Why is Weak Bass recommended here?

[Mnyb]

I agree a bit on KEF’s typical shelf they do it an all models except reference where it can be tuned by port inserts.
Or it can possibly be done with R series to ? Someone told me ?
They are trying to second guess your room gain and compensate to not make the bass a muddy mess. And on some models you have different port plugs to make some changes .
I would still prefer a more classical approach and let me fix it with my room correction.
You can possibly achieve a lower tuning with thier way of doing it , but seriously you should use subs and room correction anyway .

I must be damaged somehow, I don’t think I ever in any room in the circa 14 different places I lived in experienced “to much bass” so I also have bit of struggle to understand this tendency to design speakers with lean bass

I’m ofcourse considering bying KEf as my next speaker and simple get a model larger than I need and boost the low end until it’s subjectively ok for me

More equipment should have a real loudness function IE physiological volume control but calibrated to a level like RME does it ( otherwise it don’t really work and becomes a party button like on a 70’s stereo )

 

[ZolaIII]

Boundary reinforcement means in-room response extension is much better than the predicted response. The room and placement within it indictate much of the bass response of a given speaker.

Here is the KEF R3 in my room (F3 at 35Hz):

With a little DSP help, I can extend that down to 32Hz. That is low enough for the music I tend to enjoy:

And, here it is vs. PIR:

I would not call this weak bass and do not see a problem here.

No sorry not even Eron do it right.
Take a look at response brakedown and closed ports one in this case:

KEF R3 meta Measurements

Hi, I measured the KEF R3 meta loudspeakers this time. These speakers have just been released, but I was able to get a pre-order sample (a pair) from an audiophile in Korea. Impedance Frequency response I measure HF responses using a 5ms window gating. Therefore, the...

www.audiosciencereview.com

In some cases there is no port response register what so ever in the measurements and in other there is to much and it's not influenced by number or size of woofer in the speaker.
I propose response brakedown and port's closed response when ever possible (it won't be always possible with odd shaped port's or pasive radiators).
Honestly you can't either expect small diameter cones to go that low nor you want and need port cuffing when you are crossing it with a sub/sub's.

 

[-Matt-]

Just use Audessy or any room correction you have to correct the response back to anechoic levels.

When using speakers - yes I would correct them with Audyssey.

...But I'm talking about comparing the measurements from the speaker reviews. It wouldn't be easy to load these into my AVR!


I'd like to see an approximate visual representation of the ideal bass response (assuming a sub crossed at 80Hz) as it would appear on the predicted in-room graph. Because (due to room reinforcement) apparently my straight dashed blue lines would give far too much bass.

Maybe, just maybe, the ideal bass response is that shown by the Kef R3. And maybe it needs exactly that shelf shape to compensate for room reinforcement in an average room (whatever that is). If so, any larger speaker, in that average sized room would require eq to cut bass. If that is the case then this is the form that the theoretical curves would have. It isn't necessary that the speaker response matches these curves - if it meets or exceeds the curve for your given room size then EQ will probably be able to correct it. You would also be able to read off the graph the magnitude of EQ corrections required, and see if these were within your device's capability.


Ideally I'd like to see theoretical 80Hz crossover curves for a few different room sizes superposed over the measurements as an aid to the eye.

 

[Spocko]

Ok, and I'd like to know how a speaker's predicted in room response would need to look, in order to best compliment this.

if you crossover at the the monitor's bass roll-off as designed to compliment subwoofers, the predicted in-room response graph still goes low enough to be relevant. Using the Perlisten S4B as an example, it's designated roll-off frequency +/- 1.5db is 100Hz. Why would you want the crossover frequency to be so high? Because bad room interactions requiring corrections/treatment/DSP take place below 120Hz and limiting correction software to that cutoff leaves the monitor's performance above 120Hz mostly untouched for those who prefer that option.

The speaker's predicted in-room response below 120Hz will have zero bearing on reality because every room will have deep nulls & peaks so for all practical purposes, such information is unusable IMHO.

 

[Spocko]

I'd like to see an approximate visual representation of the ideal bass response (assuming a sub crossed at 80Hz) as it would appear on the predicted in-room graph. Because (due to room reinforcement) apparently my straight dashed blue lines would give far too much bass.

It's worse than that! You're assuming uniform room reinforcement and not taking into account isolated peaks that could see a rise of +8 dB in one specific frequency range (ie, 95Hz to 100Hz) and nowhere else.

 

[fineMen]

... talking about comparing the measurements from the speaker reviews. It wouldn't be easy to load these into my AVR!
...
Ideally I'd like to see theoretical 80Hz crossover curves for a few different room sizes superposed over the measurements as an aid to the eye.

