501

(10 replies, posted in Sinclair)

One thing I should say about using .ay files for collection purposes: it don't really suits for beeper music. I believe a new better format should be developed, at very least the old standard should be updated, before making a new music collection. Including copies (complete or partial) of the ROM into each file is really a major format design flaw. It also wouldn't hurt to add support for expanded memory and all extra sound devices ports (Covox, SD, etc), to store custom digital music and things like that.

502

(10 replies, posted in Sinclair)

Does not seem much different, the same number of actions:

- Load .bbsong into Beepola
- Export WAV
- Encode into MP3
- Upload MP3

- Load .bbsong into Beepola
- Export .bin
- Convert .bin with a tool into .ay
- Upload AY

Also, with simplistic approach these AY files wouldn't be suitable for publishing, as each one would have to contain copy of ZX48 ROM, so they would only be useful as intermediate container for MP3 conversion.

Another update. phasesynth's Detune range changed from +/- whole tone to 0..octave (with fractional numbers), this allows not just to detune it slightly, but to certain number of semitones, giving more flexibility. noisedrums added, these are simple noise-based drums with some PWM.

504

(7 replies, posted in Sinclair)

I thought why not to actually put the idea I explained above to the test, so here is proof-of-concept. I wouldn't say it sounds better, but it gives somewhat different timbres.

505

(7 replies, posted in Sinclair)

Interesting, so it is in fact one of the particular cases of what Phaser engine could do (duty control when phase reset is enabled).

More tricks the better, as they tend to sound different.

506

(7 replies, posted in Sinclair)

Haven't look into the code, but an idea for volume control I had and never tried to implement is an evolution of what has been used in Tritone: mix channels by interleaving them, but change amount of time each channel takes. Time the code in a way that it does few outputs of each channel per one loop (or maybe few copies of the loop), and force some of these outputs to zero to attenuate the channel. I.e. we have channels A and B:

ABABAB full volume
ABAB0B mid A volume
AB0B0B low A volume

and so on.

Judging by the carrier tone that is heard in the test50.sna, is it something similar to the idea?

The bundle is updated. Minor fixes, two synths optimized, two new synths added - one Earthshaker-like (I love that sound), another with customizeable 1-bit waveform. Download link is the same.

Here is a demo track written by theandroidvision. It only uses few plugins of the bundle, and does not have any extra instruments or processing: https://soundcloud.com/theandroidvision … o-vst-demo

It will be 32-bit floating point music right off the bat, even with a single plugin, because that's how VST standard works. Yes, this is certainly not a purist thing, it is more of the 'fakebit' kind.

Fresh stuff just in time for the new forum opening.

This is a bundle of VST plugins aimed for different kinds of 1-bit sound synthesis and processing. It includes six plugins, two of them for processing, two drum synths, and two tone generators, covering majority of the 1-bit sound types. Comes with source code.

http://shiru.untergrund.net/files/1bitstudio.zip

511

(7 replies, posted in Other Platforms)

To my knowledge, no, there is no access time limitations for sound registers. It is a common 'feature' (or rather an issue) of FM synth chips, but I can't remember any PSG type chip that would have this limitation.

512

(7 replies, posted in Other Platforms)

GB having Z80 is actually a common misconception. In fact it is Sharp's custom CPU that is a 8080 derivative (like Z80) and has a small fraction of Z80 extra features, but that's it. Porting Z80 code to GB CPU is about as complex as porting Z80 code to 8080 - difficult, ineffective, but has been done before.

The worst thing about GB CPU regarding 1-bit music is that is has just as many registers as 8080, i.e. it completely missing Z80's alternative register set and index registers - this will have impact on possible number of channels. Worse than that, it also missing 8080's EX DE,HL opcode that is very useful in beeper engines, and there is a major HW bug with all 16-bit increment/decrement opcodes that leads to sprite display corruption.

Still, it is certainly possible to do something, especially for CGB that has about twice more CPU horsepower, but in general the platform is not very suitable for this kind of things, so it won't perform as well as Z80-powered systems.