Sound frequency and dampening

1398342003 · Post by **1398342003** » Fri Jun 06, 2003 11:38 pm

Is lower pitched sound more difficult to dampen than higher pitched sound?

If so, than a 50DBa 60MM fan would be easier to dampen than a 50DBa 120MM fan, with an 80MM one in the middle, right?

Yomat · Post by **Yomat** » Sat Jun 07, 2003 12:14 am

Can of worms.

Generally speaking the fan relations probably holds. But some fans might have other characteristics.

As for sound dampening you know that there are 'blocking' and 'absorbing' types? Different densities of material absorb and reflect different frequencies. High freq sounds are easier to 'block' than to 'absorb'. Although some high frequencies seems to cut through blocks easier than others. Perhaps they escape and bounce through air vents more readily due to the shorter wavelength. This might be the reason why 'inside case dampening' products might not work that well.

Successful computer dampening materials has both blocking and absorbing properties. Some also have success with combining highly reflective surfaces combined with absorbing ones. This one from Mazar: http://www.mirar.org/silent-pc/.

1398342003 · Post by **1398342003** » Sat Jun 07, 2003 2:06 pm

So what would be easiest to dampen/block, a 120mm, a 92mm, an 80mm, or a 60mm? (assuming of course that each produced the same DBa)

It sounds to me that the 60mm fan would be easiest, but I'm probably wrong.

Yomat · Post by **Yomat** » Sat Jun 07, 2003 11:35 pm

I agree with Radeonman.

The bigger the fan the less nose for the same airflow. However.. if you have too big a fan you might not be able to adjust downwards as much as you like.

And as for damping. It might be easier to block high frequency sounds. But as I said they escape much easier. And low frequency sounds are MUCH more pleasant to the ear. We are all noise freaks but not all sound is bad.

A low wooshing sound can be soothing, almost. Perhaps they should market fans that varies the rpm and make it sound more like natural wind.

halcyon · Post by **halcyon** » Sun Jun 08, 2003 10:16 am

Actually basic acoustical engineering states that the depth of the damping material has to be at least one quarter of the wavelength of the frequency of sound, before the material starts to have a discernible dampening effect.

So, higher frequency sounds are easier to damp, as you need less thickness for the same amount of absorption (material being the same)

As for blocking, it is very difficult. Basically it would require an extremely rigid and dense material with absolutely no seam holes or structural vibration in the blocking material for blocking to really work. Of course, even less than perfect blocking helps, but it usually needs absorbation in conjunction to be really useful.

With fans, the rotational speed usually defines the main frequency of the pitch of the fan. If the CFM is the same, the smaller diameter fan has to have a higher RPM to push the same amount of air as the bigger one.

This means not only a higher pitch (easier to damp), but also louder noise (harder to damp).

The choice of where to put the main frequency of your fans is complicated more by the equal loudness contours of human hearing.

Two sounds with equal intensity (measured in dB Pa) do not sound equally loud to the human ear (measured in phons or sones).

To complicate matters even more, some people consider different frequencies variyingly annoying at the same audible loudness.

So, if you could pick where to place that sound frequency, you'd place it somewhere that:

- is easier to damp (higher the better)
- is not easy to hear (very high or very low, definetely not between 2,5-6kHz where the hearing is most accurate)
- is the least irritating to you personally (after the damping)

I hope that helps.

regards,
Halcyon

1398342003 · Post by **1398342003** » Mon Jun 09, 2003 4:42 pm

I wasn't thinking of the CFM/DBa ratio, I was only thinking of the effect of frequency on dampening.

Thanks for the replies, this is very nice information to have.

Post by **MikeC** » Mon Jun 09, 2003 4:58 pm

Lots of educated answers. I see that acoustical knowledge among SPCR forum members is rising.

- is not easy to hear (very high or very low, definetely not between 2,5-6kHz where the hearing is most accurate)

Agreed on all points except range: our hearing is *generally* about equally sensitive from about middle A (440Hz) or even lower all the way to ~8kHz. Notwithstanding natural deterioration due to aging.

