Why is floor bounce considered only a bass issue?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Why is floor bounce considered only a bass issue?

Shouldn't this kind of cancellation affect all frequencies with adequate vertical dispersion? :scratch2:

And also, why is the floor bounce frequency always referenced as singular and not plural? I would imagine more than one narrow band of frequencies being potentially affected.
 
Fanatic
Joined 2009
Paid Member
Last edited:
Just thinking aloud, in any room except an anechoic chamber you will have lots of bouncing around, lots of interaction, peaks and dips allover the place; the analysis being VERY complex.

Now there is *one* important response-altering effect, usually unavoidable unless you float the speaker a significant distance from the floor, which is floor bounce.

It´s both important and relatively simple to analyze, so people do.

And given the distances involved, it will happen strongly in the Bass frequencies realm, not that other frequencies are not affected but this one is very significant.
 
Well, the lowest frequency is subject to the least absorption at the floor, so it will make the deepest notch at the listening position.

Wall reflections are a different issue again - you have two sources spread out in the horizontal plane, so perfect cancellations are more rare.

Chris
 
Moderator
Joined 2002
Paid Member
Shouldn't this kind of cancellation affect all frequencies with adequate vertical dispersion? :scratch2:

It does.

The answer is in a combination of room modes and transition between half space and omni radiation.

The biggest issue obviously is the actual height of the midwoofer from the floor.

If the speaker is designed to radiate a good amount of power using the floor (down-porting, low-slung ports or woofer close to floor), the primary floor bounce dip is easily eliminated.

Other frequencies cannot be treated in the same way.

So the effect is universal, but the cure can only be applied for bass frequencies.
 
Why is floor bounce considered only a bass issue?

Shouldn't this kind of cancellation affect all frequencies with adequate vertical dispersion? :scratch2:

And also, why is the floor bounce frequency always referenced as singular and not plural? I would imagine more than one narrow band of frequencies being potentially affected.

It affects more frequencies, but at a specific frequency the reflection lags exactly half a wavelength behind the direct sound which creates a big dip in the frequency response. At higher frequencies it is more of a discrete reflection, several wavelengths behind the direct sound. At lower frecuencies the reflection goes away as the reflection is in phase with the direct sound.

One can design a system to reduce this, with a line array, or a low mounted woofer crossed to a midrange far from the floor....

Floor bounce also happens with real sources, so it is possible that our hearing is well adapted to it, and don’t perceive this as unnatural. Low frequencies is also perceived from the entire body, not just the ears, so a single point measurement may be exaggerating the effect.
 
Thanks for the clarification everyone! Makes more sense now.

One can design a system to reduce this, with a line array, or a low mounted woofer crossed to a midrange far from the floor....

Played around with a calculation for a low mounted woofer and a high mounted midrange now. Pretty neat how you can mitigate the issue this way.
Kind of hard to keep driver spacing within half a wavelength this way though.

Regarding the line array approach - I guess the idea is to cover the dip of the taller mounted drivers with the lower mounted drivers?
 
Last edited:
Moderator
Joined 2002
Paid Member
Pretty neat how you can mitigate the issue this way.
Kind of hard to keep driver spacing within half a wavelength this way though.

Something always gotta give.

Low crossover helps somewhat - 200 to 250Hz is considered the acceptable maximum.

A low port also helps. Not as much as a low slung woofer, but it does mitigate the issue somewhat.

A very large diameter bass driver also works, a 12-15" bass driver can be crossed to a typical 6-7" mid quite easily and be both close to the floor and keep a reasonable C-C spacing.

I chose a 10" woofer and favored closer C-C spacing for better vocal integrity, then used a downfiring port and overdamped the bass a bit in the crossover. This got me a good balance between low end and midbass, with a very good lower vocal range as well.

Not saying that's the only way out - there are others, such as a 2.5 way system with the lower driver down low. C-C spacing is then no longer an issue.
 
Gradient 1.4 by late Jorma Salmi. Downfire bass and a coaxial.
Finnish Gradient 1.4 loudspeaker
Gradient 1.4 loudspeaker, Jorma Salmi | Inner Magazines

gradient-14.jpg
 
Last edited:
Why is floor bounce considered only a bass issue?

Shouldn't this kind of cancellation affect all frequencies with adequate vertical dispersion? :scratch2:

And also, why is the floor bounce frequency always referenced as singular and not plural? I would imagine more than one narrow band of frequencies being potentially affected.

