Best way to mitigate floor (desk) bounce in near-field PC monitoring setup?

An ESS AMT-1 (December 2019 vintage) on- and off-axis SPL response plot. Red is on-axis, while the other traces are taken at 15 degree intervals. The blue trace is 45 degrees off-axis (IIRC).


Here is a polar sonogram, showing the need for wings below ~2 kHz.


I use my AMT-1s (single-high, then later dual-high) without wings crossed at ~700 Hz to Belle bass bins. Tilting back the top AMT-1s about 20 degrees approximates a "shaded array" [configuration (d) in the following figure from Toole's book] upwards (i.e., NOT in a floor bounce direction) to improve vertical polar response so enable a much wider vertical envelope than just the double-high AMT-1.


Not the smoothest response
I beg to differ. Those were all in-room measurements on top of a glass or hard wood surface.

With a small amount of absorption, the response gets even smoother. Note the vertical scale is in 2 dB increments--not 5 dB.

(If you're looking for an excuse to screen them out of solution, I think you're going to have to look elsewhere than SPL response.)

Mmm... another way to mitigate desk bounce, perhaps:

  • Low/no-baffle TM dipole, with high-dispersion tweeter (3/4" dome or equivalent) and relatively low-freq high-order crossover to the midrange (due to the severe "off"-axis requirement - see below)
  • Tilt the baffle toward the ceiling about 45 degrees or so. This should point the dipole null right at the desk bounce in most nearfield setups.
  • (Monopole) woofer can be right at desk level, up-firing if needed.
  • Because the speakers would be very close to the back wall, a substantial amount of absorption on the back wall would help with imaging (but not so much or so close as to actually prevent the speaker from operating as a dipole). Same with the ceiling; would be a good idea to hit the ceiling bounce with a small cloud of absorption "floating" above the tweeter/mid dipole baffle as well.

I don't know if I mentioned also about a very tight-dispersion Synergy horn as well, something like 60x40 or 90x40. Not everyone's cup of tea.

One more - using a really deep desk to make more space for an absorber right where the desk bounce happens. (You get more usable desk depth in front of the absorber in a 1:2 ratio as you increase the desk depth)
Last edited:
Was there no smoothing applied to that curve at all? Not even 1/12 octave?

It's on the plots--in the legend for each trace.

BTW: do you take in-room acoustic measurements using something like REW? Personally, I wouldn't try any of this without one of these and something like REW with a calibrated microphone for acoustic measurements.


The AMT-1 approaches I've tried are in fact spectacularly good...and all within "beer budgets". (I presently own 6 AMT-1s that I use for different purposes.).Trying to do this using compression drivers and horns to control directivity using one or more horns would increase the price to achieve similar performance rather dramatically.

And for me, direct radiator approaches are non-starters if you're trying to control directivity like you stated that you wish to control. All you wind up with is gobs of absorption and having to use large amounts of diffraction if trying to control lobe shapes to get your directivity...and limited acoustic dynamic range based on allowable modulation distortion levels at higher SPLs.