How much bass extension is really necessary?

[LogicprObe]

54 minutes ago, davewantsmoore said:

For example....  50Hz, is ~60 degrees out of phase with 100Hz.

 

Many people have argued that this phase distortion is not audible.... however the way I understand it, is that those arguments are flawed.

 

Mastering for vinyl is all in phase under 300hz and mono under 100hz with a 40hz roll off.

[frednork]

If I play tones I can hear down to around 14Hz, but realistically its probably not very noticeable in music below aound 25.  Although speaker spec is 38Hz  +-2dB in room boost takes it down to 10, Low frequency information changes the balance of the whole thing and can make a bright sounding system sound smooth. If your gear can take it allow it to go as low as it can.

 

BTW if I cut off everything below 35Hz I dont think I could be happy. Once its heard it is missed.

[davewantsmoore]

1 hour ago, LogicprObe said:

Mastering for vinyl is all in phase under 300hz and mono under 100hz with a 40hz roll off.

One would expect them to either use a shallower rolloff .... or to use linear phase filters .... so they don't cause the same outcome as the charts I posted.

[davewantsmoore]

1 hour ago, frednork said:

If I play tones I can hear down to around 14Hz

You are hearing the harmonic distortion.  ie. 28Hz, 42Hz

 

[frednork]

29 minutes ago, davewantsmoore said:

You are hearing the harmonic distortion.  ie. 28Hz, 42Hz

 

fair enough, then I cant hear harmonic distortion below around 14Hz with tones

[Al.M]

On 22/05/2020 at 10:04 AM, RoHo said:

We all know the audio spectrum is from 20Hz-20kHz.  I think it is a common understanding that for true, high fidelity reproduction the speakers/room should be capable of reproducing down to around 20-30Hz, though few can.  Most people would accept the compromise of less bass extension against the gains in speaker size and cost.    But we accept this is a compromise as we are trading off that "full-range" ideal.   

So I was thinking about the music I listen to - rock, folk, blues, country.  The lowest fundamental tone is the 42Hz from the bass guitar.  Maybe a kick-drum thud will be around that also.  But what else?  Dunno.  So would a speaker that is, say, 3 dB down at 40Hz be indistinguishable with my music to a "full range" speaker?

Sure, if you listen to large-scale classical, pipe organ or electronic music there is significant content in the lowest octave.  But for the large number of us who don't, is there anything to be gained by having a speaker that is flat to 25Hz?

Rod

For the majority of us I think we are limited by what we can commonly officially hear according to most science articles in the graph below within our age profile (middle aged men with reduced hearing to around 12-14khz but low frequency not that reduced as much); most commonly recorded CD music content not going much above those freq either.

 

Many of us have listened to many different speakers, looked at their manufacture frequency response figures, sometimes judging them correctly and other times not, trying out favourite test CDs to see if each can replicate low frequency, from there building up experience and feel for what sounds good.

 

Others have built their own speakers and subwoofer systems, designed them for the desired low frequency response, modelled and measured them and don’t rest until our favourite test CDs actually perform, getting exactly what you want.

 

For me the general experience is some speaker and subwoofer products despite having properly quoted low frequency manufacturers specs around 20-25Hz failed to pass the test in terms of low frequency bass satisfaction. In such cases one could hear I t going really low but there was no real impact or tunefulness = satisfaction.

 

In other products where the specs quoted were around 30-35Hz, even though it did not go as low, there was enough low frequency satisfaction. Some products despite having only 50-60Hz passed the satisfaction test also at only moderate loss of low perceived low frequency. 

 

Personally, I find when making my own speaker or subwoofer modelling and tuning the bass to dead flat lowest frequency doesn’t sound as good as having about 5-10Hz less extension and aiming for a slight hump or rise above the extension point.

 

Some examples of good bass - Proac speakers using Scanspeak 18w/8535 7 inch woofer in 33L ported cabinet 35Hz f3, twin Seas Excel 8 inch paper drivers in the VAF I-93, Eton 11 inch hexacone woofer in 75L ported, the venerable vintage Kef B139 oval woofer in 100L ported, many Ambience speakers using Seas 6 inch woofers ported.

Edited May 23, 2020 by Al.M

[RoHo]

15 hours ago, davewantsmoore said:

There are many things which have a lot lower frequncy content than the "fundamental" tone.

 

Short bursts of sound contain both very low and very high frequency content .... think of the "percussive" type sounds you can make hitting a string on a bass guitar.   Lots lower than 42Hz.

 

That being said, very low frequency tones themselves are not very audible.

 

Here is a simulated response of a speaker which is -3dB at 40Hz  .... If it weren't for the low frequency rolloff the phase would be zero at all frequencies.    

