal opinion to some extent.  I also do not routinely listen to car systems with > 130 dB or what not.  As you say, "110 dB == nothing impressive at all", but I think 110 dB can be very impresive.  IMO, the threshold is closer to 90 dB or even lower.  I believe a lot depends on what else is going on musically and acoustically.  If the room is dead quiet and the bass is all you can hear, you might even be able to feel below 80 dB.  But if you are in a very noisy large venue with a lot of reverb, it might take peaks > 110 dB at the seats to punch above the noise floor, especially with a full band playing at full tilt including a bass player that the kick may have to compete with. 

 

 

Forgive me, but since many meters are weighted, are you sure that you are measuring in flat response? (it's was also my doubt from my first post). Anyway i say this because with A scale, for example, 50 hz are -30 db respect to 1 khz. 

 

Another issue could be the rms level, wich is typical 10-15 db lower than peaks. When i measure music and reading 90 db on my meter (without any weighting), the peak level is around 100-105 db depending on the crest factor. With live music it's not unusual to see peaks of 20-25 db above rms.

Link to comment
Share on other sites

So the steady-state (long window) FR got worse but phase is improved?  If that is so, and mmore regions are min phase, simple EQ will likely fix the FR excursions.  

 

Will be interesting to see how it measures (FR/IR/PVL) vs what you think when you are done.

 

I improved my ETC/IR with my plinth bass traps (30 cuft pink-fluffy insulation in a semi-WAF container):

 

attachicon.gifETC Pre Sub.png

 

attachicon.gifETC Post Sub.png

 

 

 

 

 

JSS

 

In a room with less damping there will be boundary reflections from several surfaces, and with favorable speaker positions these reflections tend to statistically spread out and actually create a reasonably flat response. 

 

When most - but not all - of those reflections are damped, the remaining ones kind of stand out and produces huge dips in the response.

Since nulls can not be equalized, this has to be fixed with even more acoustic treatment.

Link to comment
Share on other sites

Forgive me, but since many meters are weighted, are you sure that you are measuring in flat response? (it's was also my doubt from my first post). Anyway i say this because with A scale, for example, 50 hz are -30 db respect to 1 khz. 

 

Another issue could be the rms level, wich is typical 10-15 db lower than peaks. When i measure music and reading 90 db on my meter (without any weighting), the peak level is around 100-105 db depending on the crest factor. With live music it's not unusual to see peaks of 20-25 db above rms.

 

It's un-weighted, RMS level, but it's a fuzzy figure (like +/- 5 dB maybe?) because it's not based directly on actual measurement.  I don't care about that though because by definition, threshold is a level that may not be noticed in every instance or by subject.  Scientific studies often aim for perception by 50% of a sample of test subjects.  It is based on calibrated playback level and subjective experience on a system with any obnoxious peaks EQed out, and I always review my results for bass with smoothing turned off.  IIRC, I had almost nothing get above +0 dB with respect to reference level.

 

If you think 110 dB is only a bit over threshold, then we must have *very* different personal thresholds.  I'm quite certain sure I can feel punch with 100 dB transients.  By that point, things are starting to feel pretty substantial.  A good drum transient can deliver a pretty nice kick to the chest at and above that level.  With more low frequency (< 80 Hz) contribution, I think the slam becomes more visceral and seems to involve the whole body more.  I do think you need a little more SPL to get feeling from lower bass that matches the upper bass, unless you are getting vibrations from your seat.  But 110 dB is serious business, IMO, and 100 dB seems to sufficient to deliver some nice hits.  Dynamics are of course good thing whether we're talking about sound or feeling.

Link to comment
Share on other sites

Let me see if I understand your question.

Improved directionality - which indicates increased sound intensity, compared to improvements on frequency response.

I can have a look at the data, maybe it is possible to see something.

 

Well what I'm really wondering is if you can show that a change in velocity response can alter tactile perception even in the absence of a change in pressure response.  I'm not sure you can easily show that though because it looks like most of the thing that improve velocity response (better directivity, lower decay times, lower early reflections) also improve pressure response.  But maybe experiments can be done in an outdoor environment.

