ESU has announced a new line of decoders, the V5. Sorry for the FB link, haven’t gotten a standard url yet https://www.facebook.com/389933821038209/posts/2222153197816253/ Some preliminary info: the new decoder will be capable of 50 khz BEMF control (today's decoders generally top out at 40khz); have a 32 bit processor, 128mb of memory, 16 bit audio with 31khz sampling (that sounds very weird as a sampling rate; 44.1 is the CD rate, and 48khz is used in a lot of audio production), and up to 32 functions. (Thx to @jdcolombo for the spec summary). I am sure we will be getting more info soon.
Being that the output for DCC sound are tiny little speakers (even for the larger scales) with mostly midrange sounds, 44.1 would be overkill as the ultra high-end sounds (which 44.1 and 48khz benefit from in the consumer & pro audio realms) couldn't be heard anyway (we're not listening to DCC sound on headphones or a 5.1 surround sound system). So a 31khz sampling rate would be more economical in terms of data. Even in the prototype world, most train sounds are low-end rumble and midrange mechanical/horn noises anyway. Ultimately, the bit rate matters the most in sound quality; a 16-bit sound decoder sounds a LOT better than an 8-bit sound decoder.
Here is a bit more official info (translated from the German announcement). The USA announcement, according to Matt Herman, will be coming soon. Conspicuously missing is any mention of the Select line.