 

 

 

You've explained what I really couldn't ask about properly!  Are there sub-harmonics and are they important?   The phase issue is something I had never considered but does make sense.

11 hours ago, Al.M said:

For the majority of us I think we are limited by what we can commonly officially hear according to most science articles in the graph below within our age profile (middle aged men with reduced hearing to around 12-14khz but low frequency not that reduced as much); most commonly recorded CD music content not going much above those freq either.

 

Many of us have listened to many different speakers, looked at their manufacture frequency response figures, sometimes judging them correctly and other times not, trying out favourite test CDs to see if each can replicate low frequency, from there building up experience and feel for what sounds good.

 

Others have built their own speakers and subwoofer systems, designed them for the desired low frequency response, modelled and measured them and don’t rest until our favourite test CDs actually perform, getting exactly what you want.

 

For me the general experience is some speaker and subwoofer products despite having properly quoted low frequency manufacturers specs around 20-25Hz failed to pass the test in terms of low frequency bass satisfaction. In such cases one could hear I t going really low but there was no real impact or tunefulness = satisfaction.

 

In other products where the specs quoted were around 30-35Hz, even though it did not go as low, there was enough low frequency satisfaction. Some products despite having only 50-60Hz passed the satisfaction test also at only moderate loss of low perceived low frequency. 

 

Personally, I find when making my own speaker or subwoofer modelling and tuning the bass to dead flat lowest frequency doesn’t sound as good as having about 5-10Hz less extension and aiming for a slight hump or rise above the extension point.

 

Some examples of good bass - Proac speakers using Scanspeak 18w/8535 7 inch woofer in 33L ported cabinet 35Hz f3, twin Seas Excel 8 inch paper drivers in the VAF I-93, Eton 11 inch hexacone woofer in 75L ported, the venerable vintage Kef B139 oval woofer in 100L ported, many Ambience speakers using Seas 6 inch woofers ported.

Terrific post, thanks for that.

[davewantsmoore]

On 23/05/2020 at 11:59 PM, Al.M said:

 In such cases one could hear I t going really low but there was no real impact or tunefulness = satisfaction.

If there is "impact" or "tunefullness" on the recording .... then it was either the speaker, or the room, removing it.

 

On 24/05/2020 at 11:37 AM, RoHo said:

You've explained what I really couldn't ask about properly!  Are there sub-harmonics and are they important?

There's just whatever is on the recording.

 

...but you can see that with a speaker that has a -3dB at 40Hz .... then 40Hz tones are more than 90 degree (ie. significantly) out of phase with tones that are higher in Hz (eg. 120Hz+).

 

If a speaker response is not flat to very LF if affects the phase response many ocatves above.    You can use DSP to correct this (which is what some modern speakers, or systems use) ..... but it has the drawback of an overall time delay.

 

 

Also, if a speaker can "only just" make it to some low frequency, then it probably starting to distort significantly at that frequency.    Low distortion bass is a very important part of realistic reproduction.

 

[Pim]

On 25/05/2020 at 5:19 PM, davewantsmoore said:

 

 

...but you can see that with a speaker that has a -3dB at 40Hz .... then 40Hz tones are more than 90 degree (ie. significantly) out of phase with tones that are higher in Hz (eg. 120Hz+).

 

If a speaker response is not flat to very LF if affects the phase response many ocatves above.    You can use DSP to correct this (which is what some modern speakers, or systems use) ..... but it has the drawback of an overall time delay.

 

 

 

I think you've just explained how Devialet's Speaker Active Matching works. Any thoughts?

[andyr]

On 25/05/2020 at 5:19 PM, davewantsmoore said:

...but you can see that with a speaker that has a -3dB at 40Hz .... then 40Hz tones are more than 90 degree (ie. significantly) out of phase with tones that are higher in Hz (eg. 120Hz+).

 

If a speaker response is not flat to very LF if affects the phase response many octaves above.    You can use DSP to correct this (which is what some modern speakers, or systems use) ..... but it has the drawback of an overall time delay.

 

Dave, do you mean by this that if, say, a spkr's natural roll-off is -3dB @ 40Hz then if you use a HP filter to roll it off at, say, 80Hz ... you will have less 'out-of-phase-ness' @ 120Hz+?

 

And surely any overall time delay can be dealt with by the 'DSP unit' - thus mitigating any drawback?

 

On 25/05/2020 at 5:19 PM, davewantsmoore said:

 

Also, if a speaker can "only just" make it to some low frequency, then it probably starting to distort significantly at that frequency.

 

 

Absolutely!    And said distortion is why it makes sense to apply a HP filter to the spkr, to roll it off at a higher frequency (and allow a sub to take over from there).

 

Andy

 

[davewantsmoore]

15 hours ago, Pim said:

Any thoughts?

Devialet's Speaker Active Matching is a fancy name for "EQ".   The result is often greater low frequency entension.

