Ann: AGM Presentation, page-66

Yes, but in the presentation image show above the colour code scale is still in dB, and its maximum coloured areas are at 32dB.

The output profile is centred on what I've assumed is the MEMS chip's axis, rather than an individual pixel's axis. I think the rest of the SW titling (i.e. separation distance scale in mm) indicates it is set up for a chip array. So the images are for individual chips, but these form part of an array that is coordinated to act as a speaker.

It looks like the software view shown allows the experimenter to explore the sound output from the individual chips in the array shown top left (working chip viewed has red dot) at different frequencies while visualising/viewing the output at different angles and planes. Output is pictured as the complex 3D coloured image in the top centre field. Different patterns will be generated for differing frequencies, giving a whole series of different cross sections such as pictured for one particular frequency, 14,481Hz.

I believe the coloured profile (centre bottom) shows a view of the sound coming out from the front of the chip as a sound intensity profile (SPL) viewed as a "top view" i.e. looking directly down from above the chip standing on its edge, but seen only as a thin cross-section slice in the horizontal yellow plane shown. This plane could be inclined away from the horizontal if desired.

The coloured profile at right is also a cross-section of the horizontal yellow plane (red line, 0deg) and again seen as a thin slice when viewed on edge from the RHS.

The Frequency Response curve on the left bottom is a "cross section of the coloured cross section" looking down through the horizontal yellow plane in the off axis vertical plane at 55.7deg (red line). The dips in the graph show the SPL of the sound intensity at different frequencies. The coloured stripes correspond to the drop-off's of sound at various frequencies, at angles from the chip axis. This graph shows the relationships between frequency and SPL that are represented by the colour coding. In other words a whole continuum of coloured cross sections exist for all frequencies, but this graph lets the user see the whole data set along a plane just for a single subset of the data.

This indicates that at 14,481Hz most of the emitted sound is travelling straight out from the front of the array, with far less being emitted off axis. That is presumably the subject of the histogram to the right that has been deliberately blurred. The orange coded rings that the off axis line crosses correspond to maximum SPL's between 10-20dB, vs >30dB on axis. 14,481Hz is the second highest peak going up the frequency range after the 32dB plateaux for lower frequencies. I am guessing that the histogram shows a few representative frequencies to show how effective the directional capabilities of the chip array are overall.

As far as I can tell, while not being familiar with this particular software, this signifies that the chip is being operated in a mode that puts out a narrowed forward facing sound beam with little other directivity. One presumes this can be adjusted to angle that beam in any direction, as indicated by the "Directivity Control" settings. It is also set to have (maximum?) effect at 1m in front of the speaker array. One can also presume that every chip can be programmed to do the same thing, or groups can do DIFFERENT things, depending on some overall setting on the ASIC. i.e. the sound can be split by the array into multiple beams that can be sent to focus on different places in the room where listeners have been detected, or send soundtracks in different languages to those people. This is all made possible by DSR.

This is a pretty good indication of where the company is at I think.