Won't bother you anymore. One last, though. Three different, quite plausible, in no way extreme listening positions in my room compared (levels recalculated):

That's a vast difference in bass versus mids balance. 10dB or so. And that's the same room, the same speaker position. And it's the R3s, no e/q, room size is like 70sqm, quite high ceiling ...

I leave it to you. I'm not 'AI' trained to argue honest people

 

[Sancus]

Kef's slight bass rollofs are likely intentional and because they're designed for European construction which tends to be denser than in North America, so you get more boundary gain. They're actually trying to make the bass flat instead of overwhelming. They offer different physical ports in the Reference series to allow customizing that.

Realistically speaking you absolutely must have the ability to EQ in order to get correct bass in any setup, so you can just EQ them flat to 40hz if they're not in your room.

I've been looking back over the reviews of some of the most recommended speakers and one thing they seem to have in common is really weak bass.

Huh? The 8361A is probably the most strongly recommended speaker on the site and it has incredible bass for its size. The KH150 is another recent review and it also has incredibly good bass for its size.

Bass is a very important component of the preference score, so if a speaker has a poor score despite having good bass then it's probably quite bad in other attributes.

But I doubt that is part of the equation for calculating "with sub" preference scores. Or is it?

"with sub" scores assume you have a 100% perfectly integrated subwoofer that goes perfectly flat to 20hz or below and that the bass response of the speaker is sufficient to create such a perfect crossover, the crossover frequency being rendered irrelevant.

 

[-Matt-]

"with sub" scores assume you have a 100% perfectly integrated subwoofer that goes perfectly flat to 20hz or below and that the bass response of the speaker is sufficient to create such a perfect crossover, the crossover frequency being rendered irrelevant.

Thing is, it isn't irrelevant at the high frequency side. I.e. It makes a difference if your sub only needs to go up to 80Hz or if it has to go all the way up to 250Hz. But I don't think this is taken into account by the "with sub" preference score.

 

[sigbergaudio]

@-Matt- your problem feels made up. Basically all speakers have a bass extension that is sufficient for an 80hz crossover to subwoofers. So you don't need to look at estimated in-room responses to try to determine that.

 

[-Matt-]

@-Matt- your problem feels made up. Basically all speakers have a bass extension that is sufficient for an 80hz crossover to subwoofers. So you don't need to look at estimated in-room responses to try to determine that.

The "problem" is then, why isn't that evident from the predicted in room measurements where the bass appears to roll-off well before 80Hz?

...and can the review measurements be used by those who don't want to use a sub to identify speakers that offer good bass extension.

 

[Sancus]

The "problem" is then, why isn't that evident from the predicted in room measurements where the bass appears to roll-off well before 80Hz?

...and can the review measurements be used by those who don't want to use a sub to identify speakers that offer good bass extension.

I think you have some misconceptions about the reviews here:
1) "with sub" is not part of the review. Amir reviews speakers as standalone products.
2) pref scores are also not part of the review, Amir does not use the score or calculate it. That's done by other forum members. His recommendation does not follow the score, but is rather his personal call based on the measurements and his listening test.
3) the "with sub" score in general just assumes that you do what you need to do to get a sub working properly. If there are factors that prevent this, then the score isn't valid. It's that simple.

The "problem" is then, why isn't that evident from the predicted in room measurements where the bass appears to roll-off well before 80Hz?

A speaker doesn't need to be flat at 80hz to cross it over, it just needs to have sufficient output. At the crossover point, sub and speaker are working together, neither has the full load. None of the speakers you've linked to in this thread have a problem with an 80hz crossover.

Yes, you need some knowledge of how to read measurements to get full value out of the reviews. That's normal.

 

[-Matt-]

I think you have some misconceptions about the reviews here:
1) "with sub" is not part of the review. Amir reviews speakers as standalone products.
2) pref scores are also not part of the review, Amir does not use the score or calculate it. That's done by other forum members. His recommendation does not follow the score, but is rather his personal call based on the measurements and his listening test.
3) the "with sub" score in general just assumes that you do what you need to do to get a sub working properly. If there are factors that prevent this, then the score isn't valid. It's that simple.



A speaker doesn't need to be flat at 80hz to cross it over, it just needs to have sufficient output. At the crossover point, sub and speaker are working together, neither has the full load. None of the speakers you've linked to in this thread have a problem with an 80hz crossover.

Yes, you need some knowledge of how to read measurements to get full value out of the reviews. That's normal.

Well, I agree with all of those points and still think it would be handy to have some curves superposed over the predicted in-room response that indicate the point at which the bass response is likely adequate.

 

[amirm]

The room dominates the response in bass region, not the speaker. You can't make any judgements about that region from measurements. The PIR gives clues as to tonal balance, not exact frequency response in a real room. It represents a room with no room modes.