 

A note on decay:

What is important here is early decay - late reverb can not have any significant contribution to any tactile feel, because the level is down by at least 10-20dB or more, and that is too low in level to contribute any significant changes in the sound field intensity or velocity.

 

I'm not saying late reverb most certainly does not contributes positively to tactile feel.  I'm saying it may contribute negatively to tactile feel by masking it.  With hearing, I believe this is the case, even for signals at quite a low level.  It doesn't usually cause distortion or annoyance.  It is masking, so it primarily causes details to sort of fade into a blur.  IMO, It can still have a profound effect on the perception and It has the effect of raising the effective noise floor, especially if multiple bass instruments are playing at once.  It's worse in rooms with poor control of modal behavior, because the decay tends to be very un-even with frequency and the peaks correspond with high Q resonances that can take a very long time to decay compared to the rest of the sound.  Larger rooms with diffuse bass tend to be better, but there's still a loss of fidelity of the initial waveform in many circumstances, depending on the nature of the sound and what else is playing.  Anyone heard a rock band try to perform in a large concert hall designed for acoustic performance?

 

As for the 10-20 dB SPL figure you quoted, we have to be careful how we define level because you can get different numbers depending on what kind of analysis you are doing of your reflections.  If you are looking at the impulse response or ETC, you can't speak about the strength of reflections by looking at the numbers on those plots because it under-emphasizes low frequencies.  The strength of a particular reflection for a particular frequency depends on the *area under the curve* in the temporal vicinity of that reflection in the impulse response.  The width of that region increases as frequency decreases.  So ETC and IR are very good for seeing high frequency reflections and for getting a very accurate reading on the time it took for that reflection to arrive.  Low frequencies, don't show up unless the high frequency response is very clean because the absolute level in the IR is a lot lower.  The difference is that it doesn't oscillate as much, so the area under the curve can be just as great because the region is spread out over a wider region of time.  Does this make sense?  I realize this is a bit complicated, but the take away is that ETC and IR suck for low frequency analysis unless you use some kind of low-pass filter on the data.

 

Another way to get a dB SPL figure is to window the direct response up to before the first reflection and compare the smoothed SPLs between the direct and long-term steady-state responses.  That gives you a way to back-calculate the RMS level of the reflected energy in the steady state.  Or you could do some kind of time-frequency analysis like the spectrogram, waterfall plot, or decay plot and get dB values from there.  The actual values vary a bit, depending on window settings and things, but I find those tools to be very informative about what's going on acoustically across multiple frequencies.

 

Depends on the frequency, distance to the source must be large compared to the wavelength to be far field.

 

I will read through the rest of your post again later, and see if I have something meaningful to add, thinking about the perception part.

 

Yes you are right in this case.  We all have to be aware that the terms "near-field" and "far-field" are heavily overloaded in the sense that they mean different things in different contexts.  The differences are often subtle too.  I did some of my own research and educated myself on the theory of the acoustically small monopole radiator in free space.  Note that free space means no interference from boundaries.  So it's still relevant to ground-plane because the boundaries don't really interfere.  They just reduce the problem to a small subset of 3D space. 

 

With respect to the pressure response, there is no near-field or far-field response.  The statement "acoustically small" implies the radiator is much smaller than the wavelength so the radiation will be essentially omnidirectional.  This is what I was trying to say before, but I completely ignored the velocity response that discussion.  Now that I know better, I can confirm that the velocity response does in fact vary in a frequency-dependent manner through space.  Thus, there is a near-field and far-field.

 

In the near-field, acoustic impedance becomes substantially lower than the *characteristic impedance* of the medium.  The characteristic impedance of a medium is the acoustic impedance of a free-space monopole radiator in the far-field.  The characteristic impedance in a medium like air is equal to rho*c with rho being the density and c being the speed of sound in the medium.  Liquids and solids are much more dense than gases like air, and I believe most have higher speeds of sound as well.  Thus, liquids and solids typically have much higher characteristic acoustic impedance.  As one approaches the near-field of the monopole, particle velocity increases substantially.  Pressure increases also, but particle velocity increases *faster*.

