Why We Love MORE BASSSS!

[almikel]

On 01/12/2020 at 1:45 PM, Satanica said:

 

I'm not understanding you.

 

Equal-loudness contours are curves that would be needed to correct to flat for perceived human hearing.

 

 

As per above: The ear is less sensitive to low frequencies, and this discrimination against lows becomes steeper for softer sounds.

Hi Satanica,

 

@odb123 was quoting @davewantsmoore in his post...what Dave is saying is if you look at the equal loudness curves below 100Hz, they get closer together as the frequency drops - so that if you were listening at say 90dB it only takes a small bump/dip to make it sound louder of softer to the ear, ie the ear is more sensitive to changes (bumps/dips) in the frequency response at lower frequencies

 

cheers

mike

[davewantsmoore]

On 01/12/2020 at 2:45 PM, Satanica said:

As per above: The ear is less sensitive to low frequencies, and this discrimination against lows becomes steeper for softer sounds.

Yes... but this usually confuses people no end.

 

A low frequency sound needs to be louder (than one at higher frequncy) to sound the same loudness.... or to indeed even be hear.

 

You can see from the chart that very LF at 80dB is inaudible.... barely audible at 90dB and .... still quiet at even 100dB..... This is before we even consider the masking effect of (much louder) sounds at 50Hz, or 100Hz, masking the practically inaudible sounds at 25Hz (for example).....    what people perceive as low bass at these volumes is mainly psychoacoustic.

 

BUT.

 

The real take away from this chart is that the ear is very sensitive to level differences at low frequencies.

 

 The range of loudness in the chart is spread over ~120dB at 1khz ... and only 70dB at 20Hz.   This means that bass is a million times more sensitive to changes in level (for the same change in perceived loudness).

 

Eg. a change in level of ~20dB at 20Hz.... is equivalent to a change in 50 phon (phon is what you hear).

 

 

SIMILARLY.... The claim that the ear is most sensitive at 3khz.   Yes, 3khz needs less energy to sound the same loudness as other frequenies.....   But it is (relatively) low sensitivity to changes in energy level in terms of phon.

 

Sound (information) perception is all about changes in level, measured in phon.    So, these comments essentially get it backwards (for people concerned with audio reproduction quality).

 

 

This leads to a fairly big paradox that what we can accurately measure and easily fix for a speaker in a room, actually isn't very significant (in terms of phon) ..... and what is a big deal (bass, eg. << 500Hz) is both dificult to accurately measure, and to fix.

 

 

 

[Satanica]

17 hours ago, almikel said:

 

@odb123 was quoting @davewantsmoore in his post...what Dave is saying is if you look at the equal loudness curves below 100Hz, they get closer together as the frequency drops - so that if you were listening at say 90dB it only takes a small bump/dip to make it sound louder of softer to the ear, ie the ear is more sensitive to changes (bumps/dips) in the frequency response at lower frequencies.

 

Hi Mike, yes I think I understand that but where did Dave specifically mention "equal loudness curves below 100Hz" ?

[Satanica]

3 hours ago, davewantsmoore said:

Most people don't understand the consequence of this curve.  (aka Fletcher Munson Curve) https://en.wikipedia.org/wiki/File:Lindos1.svg

 

..... and lots of people say things which gives the opposite impression from what is true.

 

Apparently the above is quoted from you.

 

So I'm understanding you, can you please give me an example of what you mean in bold?

If it is something I said/wrote then quote me on it.

[akjono]

6 hours ago, davewantsmoore said:

Yes... but this usually confuses people no end.

 

A low frequency sound needs to be louder (than one at higher frequncy) to sound the same loudness.... or to indeed even be hear.

 

You can see from the chart that very LF at 80dB is inaudible.... barely audible at 90dB and .... still quiet at even 100dB..... This is before we even consider the masking effect of (much louder) sounds at 50Hz, or 100Hz, masking the practically inaudible sounds at 25Hz (for example).....    what people perceive as low bass at these volumes is mainly psychoacoustic.

 

BUT.