Bass response has been corrected in Klippel for probably 2 years worth of reviews. It was an early issue that was caught and fixed. Now it produces more accurate response than anechoic chambers. And at any rate, it only impacts towers with distance between ports and speakers, not bookshelves and such.

 

[AwesomeSauce2015]

Looking back at OP's original post, I would like to make a few things clear:
The reason why the B&W speaker wasn't recommended is because the directivity, and on-axis frequency response were horrible. On the contrary, Genelec, KEF, and Revel's speakers have good directivity, and good on axis frequency response.
In my opinion, and in the opinions of many listeners, having a speaker with good directivity, and a flat on axis frequency response is better than having a speaker put out 20hz lower bass, especially when subwoofers are a thing.

One other note is that KEF and Revel design their speakers to be used with subwoofers, with the exception of the KEF Blades, and probably the Revel Salons. That is why the bass rolls off early, since they are trading deep bass extension for more SPL capacity. Bowers and Wilkins probably would rather have 10hz lower extension than having the SPL capacity, since any high output levels will never be used anyways (since the B&W speaker sounds bad).

Amir's recommendations seem to be based on whether or not a speaker is a good speaker. Basically, does it have a good (flat) frequency response, and does it have good directivity. After those are considered, then things like dynamics, bass output, and other factors can come into play.

So basically, weak bass is not always recommended here. I, and many others, are advocates for having a fairly strong bass presence, and when I personally tune my systems I will run the subs a bit higher than the technical "reference" level, because I like how it sounds.

 

[Colonel7]

Thing is, it isn't irrelevant at the high frequency side. I.e. It makes a difference if your sub only needs to go up to 80Hz or if it has to go all the way up to 250Hz. But I don't think this is taken into account by the "with sub" preference score.

This has the formulas that go into the preference score
Thread 'Speaker Equivalent SINAD Discussion'
https://audiosciencereview.com/forum/index.php?threads/speaker-equivalent-sinad-discussion.10818/

 

[RayDunzl]

Raw vs flat EQ here at home at the listening position, 1/3 smoothing, left and right, 15" wide electrostatic panel with 180Hz passive crossover to 12" sealed woofers, no subs.

I don't know if the hump is only the room (don't think so) or not. Don't have other measures of this exact speaker.

The spec (from 1998) says they give 23Hz, and they do.

 

[-Matt-]

This has the formulas that go into the preference score
Thread 'Speaker Equivalent SINAD Discussion'
https://audiosciencereview.com/forum/index.php?threads/speaker-equivalent-sinad-discussion.10818/

The basic preference score does include a term for bass extension, which is good.

But unless I miss it, that link doesn't say anything about how the "with sub" score is calculated.

If the assumption is that a perfect sub will take over at whichever point the speaker response rolls off, then my hypothetical, very smooth, tweeter only speaker will still be able to get a very high "with sub" preference score.

By simply not producing bass or midrange at all, the hypothetical speaker benefits from a "with sub" preference score that assumes perfect bass and even perfect midrange. Unless someone can refute this by showing how the equation for "with sub" preference score specifies a maximum upper limit for the assumed ideal sub output?

 

[JustJones]

The basic preference score does include a term for bass extension, which is good.

But unless I miss it, that link doesn't say anything about how the "with sub" score is calculated.

If the assumption is that a perfect sub will take over at whichever point the speaker response rolls off, then my hypothetical, very smooth, tweeter only speaker will still be able to get a very high "with sub" preference score.

By simply not producing bass or midrange at all, the hypothetical speaker benefits from a "with sub" preference score that assumes perfect bass and even perfect midrange. Unless someone can refute this by showing how the equation for "with sub" preference score specifies a maximum upper limit for the assumed ideal sub output?

My speakers cross the tweeters at 3.5 Khz. I'm not sure I understand what you mean by high " with sub" preference score. I don't see how you come to your conclusion when the speaker is only being measured above 3.5 Khz. I don't know of a subwoofer that goes that high. I guess I'm missing something.

 

[NTK]

I believe this is how @pierre's with-sub preference score is calculated:

spinorama/src/spinorama/compute_scores.py at master · pierreaubert/spinorama

A library to display and compare spinorama (speakers measurements) graphs. - pierreaubert/spinorama

github.com

The lf_x (low frequency extension) for the with-sub score is fixed at -4.31 * log10(14.5) = -5.0 for all speakers.

[Edit] Added the missing negative sign to the formula to calculate the lf_x score. With the negative sign, a speaker with a lower low frequency extension (the -6 dB point) will give a larger (less negative) lf_x number, which raises the preference score.

 

[jonfitch]

Even if bass performance is not amazing, it's still one of the easiest things to correct for since the bass frequencies are near omni-directional.