 

This much would seem to imply that, in the outdoors where conditions are more like free-field (vs. inside, which is a totally different ball-game), tactile sensation may be better, for the same SPL, if one is standing in the far-field of the source.  However, we're not really done here.

 

The first problem is that impedance isn't just a single quantity.  When I'm using the words "higher" and "lower" above, I'm talking about the acoustic impedance *magnitude*.  Acoustic impedance has phase too.  For a good impedance match at a given frequency, the phases should be similar as well.  Out-of-phase is the worst case for transmission because the impedances are opposite each other.  As with electronic impedance, acoustic impedance can be described in terms of a resistive and reactive component.  In terms of circuits, resistors would be resistive and capacitors and inductors would be reactive.  This is another way of looking at the acoustic impedance and arises from a useful mathematical transformation.

 

The second problem is that we're not interested in the *percentage* of energy that is transmitted into the body.  Acoustic impedance matching affects the percentage of energy that is transmitted vs. reflected/diffracted at the interface between media.  However, I think we want to know the actual *quantity* of energy transmitted or intensity at the air-body interface, for a given SPL.  This latter point is important because we can always get higher intensity by turning up the playback level, adding more woofers, etc.  The application is room-treatment and/or other technology to improve in-room tactile perceptual response.

 

The third problem is that the monopole radiator is modelled as totally anechoic, and we are interested in modelling enclosed spaces.  Acoustically speaking, large boundaries act almost like ginormous acoustic radiators.  The consequences are that their near-field effects can extend well into the room.  The boundaries completely alter the parameters of the problem leading to a totally different wave equation solution and sound field that depends on the acoustic environment.  I'm still leaning toward thinking that acoustic impedance increases indoors.  With respect to problem #2, however, the sound intensity may drop faster, in which case, the energy transfer may still be lower in the high impedance area.  This is starting to get real tricky real fast.

 

Alright then.  Call me a big skeptic on the perceptual relevance of particle velocity, but I do have to admit that the physics are kind of intriguing while I'm learning.  I think we could dive into these problems in more depth with further discussion.  It just takes a lot of time for me to discuss these topics in the required depth.  Anyone, feel free to ask me for clarification on this or to point out where you think I may have gone wrong.  This stuff is getting pretty technical.

Link to comment
Share on other sites

In a room with less damping there will be boundary reflections from several surfaces, and with favorable speaker positions these reflections tend to statistically spread out and actually create a reasonably flat response. 

 

When most - but not all - of those reflections are damped, the remaining ones kind of stand out and produces huge dips in the response.

Since nulls can not be equalized, this has to be fixed with even more acoustic treatment.

 

Do you need more absorption or do you just need absorption in the right locations?  Interferences in frequency response due to early arriving energy tend to be relatively broad spectrum.  Wide peaks and dips in the bass suggest substantial early reflection energy.  Even with treatment, it's hard to absorb enough bass, but I remember in your case that you had a large window (?) on one side and couldn't possible do any side-wall treatments because the room was so narrow.

Link to comment
Share on other sites

Do you need more absorption or do you just need absorption in the right locations?  Interferences in frequency response due to early arriving energy tend to be relatively broad spectrum.  Wide peaks and dips in the bass suggest substantial early reflection energy.  Even with treatment, it's hard to absorb enough bass, but I remember in your case that you had a large window (?) on one side and couldn't possible do any side-wall treatments because the room was so narrow.

 

Absorption in the right places will improve things, how much remains to be seen.

The radiation pattern of the speaker I am using in this room now is difficult, this speaker is a down-sized version of something that works a lot better in the lower midrange.

 

I have already installed side wall absorption, and this has some effect on low frequency decay, but it does not improve the frequency response across the midrange.

A possible solution is to reduce the thickness of this absorption, some absorption is necessary because the direct sound from the opposing speaker illuminates the side walls.

 

The room must perform well from 100hz an up, the low bass already has good decay, and the bass response will be fixed by the bass system.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...