Since floor-bounce is an echo like any other, following test was developed to verify how much floor-bounce a dipole planar magnetic really produces.
A "full range" cone speaker in sealed enclosure was positioned 20" high at 2 meter from the microphone. The height of the microphone was 1 meter. This measurement was performed indoor, size of the room is 40D x 25 W x 20 feet high. A regular MLS impulse measurement was performed, the impulse response was subsequently magnified and analysed in search for the floor-bounce echo. Simple math shows that the echo can be estimated to arrive less than 2 milliseconds later. Since floor bounce is not particular pin-point precise, some tolerances can be expected.
The MLS impulse shows the first major echo arriving at 1.833 milliseconds after the main impulse. (delta T from top to top.) Observe how the floor echo follows the first pulse trend, only at a much lower magnitude.
A second impulse response was taken with the full range cone speaker placed 12" higher. A quick guestimate tells us that the the time shift should be less then .7 milliseconds. The second plot shows us the suspected floor bounce echo at 2.5208 milliseconds after the first impulse. A delta T of 0.68783 milliseconds. So far correlation, it is possible to figure out were to expect the floor bounce echo. (Be aware that the time scale is different both plots.)
The full range speaker was replaced with a RD75 on baffle. Center of the ribbon was approximately at the same location of the previous test with the speaker. One would expect to see any sign of floor bounce showing up in the same range. The plot below shows us the first 4 milliseconds of the ribbon impulse response. Magnification of the impulse response is the same for all plots.
No sign of a floor bounce echo. Which confirms my previous experiences with planar magnetic ribbons. The cilindrical wave front remains cylindrical pretty far out there. The above measurement was performed at 2 meter. Prior observations have shown that some bounce can be expected at 4-5 meter distances with ribbons that are installed straight up.. (microphone at 1 meter high.) Slanting the ribbons slightly backwards is sufficient to remove this aural cue and is more SAF than soundproofing the floor in front of the dipoles. Slanting back has the additional advantage of smoothening out the frequency response between 1 and 2.5 kHz. The amount of smoothing however is function of listening distance to the ribbon and width of the baffle.