 

 

Treating the speaker, amplifier, and EQ as one "closed system" like Devialet (and others) do, is very very sensible.    I think it's quite sad that audiophlies are left to match speaker and amps, and cables, and EQ (or unfortuntely likely not EQ) in the name of "system synergy" and other snowflake-ism.

[davewantsmoore]

One aspect of this that often surprises people.

 

As the wave/sound gets shorter (eg. as we approach a really short and sharp "click") .... it's bandwidth gets wider (contains more frequencies).

 

People often think of low frequencies as slow and long ..... but a very short sound .... contains lots of low frequency content.

 

 

As the length of the sound wave approaches zero .... the frequencies represented approaches infinity  (ie. right down to 0Hz)

[davewantsmoore]

13 hours ago, andyr said:

 

Dave, do you mean by this that if, say, a spkr's natural roll-off is -3dB @ 40Hz then if you use a HP filter to roll it off at, say, 80Hz ... you will have less 'out-of-phase-ness' @ 120Hz+?

No, you will have more (it will be worse).

 

Frequency and phase are just two different views of the same thing....  so when you bend the frequency more, you bend the phase more.

 

That is unless you employ filters (aka.  Linear Phase filters / Mixed Phase filters) which are able to modify the frequency and phase independantly.    Then you can correct the phase response to be flat, for any given frequency reponse.

 

 

 

 

13 hours ago, andyr said:

apply a HP filter to the spkr, to roll it off at a higher frequency (and allow a sub to take over from there).

Sure, same issue applies (wrt phase).... obviously in a well summing XO, the frequency response is flat.

[cazzesman]

Realistically,

 

How many of us could do critical listening from the sweet spot for 3 CD's with the vol over 85db?   110db of bass thumping your chest for 1 entire cd is just plain silly for us 50+ folk.    We are going deaf as it is let alone having Acadaca at 100db for an hour in our lounge room.

 

Honestly not many would survive the fatigue factor.     I doubt I could listen to an entire Tool CD at 100db in my sweet spot without passing out. 

 

IMHO opinion, for music audio only, it is not about the DB it is about the lack of distortion.   My Kii 3's at 85db are clean, clear concise.  That is plenty.    Carey still hits the sweet spot without caving in my chest.

 

Regards Cazzesman

 

PS......I am not talking about HT, just music listening.

Edited June 6, 2020 by cazzesman

[davewantsmoore]

8 hours ago, cazzesman said:

IMHO opinion, for music audio only, it is not about the DB it is about the lack of distortion.   My Kii 3's at 85db are clean, clear concise.

They sound clear not due to the lack of "THD" distortion..... but due to the axial frequncy response accuracy.   ie. "constant directivity".

 

8 hours ago, cazzesman said:

Carey still hits the sweet spot without caving in my chest.

   Isn't he marvellous!?  

 

 

This is why I think some sort of loudness EQ is very important for playback.   Something like ISO 226:2003 ..... so the bass level is kept appropriate vs the overall SPL.     On most systems ones you turn the wick up to get good/realistic sounding bass, then everything > a few hundred Hz is too loud, and above 1khz tears you a new one.  Ouch.

[davewantsmoore]

8 hours ago, cazzesman said:

distortion

In fact I'd never even looked at the THD type measurement for the Kii Three before ..... because THD is basically that irrelevant (it might show a big tell for bad design if it's out of control, but I know that's not going to happen on this speaker).

 

3.4% @ 90dB @ 100Hz.   Nothing special at all..... but that's about what you'd expect from inexpensive, small drivers.

 

As I said, the magic is in the frequency response (vs angle).

[GregoryS]

On 03/06/2020 at 11:38 AM, davewantsmoore said:

One aspect of this that often surprises people.

 

As the wave/sound gets shorter (eg. as we approach a really short and sharp "click") .... it's bandwidth gets wider (contains more frequencies).

 

People often think of low frequencies as slow and long ..... but a very short sound .... contains lots of low frequency content.

 

 

As the length of the sound wave approaches zero .... the frequencies represented approaches infinity  (ie. right down to 0Hz)

is this why clicks and pops (of various sources) are often full range?

[davewantsmoore]

19 hours ago, GregoryS said:

is this why clicks and pops (of various sources) are often full range?

Yes.

 

A very short sound contains a large number of different frequencies.    A very long sound contains a low number of different frequecnies.

 

At the extremes.....

 

An "infinitely short" sound (a "dirac pulse") contains all frequencies (in theory from zero to infinity Hz)

An "infinitely long" sound (for example a "sine wave") contains one, and only one, frequency.

[warweary]

Do the notes that we can't hear interact with notes we can hear? Like waves summing?

[hammerharry]

Is bass to sound what black is to colour?