 

The real take away from this chart is that the ear is very sensitive to level differences at low frequencies.

 

 The range of loudness in the chart is spread over ~120dB at 1khz ... and only 70dB at 20Hz.   This means that bass is a million times more sensitive to changes in level (for the same change in perceived loudness).

 

Eg. a change in level of ~20dB at 20Hz.... is equivalent to a change in 50 phon (phon is what you hear).

 

 

SIMILARLY.... The claim that the ear is most sensitive at 3khz.   Yes, 3khz needs less energy to sound the same loudness as other frequenies.....   But it is (relatively) low sensitivity to changes in energy level in terms of phon.

 

Sound (information) perception is all about changes in level, measured in phon.    So, these comments essentially get it backwards (for people concerned with audio reproduction quality).

 

 

This leads to a fairly big paradox that what we can accurately measure and easily fix for a speaker in a room, actually isn't very significant (in terms of phon) ..... and what is a big deal (bass, eg. << 500Hz) is both dificult to accurately measure, and to fix.

 

 

 

Exactly! What....?? You raised many points Dave, I will just address this: Yes I totally agree the ear is very sensitive to level differences at low frequencies. And the "Claim" that the ear is most sensitive at 3 khz is an average, or median measurement. Like the seats in airplanes. Does not fit any one human on the planet, but is engineered for an "average" of the total human stature possibilities. In other words=fits no one. 

 

Also very cool experiment @QuinnInSydney. .. See I am not crazy, or... we are both crazy in the same way! hehehe. 

 

The paradox is that every human experiences sound in a different way, and that for some of us we need 20 Inch 1000Watt Subwoofers to feel the Bass (20-40Hz I will just say). Like a hit on the head with a 4X6 ! for some of us, and others will enjoy 12 inch woofers that dig to 38Hz, but punchy and clean. 

 

Psychoacoustics may be another way to say one "feels" the bass. And hence the music. 

 

Righteous. 

 

[Satanica]

7 hours ago, davewantsmoore said:

BUT.

 

The real take away from this chart is that the ear is very sensitive to level differences at low frequencies.

 

 The range of loudness in the chart is spread over ~120dB at 1khz ... and only 70dB at 20Hz.   This means that bass is a million times more sensitive to changes in level (for the same change in perceived loudness).

 

Eg. a change in level of ~20dB at 20Hz.... is equivalent to a change in 50 phon (phon is what you hear).

 

 

SIMILARLY.... The claim that the ear is most sensitive at 3khz.   Yes, 3khz needs less energy to sound the same loudness as other frequenies.....   But it is (relatively) low sensitivity to changes in energy level in terms of phon.

 

Sound (information) perception is all about changes in level, measured in phon.    So, these comments essentially get it backwards (for people concerned with audio reproduction quality).

 

 

This leads to a fairly big paradox that what we can accurately measure and easily fix for a speaker in a room, actually isn't very significant (in terms of phon) ..... and what is a big deal (bass, eg. << 500Hz) is both dificult to accurately measure, and to fix.

 

Thinking about this some more I presume you mean the ear is is more sensitive to frequency response differences at low frequencies than at non-low frequencies.

If so, I think I'm understanding you and combined with the fact that frequencies below the transition region\zone (300Hz?) are dominated by the room rather than the speaker is why good bass as perceived by the ear is harder to achieve than non-bass frequencies.

 

BUT I don't think this is the point or purpose of the ISO 226:2003 curves as inspired by Fletcher-Munson.

The point is that even if you have a speaker system with a flat frequency response as measured by an acceptably accurate microphone the human ear won't perceive it as such.

Rather, how the human ear will perceive it will depend on how loud (phons) it is and an averaged approximation graph has been developed as such.

 

[almikel]

6 hours ago, Satanica said:

 

Hi Mike, yes I think I understand that but where did Dave specifically mention "equal loudness curves below 100Hz" ?

he didn't - that was me trying to explain Dave's point that people often miss the point that the human ear is more sensitive to changes in level for lower frequencies. I picked 100Hz because the equal loudness curves clearly get closer together below 100Hz, and the closer together they get, the more sensitive the ear is to changes in level.

