[Milkymist-devel] [Democompo] Status

[Bengt Sjölén]

Hi Sebastien,

yes certainly it's not ready yet, I just wanted to share that the  
sequencer was running and playing stuff:)

About the rythm in 0 and 3, this is all about non-implemented commands/ 
effects.

0 was just that "set tempo" was not implemented and some delays in the  
song were in the samples themselves - so playing at wrong tempo the  
timing breaks - also the fine tuning commands and settings are not  
implemented yet so the actual sample playback speed might be different  
than intended so things might be out of tune and offset in time  
because of this too.
I implemented "set tempo" now at least so this one I think is playing  
ok now - haven't listened to all of it though.

3 is a really advanced module using lots of "set sample offset" to  
play from within a sample and multiple "set tempo" and "pattern break"  
and "position jump", none of which was implemented either before -  
there it even stopped after the intro because it should have jumped  
elsewhere in the song. These are now implemented too so this plays a  
lot better - might still be other issues with tuning, arpeggios and  
all the other stuff like that you can even put loops in a pattern...

About the API I havent looked at all at this yet, I just stuck my  
stuff inside of the testbed that Fabien set up but this is of course a  
trivial change - especially since all is already parsed and played  
directly from the memory mapped file.

B



On 13 sep 2009, at 19.12, Sébastien Bourdeauducq wrote:

> Hi!
>
> On Sunday 13 September 2009 17:21:18 Bengt Sjölén wrote:
>> I just pushed some major changes into the modsynth repo. It's playing
>> the mod files in the data directory fairly successfully now.
>
> Cool :) Thanks for the contribution.
> Just two things:
> - the API must expect a buffer containing the MOD data, not a  
> filename from
> which it reads the file. So the function should be something like
> int mod_load_buffer(mod_context_t**, void*);
> instead of
> int mod_load_file(mod_context_t**, const char*);
> - severe problems can be heard on 3.mod, and on 0.mod the rhythm is  
> not
> correct (this problem is perhaps linked to those with 3.mod)
>
> Regards,
> Sébastien
>