[50jess]

Has anyone tried putting rockwool in-between speakers or on floor?
http://www.harbeth.co.uk/usergroup/showthread.php?1295-Treating-floor-bounce-from-your-speakers-to-your-ear

[JohnR]

I tried putting down a couple of cotton bolsters, and it seemed to make response (measurably) worse. I didn't label the measurements but I'm pretty sure this was it - nothing on floor in blue, with bolsters in red:



I can't explain the result, but my take away from this little experiment is that a complete approach will be needed - fiddling with things like reflections here or there isn't going to be the way to do it.

[TRADERXFAN]

Just wondering if that bounce was factored in to the crossover design. Thus taking it away lowering output?

[Vapor Audio]

I put panels of Roxul AFB down on the floor when taking measurements, usually about 12" - 16" thick is what it takes to get rid of floor bounce.  And it will get rid of it completely. 

But I've tried treating the floor for listening, and have trouble hearing any difference at all.  The perceptual studies explain why, because the first impulse is much more important in how we perceive sound. 

[neekomax]

Also, we've lived with floor bounce all our lives... maybe we're able to filter it out?

[bpape]

Just wondering if that bounce was factored in to the crossover design. Thus taking it away lowering output?

I would think that any serious manufacturer would design the speakers to account for the fact that they'll either be on the floor or on a stand of a recommended height.  They can't know much else about a specific room but that should be a given. 

Bryan

[JohnR]

In my case they're not off the shelf. The crossover are at 100 and 800 but the region affected is in the middle of the dipole compensation eq. I suppose I may have done it differently but it's hard to see how without actually doing it.

I'm just providing a data point

I put panels of Roxul AFB down on the floor when taking measurements, usually about 12" - 16" thick is what it takes to get rid of floor bounce.  And it will get rid of it completely. 

After seeing your post I had a look at the impulse response. The bolsters made a difference but didn't remove it completely (around 46.5 ms). The reflection at 48.5 is bigger though - I can't tell at the moment where that is, perhaps the ceiling.

[Vapor Audio]

To tell which impulse wiggle belongs to whatever surface, use V=X/T where V is velocity, X is distance, T is time.  Speed of sound is 340 m/s, plug in whatever time in milliseconds seen in the impulse to get your distance.

That's a very busy impulse response btw.  Looks like a 3-way with very large offset differences.  I don't know anything about the speaker you're measuring, but just looking at that impulse I'd say there's a lot of potential performance gain to be had by cleaning things up in the time domain.  Tons of reflections in that impulse too.

[JohnR]

It's a four-way dipole.

Thanks for the math lesson.

[face]

I put panels of Roxul AFB down on the floor when taking measurements, usually about 12" - 16" thick is what it takes to get rid of floor bounce.  And it will get rid of it completely. 

Take a 6" bass trap and raise it off the floor another 6", it will have almost the same effect.  Works great for me as I just take one off the wall and put it back when done measuring. 

But I've tried treating the floor for listening, and have trouble hearing any difference at all.  The perceptual studies explain why, because the first impulse is much more important in how we perceive sound.

I have the opposite results.  Leaving said bass traps on the floor in front of my speakers makes everything sound dull.  Your results may be different due to the different dispersion characteristics of ribbons.