As mentioned by halcyon, people have their pet peeves with noise. I have found that while the high pitched whining of a hard drive or small fast fan is very annoying, the droning hum of a 120mm fan whose vibration is getting into the case is not far behind for annoyance, even a fairly low levels. Think about the rhythmic boom boom or thud thud sound of rap or similar from a stereo that is being played as much as a house or two away, but on a very bass-capable system. Only the bass sound is audible. Even when it is not loud, after a while, it is enuf to drive one bonkers.

tragus · Post by **tragus** » Mon Jun 09, 2003 5:47 pm

MikeC wrote:Agreed on all points except range: our hearing is *generally* about equally sensitive from about middle A (440Hz) or even lower all the way to ~8kHz.

To amplify

what halcyon mentioned and MikeC alluded to, let me suggest to anyone feeling especially geeky about the ear's sensitivity to different frequencies check out the canonical isoloudness curve set that graphically demonstrates the human auditory system's perception of different frequencies at different dB SPL (sound levels).

Note, however, that these curves were made using pure tones. Noise, by its nature, is a complex sound made of *many* frequencies. As several folks point out, the annoyance factor varies by person and by context. To wit, "noise" has a technical meaning of some semi-dense set of frequency components with randomly varying phases and amplitudes all summed together. It also has a psychological meaning of some sound that is perceived as annoying, grating, or otherwise unpleasant or even painful. Colloquially, we tend to mix the two meanings though it's useful to remember that one is a physical stimulus and the other a perceptual phenomenon. Similarly, level and loudness are respectively the physically measurable sound pressure change (dB SPL) and the human perception of the "intensity" of the sound (phons).

That said, we can look at the isoloudness curves to get a general sense of which frequency region is likely to be most detectable and/or perceptually salient. Note that our loudness perception flattens across frequency as level increases. I'll put on my really geeky hat and say that a major piece of information missing from all reviews everywhere in computer-land is spectrum analysis of the purported sound. "34 dB(A)" helps, though it simply sums the spectral amplitudes of the sound using the isoloudness curves as relative weights (strictly, often reporting RMS); it doesn't give you the exact distribution of the frequency components, however, which will greatly affect the quality or timbre of the sound. Thus one 34 dBA may be background and not unpleasant, but another 34 dBA will be unbearable. I could make a much more informed judgement and provide quantitative comparisons if provided with the spectral profiles of products.

halcyon · Post by **halcyon** » Mon Jun 09, 2003 11:28 pm

To nitpick (OffTopic):

Actually the range from c. 2,5 kHz to 6 kHz is the most sensitive for all intensitity levels (as currently studies). It is usually 3 - 10 dB more sensitive than from 400 Hz - 2,5 kHz or from 6-8 kHz, varying as a function of intensity.

This is where the outer ear resonances are placed, which amplify the sound for those frequencies. Naturally the outer ear and especially the ear canal length is an anatomically variant feature and as such, there is no exact figure for the most sensitive region for all people. Hence the quite wide range of 2,5 to 6 kHz. One could also say 1,5 - 6 kHz. It's a matter of preference and/or application where the dB difference is considered significant enough.

Or to be more precise the ear canal resonance frequency is between 3-4 kHz (the utmost most sensitive hearing region for all intensity levels), but outer ear amplifies sounds from between 1,5 till 7 kHz. This is further complicated by the basilar membrane tuning in the innear ear, which complicates matters for sounds above 6 kHz (after which a sharp decrease in treshold of basilar membrane activity occurs).

However, if you study the Fletcher-Munson equal loudness curves and the later ISO variants, the most sensitive reagion becomes apparent. What is interesting to note is that the 500 Hz dip (lowering of a treshold) does not occur in the studies of other researchers after Fletcher-Munson. As such, even the currently often cited ISO 226 equal-loudness contours appear not to be correct for the lower mid-region of sounds. The preliminary revised ISO curves also show this. This also suggests that the 400-500 Hz region is not as sensitive at equal levels as the approx. 2,5 - 6 kHz region.