"Floor bounce" is the term used to refer to the interaction of the direct sound and the first reflection off of the floor. I think this became a "popular thing" when people started measuring loudspeakers and found this huge dip in the lower midbass and then thought "Holy cow, this is a seriously BAD response!". But, as many people have pointed out, the sound undergoes many reflections and many returns from the room reach the listener. This tends to even out the floor bounce when you consider the hearing process and what is perceived by the brain. A microphone measurement is not the same thing as the ear-brain hearing process and this is a great example of when the "mic" is not a good indicator of what you hear. So don't worry about it too much. IMHO room modes are more of a problem.

You are correct in that this phenomenon can effect all drivers. But it depends on how far above the floor they are, and what their passband is. Take a madrange - if this is located high above the floor and the midrange is also crossed high, the dip frequency is below the lower edge of the passband and the midrange response would be completely unaffected. Another way to think about is that it is affected, but below the driver's rolloff, and therefore doesn't impact the passband.

Roy Allison (I think it was him) once outlined a design principle in which you put the woofer at the floor, and the midrange as high as possible, e.g. 40" above the floor or more, and cross over between W and M at 350Hz or above. Just as the floor bounce would be below the midrange's passband, the woofer's floor bounce will be too high in frequency and out of it's passband as well. The frequency of the bounce is inversely proportional to the distance above the floor, and also the woofer's diameter spans from the floor up to 10" or 12" above the floor and this vertical "length" helps to smear out the floor bounce making any dip shallow to non-existent. This is also the case I believe when multiple woofers are stacked vertically and adjacent to one another. The only downside to the Allison design is that the W and M are acoustically far apart at the crossover point, so instead of floor bounce you now have lobing in the crossover region to worry about.
 
Last edited:
Only rarely does a topic here interest me these days, but floor bounce is a goodie. :)

Amidst all the trade-offs of speaker performance, getting the big concert-hall sound in a smallish room is highly interesting, IMO.

A historical note about floor bounce comes from Roy Allison who gave it some thought. TBH, not knowing about Roy Allison and Edgar Villchur in speakers is like not knowing about Beethoven in music.

A Glorious Time: AR's Edgar Villchur and Roy Allison Allison Part 1 | Stereophile.com

The accepted wisdom is you put the woofer near the floor and cross it around 300Hz. You then mount the mid and tweeter section much higher up, where it is clear of the floor bounce or "Allison Effect".

Now your midrange is purer and clearer. An additional trick is to use an MTTM dispersion for the projecting PA-type sound, which has less vertical dispersion, so more immune to refection off the floor and ceiling.

A lovely wall-mounted Allison IC20 speaker on theoretical grounds:

427330d1404675679-quintessentially-german-loudspeaker-70th-modern-interpretation-allison-ic20-speaker-jpg


427331d1404675679-quintessentially-german-loudspeaker-70th-modern-interpretation-allison-crossover-jpg


Much to like there.
 
  • Like
Reactions: 1 user
Remember the effect of listening distance too!

I did just that in my clone, LR2 xo at 300Hz. Room measurements with long gating always show multiple reflections and modes...

The accepted wisdom is you put the woofer near the floor and cross it around 300Hz. You then mount the mid and tweeter section much higher up, where it is clear of the floor bounce or "Allison Effect"

With a 1 meter listening height and midrange placed at that same height, floor bounce cancellation happens at between 250 Hz and 350 Hz with a typical listening distance of 2,5 to 3,5 meters.

This means the woofer should be crossed a bit above this point, probably around 500 Hz to cover it. Which is why I'm puzzled by Gradient's low crossover at 200 Hz.

Crossing over at ~500 Hz would also be right in the area where you would get baffle step loss with a typical sized baffle, so you could fix that as well.

Any input?

An additional trick is to use an MTTM dispersion for the projecting PA-type sound, which has less vertical dispersion, so more immune to refection off the floor and ceiling.

Why MTTM and not MTM?
Won't the extra tweeter do more trouble than good? (lobing)
 
Folks, floor bounce just is not that much of a problem unless your listening space is a huge, huge space and speaker are far, far away from walls and ceiling.

In a domestic listening space, the additional multiple reflections off the walls and ceiling "fills in" the dip that you get when you only consider the floor reflection.

Calling "floor bounce" a problem is mostly an academic argument that only applies to certain listening spaces that are not typical of a domestic listening environment.
 
  • Like
Reactions: 1 user
A lovely wall-mounted Allison IC20 speaker on theoretical grounds:

427330d1404675679-quintessentially-german-loudspeaker-70th-modern-interpretation-allison-ic20-speaker-jpg

That speaker has many good design features, like using two woofer with one reversed. But the two-tweeter thing is really not good!

Also, I believe the general consensus is that speakers on/in walls have poor soundstaging and this stems from the lack of the same reflections from room boundaries that give rise to the various dreaded "bounces" and cancellations...

So, on balance, the design has got it's warts, too.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.