So if I am to summarise the current state of play in real world...

So if I am to summarise the current state of play in real world terms, the chip delivers single tones to 1 trilling cycle reliability up to 100Hz. What about it's reliability above the 100Hz threshold announced. I think AKP have problems with cone breakup and reliability above 100Hz.

What is cone breakup you ask?

Cone breakup refers to way in which piston drivers bend at the top end of their bandwidth. At lower frequencies the piston moves as a unit, but at a certain point based on the mass, stiffness and geometry bending modes cause sharp peaks in response and resonant ringing. At this point the driver will most likely become unusable. Stiff cones tend to have sharper breakup with a high Q peak that is easily seen in the response. Soft cones provide damping of cone breakup so that it may in some cases not be visible in the response.

 

The individual pixels in the wafer are still bound by the same laws of physics that affect cone breakup, just at the micro level.

I think beyond the 100Hz point the individual pixels experience "cone breakup" where the contacts meet the piston/diaphragm, and the chip is unable to produce high frequency sounds in time accordingly. I think this has always been an issue and the stiction was resolved by reducing mass and consequently this has made the piston weak and unable to be pushed/pulled at higher frequencies (around the edges where the connection is made, not the diaphragm itself). Imagine only half the pixel moving (i.e. moving sideways) producing an unreliable pulse that then impacts on the adjacent pixels, that are also doing their own thing. Depending on material consistency and resistance in contacts at the micro level, each pixel is literally doing its own thing. In large scale applications this this can be dealt with using accelerometers and/or lasers to correct, but the AKP chip has no such feedback circuitry, so AKP have to use a brute-force approach to try and resolve this issue.

It should also be noted that I think the chip is being used to produce simple waveforms, so the pixels are travelling in unison. This is no different to how USound and other conventional speakers work; so to say that the chip performs at unrivalled levels is misleading, and potentially false without some more information, not to mention it is not operating in the digital state with the 196kHz bandwidth it's supposed to have (i.e. it should be able to play a number of tones simultaneously and independently). I note that the USound chip produces 115dB output down to 10Hz flat to 1kHz, but as we know the USound chip is not an analog speaker - and AKP in their announcement have been specific in stating that, to exclude a comparison between USound and AKP. That is, AKP have compared apples to oranges to intentionally be misleading.

We're not there yet, AKP haven't even touched the surface when it comes to the algorithms and control system to be able to split and handle complex sounds. Remember, analog sound is fairly simple, a waveform goes into a speaker and it tries to follow that as best as possible, in the AKP world, the sound has to be processed and the ASIC has to decide how to best reproduce that sound. I still maintain that the chip is some way off from a demonstration. I note in 11/12/2012 announcement "As stated in the Chairman’s address at the Annual General Meeting held on 29 May 2012, theCompany is still on track and expects to have a playing demonstration unit available later thismonth." and 25/08/2014 "The company is progressing according to its plan to share and demonstrate the results,including music playing capabilities to select industry related audiences worldwide, during the2nd half of this year."

Yes, yes, I get it, we're closer, we're closer, but come on people - ask the tough questions! Each time there's a good news announcement AKP are doing a CR right before or right after, literally within a couple of weeks of each good news announcement, each time on a promise of some form. This time "While development is not yet complete, we believe these breakthroughs indicate significant progress the company has made toward delivering our breakthrough technology to the market." No promises here, yes they've learnt their lesson, underpromise and over-deliver, but if we are truly THAT close, the SP and volume would be much higher; rather than the FOMO behaviour it is currently experiencing. 

Criticise me all you want for stating the above, it is all my own opinion, and I do wish this technology to work, but the behavior of AKP has me asking these questions and looking deeper. Always interested in people's feedback but let's not count the chickens before we hatch the eggs (or however the saying goes).