Furthermore, the middle ear acoustical reflex can attenuate sounds below 1 kHz up to 14 dB (after 80 dB excitation). This does not affect hearing at same degree at the most sensitive region, raising the relative difference between these two regions even further

Please also remember that the human sensitivy for intensity changes (or apparent loudness) varies as a function of base intensity, tone duration and frequency.

The often misquoted audible difference treshold of 1 dB is really an overgeneralisation. Human auditory system can detect intensity differenced down to 0.1 dB, depending on the intensity of the sound range and the frequency/duration of sound. This is applicable for higher loudness levels however and would probably not be important for fan selection, unless you're using industrial type 240 volt fans with 80 + dB of noise :)

As a last point (I promise not to embark on these rambling in the future), the audibility of a complex sound (containing more than one sinusoidal - like almost all natural sounds) is also dependent on the frequency width of the sound, other variables being equal.

In practise, a sound that is spread over a larger frequency spectrum is usually heard more easily than a narrow frequency sound. However a wide frequency sound may be less annoying alone than a ringing narrow frequency sound. This is also relate to what Tragus was discussing (the frequency distribution of complex sounds).

Acoustics and especially psychoacoustics is really interesting, but after a while it makes my head hurt. And I'm seriously thinking of going into a doctoral program in this field :)

Cheers,
Halcyon

Refs:

Precise and full-range determination of two-dimensional equal loudness contours, http://www.nedo.go.jp/itd/grant-e/kokusai/00is1-e.htm

Factors influencing equal-loudness level contours,
http://www.physik.uni-oldenburg.de/docs ... 82328.html

An introduction to Psychology of Hearing, Brian CJ Moore, Academic Press, 2003.

Psychoacoustics, Zwicker, Fastl, 2nd ed., Springer-Verlag, 1999.

Introduction to Audiology, Martin, Clark. Pearson, 2003.

Signals and Perception, Roberts (ed), Open University, 2002.

Post by **MikeC** » Mon Jun 09, 2003 11:34 pm

Thanks for the enlightenment, halcyon. I stand cheerfully corrected.

tragus · Post by **tragus** » Tue Jun 10, 2003 5:06 am

Hmmm, halcyon, you're singing my tune.

The first and most important lesson I teach my students:

For all X, X is complicated.
Corollary: For all X that's complicated, it's even more complicated than you think it is.

Indeed, my link above was to the (slightly updated) Fletcher-Munson curves since it's the most cited (hence canonical) even if it's now shown to be inaccurate. A friend of mine is actually on the ISO loudness committee and that Japanese work has been a driving force for labs around the world to re-examine the standards. As you point out, the new isoloudness curves are still a work in progress.

To one point in halycon's excellent post, "loudness" tends to increase with bandwidth beyond a critical band. Audibility (aka masking) depends on a large number of factors, not just bandwidth of the signal. To wit, common modulation (either AM or FM) of signal, level and frequency distribution of masking/background noise, spectro-temporal characteristics of the signal itself, etc. The Moore book that halcyon cited is the probably best overall survey of the field and the set of issues, while the Zwicker book is a bit denser and getting a tad dated. One bottom line example: you'll hear your fan much more as it spins up OR down versus when it's at a steady state. If it's RPM varies systematically over relatively short time (e.g., voltage variation), you'll also hear it more easily.

If ya want some experiential advice on graduate program in psychoacoustics, PM me. As the rest of the readership can see, there's a reason psychoacousticians have no friends

1398342003 · Post by **1398342003** » Thu Jun 12, 2003 12:36 am

A little off from the topic, but where would dampening be best used, on the intake side or the exaust side?

Sound frequency and dampening

Sound frequency and dampening

Equal loudness Contours, just noticeable difference (ampl.)