How to determine perceived volume

[brownt]

I would like to determine the perceived volume for a range of sound effects with differing filters, resonance, harmonics etc. I suppose looking at the amplitude on an oscilloscope is no good because of the ears differing response to frequency.

Is there a way the perceived volume can be determined?

 

[c_mitra]

determine the perceived volume for a range of sound effects determine the perceived volume for a range of sound effects...

Do you mean the intensity (sound pressure difference) when you speak about volume? Also, perceived by whom? A normal untrained ear or a calibrated microphone? You want to consider only the range of audible sounds?

 

[brownt]

Do you mean the intensity (sound pressure difference) when you speak about volume? Also, perceived by whom? A normal untrained ear or a calibrated microphone? You want to consider only the range of audible sounds?

I mean the perception through a human ear of how loud a sound is. Frequencies are guitar based, so the fundamentals are about 80 - 1400Hz. But with distortion effects and other effects the harmonics and different waveshapes are many.

 

[betwixt]

The usual lab method is to use a microphone and linear voltage measurement through a narrow filter then to mathematically apply a weighting figure according to the filter frequency. For example, to measure at 12 spot frequencies and apply the corresponding adjustment to each so you can derive an overall figure or a graph of perceived audio spectrum.

If you have the luxury of a signal generator as the source, it is easy because you can omit the filter and use a constant level sine tone instead but if you have to measure under 'live' conditions, you might have to average the readings over a second or so at each frequency.

The other method, used on most small sound pressure meters is just a microphone, a shaped filter to mimic human hearing response then a peak detector.

There are several different models for human hearing but they only differ slightly, you should be able to find one with an internet search.

Brian.

 

[c_mitra]

perception through a human ear of how loud a sound is....

Perception is difficult to quantify but the response of the human ear is approximately logarithmic. A sine wave at 100Hz with 1V amplitude and another sine wave at 1000Hz with 1V amplitude will not appear equally "loud".

The response beyond 1kHz is more complicated but that is where the music lies.

In your range of interest, you can assume that the response is logarithmic. But "real" sound effects are not single frequency and you need to use a spectrum analyzer to study the distortions.

 

[FvM]

I understand that the question is primarily about frequency dependent loudness perception. It can be represented by equal loudness curves https://en.m.wikipedia.org/wiki/Equal-loudness_contour
Sound level meters are using respective weighting curves, most popular A-weighting.

The contour curves are for pure sine tones, loudness perception of complex sound signals is more complicated. It should be also mentioned that the level of electrical signals does not necessarily produce a proportional sound pressure, you need to consider the transducer frequency response.

 

[Audioguru]

Your age and quality of your hearing also affect the response of distortion and harmonic effects. Old and deafened people cannot hear high frequency distortion and harmonics without hearing aids.
I suspect that many rock music producers are deaf to high frequencies and put too much distortion and harmonics into their music making it sound like noise.

My hearing was tested to be normal for my age when I was 69 (no high frequencies). Hearing aids make my hearing good like when I was young. Rock music sounds good without my hearing aids but is noise with them. Normal music sounds great with my hearing aids.

 

[c_mitra]

Old and deafened people cannot hear high frequency distortion...

I am only 68 but I cannot hear anything above 9kHz...

 

[crutschow]

Sound level meters, such as the one Radio Shack used to sell, can give an sound level reading in dB that approximates what the ear hears in loudness, allowing you to make a reasonable comparison of the sound effect levels.
They typically have two different weighing functions (filtering) labeled A & C, depending upon the meter application.

 

[Audioguru]

I am only 68 but I cannot hear anything above 9kHz...

Your high frequency hearing loss is normal or worse for your age. Here is a graph of normal hearing loss with age. The loss is worse for people who were exposed to loud noise or loud rock "music":

 

[FredCailloux]

I would like to determine the perceived volume for a range of sound effects with differing filters, resonance, harmonics etc.

I use to sell sound analizing equipments year ago. From what I remember, you would need an audio range spectrum analyser to gather the raw data. Then with the specturm data you could create your own computer analysis with a simple microcontroler such as arduino. The tricky part of your project would be the calibration. Determining what is loud and what is not would greatly vary from one individual to another. Like others have mentionned, perception is very individual dependent. Still, It would certainly be possible to come up with an apparatus that delivers at least a gross analysis of the situation.
I would attack the task as such:
1- Microphone pick-up
2- Spectrum analyzer (could be done with a high speed mControler and lots of math) ( portable spectrum analyzers are expensives)
3- MicroController to analyze the raw data. This is where you would implement your analysis logic, such as peak perception, frequency range, pulse duration, tone monotony, back ground noise...
4- Some kind of graphic output such as an inexpensive little RGB display.

Another much simpler way to implement such "Meter" could be to build an active 5 bands microphone logarithmic function amplifier and give a measurement of each center frequency in real time. Although simpler, it would be usefull as a primitive audio evaluation tool.
Attached is an example of a Log function OpAmp implementation. May be usefull for your project (this was taken from Wikipedia)
https://en.wikipedia.org/wiki/Log_amplifier

 

[c_mitra]

Your high frequency hearing loss is normal or worse for your age. Here is a graph of normal hearing loss with age....

Thank you and thank you again. I was wondering whether I am going deaf...

The curves show that women make excellent hearers (even with age, far better than men) and they would make wonderful listeners...

- - - Updated - - -

1- Microphone pick-up
2- Spectrum analyzer (could be done with a high speed mControler and lots of math) ( portable spectrum analyzers are expensives)
3- MicroController to analyze the raw data. This is where you would implement your analysis logic, such as peak perception, frequency range, pulse duration, tone monotony, back ground noise...
4- Some kind of graphic output such as an inexpensive little RGB display.

Except for the first, the rest can be done in software- even a PC or Laptop should be fine...

Perhaps audacity (I used it very long ago) can be used with some tweaking...