 

Mike.  

[almikel]

6 hours ago, Satanica said:

Apparently the above is quoted from you.

 

Dave wasn't quoting me, but @davewantsmoore will often quote a post to make his point - and he wasn't disagreeing with anything in the post, just using it to highlight that people don't understand the subtleties of the Fletcher Munson curves - ie they understand that bass needs to be louder to be of equal loudness, but they miss that the equal loudness curves get closer together at lower frequencies, which means the ear is more sensitive to changes in level as frequency drops....

...I think Dave answered everything in his post

 

mike

[almikel]

2 hours ago, Satanica said:

Thinking about this some more I presume you mean the ear is is more sensitive to frequency response differences at low frequencies than at non-low frequencies.

yes 

 

2 hours ago, Satanica said:

If so, I think I'm understanding you and combined with the fact that frequencies below the transition region\zone (300Hz?) are dominated by the room rather than the speaker is why good bass as perceived by the ear is harder to achieve than non-bass frequencies.

reading @davewantsmoore's post - I agree that's what he's saying - for bass frequencies the room makes a really big impact on the dB phon levels, with the ear being very sensitive to small changes in dB phon levels...but also that we can't measure very accurately at low frequencies.

 

 

Edited December 4, 2020 by almikel
clarification from Dave

[davewantsmoore]

2 minutes ago, almikel said:

for bass frequencies the room makes a really big impact on the phon levels

You mean to say ....   "a really big impact on dB levels".

 

Increasing the sound loudness by X phon sounds like the same increase in loudness at every frequency.... that is the whole point of a Phon.

[almikel]

1 minute ago, davewantsmoore said:

You mean to say ....   "a really big impact on dB levels".

 

Increasing the sound loudness by X phon sounds like the same increase in loudness at every frequency.... that is the whole point of a Phon.

thanks Dave

[davewantsmoore]

6 hours ago, Satanica said:

So I'm understanding you, can you please give me an example of what you mean in bold?

 

People say that the "ear is more sensitive through 1 to 3Khz" or whatever ......  and they take from that that any differences in the frequency response between playback systems are most important in this region (due to the sensitivity of the ear).     But this is not true.   The ear is least sensitive to features of the frequency response in this region  (ie. the complete opposite).

 

For example hifi reviewers will be talking about a speaker, and saying well there's a bump or a dip, or whatevr in the "critical treble range", where the "ear is most sensitive" (so it will definitely be noticable).   For example.

 

I'm not sure if you said anything specific.

[davewantsmoore]

1 hour ago, Satanica said:

BUT I don't think this is the point or purpose of the ISO 226:2003 curves as inspired by Fletcher-Munson.

The point is that even if you have a speaker system with a flat frequency response as measured by an acceptably accurate microphone the human ear won't perceive it as such.

 

The purpose of it is to demontrate the relationship between Phon and dB .... and it tell us a number of different useful things.

 

Yes... one is that the lines are not flat.   This means that for the sound to be perceived as flat, then when represetned in dB it should not be flat.   It will need to tilt up at LF and HF.

 

Also ... it shows that it will need to tilt more when the sound is quieter.... which most people are familiar with too.

 

Another implication of this, is that changing the level in dB results in the least amount of change in Phon through the upper-midband..... from about 800 to 4000Hz..... meaning that although it takes the lest amount of sound energy to reach some perceived level (Phon) at these frequencies, that the ear is actually least sensitve to changes in the level at these frequencies.

[davewantsmoore]

3 hours ago, akjono said:

The paradox is that every human experiences sound in a different way

 

That hinges very much on what you define as "experience".

 

If you mean "the phon scale fits nobody"   ...ie. we all perceive levels differently..... then I think that is much more nonsense than truth.    (of course, there are always special snowflakes)

 

What people "like" is a different matter entirely 

 

[almikel]

1 hour ago, Satanica said:

BUT I don't think this is the point or purpose of the ISO 226:2003 curves as inspired by Fletcher-Munson.

