[hatari-devel] Re: Hatari YM conversion

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

As this might be of generic interest, I'm answering to the list.

On lauantai 26 lokakuu 2013, Douglas Little wrote:
> Do you know much about the YM software synth inside Hatari?
> 
> I'm curious if it can be used standalone, to generate raw WAV data into
> a buffer.

Hatari can save the played data to a file in WAV format, and like
other things, you can control this from within emulation with XBios(255).
- Sound recording is toggled on/off with "hatari-shortcut recsound"
- Sound save file is set with "hatari-path soundout <path/filename>"

Save format is decided based on filename extension, it should be
either ".wav" or ".ym". 


For standalone YM conversion, it might be better to use some of the projects
from which Hatari's code originates from, though.  According to sound.c:
-----------
  The new version of the sound core uses some code/ideas from the following 
GPL projects :
    - tone and noise steps computations are from StSound 1.2 by Arnaud Carré 
(Leonard/Oxygene)
    - 5 bits volume table and 16*16*16 combinations of all volume are from 
Sc68 by Benjamin Gerard
    - 4 bits to 5 bits volume interpolation from 16*16*16 to 32*32*32 from 
YM blep synthesis by Antti Lankila
-----------


> I have an idea for a genetic algorithm which can 'grow' very
> short blocks of sound using YM, based on spectral error feedback vs
> natural sounds (error from signal power and SNR).
> 
> It would be interesting to see how closely the YM could match natural
> sounds using genetic selection as the driver.
> 
> I have a few other ideas to try with it but this one interests me the
> most because I think it has not been tested, at least not on YM.


	- Eero