Printed Amstrad Addict magazine announced, check it out here!

Main Menu

Instrument pitching in STarKos and ST128

Started by Nworc, 18:18, 18 January 22

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.


Good to see that STarKos is somewhat in the tradition to ST128 by means of look'n'feel, in the way it is controlled by the user, but also in the way it works. The latter might be a compromise to stay compatible in a way.

The point that I never understood in Soundtrakker, and which I see being implemented in the same way also in STarKos is the handling of the pitching for an instrument. I am talking about the pitch function of the instrument editor.

Sometimes you want to add nice tricks to an instrument, e.g. to change the pitch over time - like you can do with the loudness or noise effect. And to my thinking, the most logical way to implement that in a musical fashion would be to provide a modification in terms of division of the current sound note. Like in e.g. 1/8 of a note distance to the next tone, from C to D for instance, so to stay within this example, a pitch of +8 would make a D out of a C.

But what is implemented is apparently independent of the current tone, it is just a fixed number added to the current frequency, meaning that the pitching distance you achieve depends on the note, or octave you play. This is ridiculous - I mean, that way the pitching setting can only be used for a single tone as it sounds awkward in different tones.

What do you think about that / what is your experience.

BTW: STarKos is a great tool, and migration is made easy for users that are accustomed to the classic music composer.


On Starkos you can use the arp column (+2 in your exemple C -> D) for an appoggiatura.
Tough, you cannot encode the glissando effect a more precise resolution would allow.

The reason behind STarKos and St128's choice is technical (same battle for vibrato):
- They first work with semi-tones (so arpeggio and transposition are easy), but don't use higher resolution (e.g. 1/16th of tone), which would require working with 16 bits rather than 8.
- Then the note is converted to PSG period (proportionally invert of the frequency, that's why the higher the note, the more dramatic the effect). Then it would be too much computation to apply adjusted pitches.
One could argue that's also fine from a musical point of view (more ample vibrato on higher notes mitigates the timbre's aggressiveness --cf Tim Follin's works of art).

Tl;dr: The musician point of view is slightly sacrificed to have very efficient player routines.

Ayane will remove those restrictions, since nowadays we can convert any modules to even faster players based on stream replay.
(well, the internal in-tool player still must remain fast enough, especially for 300Hz replay, but that's covered!)


@Nworc If you have any desiderata (*), feel free to log them here:
Would be very happy to hear!

The project has been resurrected, and the first alpha is planned on June.

[size=78%] (*) [/size]


@m_dr_m: Thank you for so much your clarifiying words - you can't imagine the relieve, if after having wondered about that for years, you finally see that there are people out there who do think the same.

While computation time might have been one of the reasons (besides maybe tradition, as the prototype of these trackers, the Amiga Soundtracker, worked also that way but I haven't checked that), I think that today you want to get the maximum out of what is possible with the AY, even if it would cost you a couple of cycles, just to thrill the ears a bit more.

Being able to do a perfect glissando / portamento would be a great thing, among the possibility to use mini-samples! I think I fully understand what you intend to do, I do believe you are not just thinking about cheap digi-samples, but samples that make full use of the PSG, and that would be really great!

I had a good look on your feature list - I like the name of the project and I like what you are doing, and really hope to see your project hitting the stage in the near future. The best help that I can give you is just a simple advice: don't try do too much features, don't do too much at the same time. The simple is the strong. Just make sure that the features you decide to do, are 100% perfect. You can still add more features (on reserved commands) in the future.


Quote from: Nworc on 20:53, 23 January 22don't try do too much features

Yes I plan to do exactly that!

Just able to read STK files and pimp sounds to add "sid voice" effect at first.

Regarding portamento:
It could be encoded in the sound with fractions of tone as you suggest.
I'd also like to let the user encode it in a pattern (when sporadic) by putting the attacking note + option (e.g. D4 Pt) and then the target note (e.g. C4).

So the duration of portamento would be exactly the one between the two notes, and the steps would be calculated automatically.

Powered by SMFPacks Menu Editor Mod