The point is that even if you have a speaker system with a flat frequency response as measured by an acceptably accurate microphone the human ear won't perceive it as such.

Rather, how the human ear will perceive it will depend on how loud (phons) it is and an averaged approximation graph has been developed as such.

this delves into the issue described by Toole and Olive as the "Circle of Confusion" - paraphrased -  "that there can be no standard of sound reproduction since there is no standard of production from the recording and sound engineering end.... If the audio systems that were used to mix and master the recordings had a poor frequency response, the sound engineers will be creating a sound mix that sounds good on that particular system, but when that sound mix is reproduced on a system that is far more accurate and linear, it can sound odd."

 

Mixing engineers will always adjust the levels to cater for Fletcher Munson - but we don't know how flat the FR was in the mastering room

 

[almikel]

2 hours ago, Satanica said:

The point is that even if you have a speaker system with a flat frequency response as measured by an acceptably accurate microphone the human ear won't perceive it as such.

As audio enthusiasts with control only over the "reproduction" end of the process, I still think a flat room frequency response is a good goal...obviously difficult to achieve.

 

Good recordings will typically be recorded/mixed in good recording/mixing rooms - and we all have examples of "well recorded" music, which IME get played on my system far more often than poorer recordings.

 

Of course there are metrics other than a flat FR which are important to great "in room" sound - eg IMHO the room's time domain response (ie ringing/reverb times), and a smooth "off-axis" response of the speakers are very important.

 

cheers,

Mike

[Satanica]

On 04/12/2020 at 6:18 PM, almikel said:

Mixing engineers will always adjust the levels to cater for Fletcher Munson - but we don't know how flat the FR was in the mastering room

 

 

For sure. Regarding Fletcher-Munson, what I gathered from Wiki was that experiment was done on headphones in the 1940's and the most modern ISO 226:2003 from 2003 is quite different and was done with speakers. Both of these curves should really be referred to as Equal-Loudness Contours.

[Pin]

Improve the bass and the mids and highs improve disproportionately 

[almikel]

21 hours ago, Pin said:

Improve the bass and the mids and highs improve disproportionately 

That's certainly been my experience - get the "in room" bass right, and as long as you haven't killed the top end during the process (ie soaked up too much treble), you can sit back and enjoy clean/tight bass and further room treatment isn't required once the "in room" bass is under control...

 

I will make a clarification on my post above

On 04/12/2020 at 7:25 PM, almikel said:

I still think a flat room frequency response is a good goal...obviously difficult to achieve.

I meant an "even" FR - I run a room curve with a downward tilt - not flat - around 6dB of boost down low reducing to around -6dB of cut up high from 20Hz to 20kHz, fairly straight...

 

In my lightly constructed room, all the low bass leaks out, so with the absorption I have in my room, I can achieve "reasonable in room bass" with the addition of a few bands of EQ cut below 150Hz or so.

If I had a more rigid room that reflected low bass back into the room, I'd have much bigger challenges to manage the bass in my room.

 

IMO target absorption treatment at the lower frequencies as the only priority - avoid absorption treatment at 1st reflection points and straddle as many corners with as deep/wide absorption as you can get away with - this will put the absorption where it's soaking up bass, but not killing the treble.

 

Absorption traps get very large to absorb low frequencies - and along the way of applying enough absorption to manage the "in room" bass in a room , you could easily have absorbed too much top end.

 

Covering corner traps with a membrane (eg builders plastic) or timber slats in a 1D BAD pattern will reflect treble back into the room, but continue to absorb bass.

 

The more rigid the room boundaries are, the harder it is to achieve "smooth/tight in room bass"

 

IMO great "in room" bass is awesome - tight/dry/accurate bass with no overhang/boominess is fantastic...loose/boomy/ringing bass just gets turned down.

 

cheers,

Mike

[Pin]

Absolute sound noted that records in the seventies had rolled off bass and with the variance between channels really low bass was not an issue

[Pin]

Sorry it was stereofile originally published in 1975