SIDBasic by IstvanV, and me ( rest of work and ideas by Istvanv), software SID music player, it plays converted SID music on a pure Enterprise 128KB machine without any hardware addon.
The software is EXOS compatible. If your config has any file handling device ( floppy disk controller, SD Card ) then after pressing ESC file selection menu appears, if you have tape config just load the chosen M64 file.
It decompress 8KB data blocks during playing music, this was the key to be able playing full musics on a 128 KB machine.
Limitations: maximum M64 file length is 24320 bytes which can be loaded (5f00h), SYNC, High Pass, Low Pass filters are not implemented, digi effects can not be played back.
Combined waveforms are not working correctly: in this case the conversion is done with priority noise > sawtooth > triangle > square wave.

Controls:
1 - play music with internal Dave chip
2 - play music with external 8bit DAC card if exist
3 - raster bar on (it works only 50Hz playback speed musics)
4 - raster bar off
Esc - file selection

Music control speed is done by machine code cycles in this way control frequency is much more flexible than using Nick interrupts for this purpose.
Playback speed on 4MHz machine:
12500 Hz playback speed up to 50 Hz music control speed
11905 Hz playback speed 50 - 120 Hz music control speed
11364 Hz playback speed above 120 Hz music control speed

Youtube

[ Guests cannot view attachments ]

Hi!! Regards from your 8bit  CPC cousins at cpc-wiki forum....
on CPC system we are users of your old SIDplayer...
Nowadays I read about this new M64 files AY soundchip player on speccy side...

What about a CPC version ? You think port it is complicated???

Thanks a lot for your great work and research.

Long live to 8bit. Z80 cpu power forever!!!

It could be fast enough with some loss of resolution, perhaps with only 2 AY channels instead of 3. But this may still be slower than the Spectrum version, the official method would require 12 port writes. One channel can be written fast, however, the resolution would be worse than on the Plus/4. Another idea is to use one channel, but write it multiple times per sample to increase the effective resolution, this may or may not work in practice.

What happens if one channel is written in equal time periods ?

If done frequently enough, it is similar to writing separate channels, but not as loud. However, if the time period of the whole loop is too long, it produces an audible frequency. Three channels multiplexed at 1/30000 second intervals can be heard as 10 kHz.