Well, the 1st part is the personal preference. Due to the chip's given soundstyle "from scratch", people who liked more high sounds, preferred POKEY, while people who preferred the "fat base sound" , preferred SID.
The "bass" sounds , used in POKEY tunes, have been some "note similar sounding" polycounter distortion sounds, that sounded most stable when calling the registers directly. Or the 16 bit poly has been used, dropping a channel and get a good config for a big amount of POKEY tunes. The "best fitting pitch basses" always had the problem, that they sound different each "call" . So they were not much usable. The solution here is rather simple: If they sound different each call, then call them more often, to have an approximated sound, deeper than standard basses and more pitch correct.
But here is the crux. How to get coders and musicians, that prefer the "high notes", making tools and music, supporting the lower sounds? Development always gets haltet by personal preferences, which explains why RMT never supported this directly.
2 different pitches were played at the A8's resolution in two different octaves. Bt the pitches were not really a faktor, due to the 8 bit res. with the modulations, you get 2 different octaves at the 100% same pitch (of 2 different octaves). Higher notes need faster corrections, as done in the "double speed" tune.
This one uses 3 "modulated" channels at the same time. Due to the real low bass usage, the filtered channel plays at mid range, manipulating the background voice. The background voice automatically gets tunes to the bass by this.
To make something clear. The Videos were done, using the emulation. Pictures and Music were taken from Atari800Win+4. You know, the emulations sound different from the real machine. Particular the modulations are very timing accuracy depending. So they vary in the needed timing, and the sound result is about 95% correctness. Some parts the real POKEY plays better, some it plays worse. But the shaping of waveforms is clearly better on the real POKEY.
This one is recorded using Altirra. It seems the new version gets closer to the original sound, so I gave it a try. As you can see, it's still simple VBI programming.
The most important point: This one uses no frequency vibrato. It uses a waveform-vibrato.
@emkay, I've a question as a musician. You're showing impressive examples, but I'm missing some details. What you're exactly doing with RMT song - and/or, maybe, player. What do you mean talkin about "orbit" manipulation (orbit means shape?) and what's exactly "restart" trick.
I'm asking cause I'd like to know some details and also start experiments by myself :)
To the "Orbit" programming. As no Tracker exists for resetting POKEY's registers, it helps to set the channel to a different generator and to play the lowest pitch. Playing pitch 41 with gen. C would look like (for example)
CCCCCACCCC 44444F4444 11111F1111
Speed of repetition (length of the envelope) and position of the "a" command has to depend on the resulting sound for the tune. The sound will "warble" to a resulting "new wave" . reducing of volume could be needed.
This tune makes it a bit more clear, I guess. If you listen to the bass, it has a faster "warbling" sound (running at 2x VBI). The resulting sound stands almost stable and you can use a better fitting pitch for the basses than from the bass table. The short playing of the "gen A" makes the polycounter playing from a different start...
the other two are just "imaginated" values on the bass generator (means the value you want to play) and the gen A has to be played at lowest pitch. A supporting Tracker had to reset Pokey's registers at demand. And the envelope reset (playing gen a at lowest pitch)should be enabled in some cpu cycles, to make it unhearable. In RMT it's the only possibility to set it in the envelope.
You have to understand simple physics for that, and how POKEY works. 2 high pitch tones played at the same time will result in a loud high pitch tone. 2 high pitch tones played close together will result in a lower tone (harmonic)because you don't recognize the switch between the high pitches. for basses it means, to set the chance for POKEY of playing more than one high pitch at the same time with bass gerators, you have to spread the generator peaks from each other. This is granted when playing the lowest pitch on the used channel(s) For creating effects in higher ranges, you have to find a way to set the peaks as close as possible. This is done playing the highest pitch on the filter voices, until they get synchronous. In RMT it means playing 2 times the highest pitch on the envelope . Then you can adjust the offset of the filter by playing a lower pitch and set it back .... thats all.
The adjustments in RMT are rather coarse, because it's a long time between the adjustments. The shorter the time for adjustments, the sharper the effect will get. That's why double speed sounds even sharper than single speed tunes.
The music from 27th post is horrible and awful! it doesn't stay the tune. The lead voice is out of the key and background sounds are pitch shifted less than an 1/4 of note to the lead.
I atached an example of good practice of modulation and pitch-bending. Listen to, from about 44 second. Every notes are almost perfect in tune and the pitch is changing very gently and carefully.
You know, it took more than 20 years people thought POKEY is not able of doing (controlled) modulations. POKEY is now ~35 Years old, and no Tracker supports all features right now. So I'm happy with the results....
You know what happened last year. No chance for Raster to support his tracker from the place he belongs to now.
We'd need a new Tracker and someone who puts the relevant features into it. Actually, someone has started a project. But it may take a long time till a useful release may appear.
Bob Yannes himself explained it. If you have a look at the Noise and the Pulse generator.... He actually had a look at the "Sounddevices" from other companies (incl. Atari) and built the Sid upon that. The part with POKEY simply missed the coordinated programming of the pulse width, which has been handled right now. And it makes 2 operator modulations possible. Let's have a supporting Tracker/Sound development tool and and Oscilloscope , and make real POKEY music in the future :)
@emkay, do you have a quote on that from Bob Yannes? Because from what I can find his "base" was that devices he found in homecomputers of his time were... terrible and he said their creators obviously had no idea about music. Hence he wanted to release a proper synthesizer, or at least what Commodore has let him to release, because what he was intending was not SID we know - it was much more than that.
And surely POKEY has better noise generator.
As far I'm aware - but feel free to correct me - on POKEY you need 2 voices to generate somewhat proper PWM, and while I agree it's good to have a player which would support it properly, this will effectively render the chip dual-voice. This still makes sense with two POKEYs, but does not make sense to have them stereo in this way.
@booker: Trackers as RMT already support creating instruments involving 2 channels in one (of course withing one pokey). No one said that it's always needed to have it on all 4. See demo track attached to RMT called like_c64 - actually it just contains few notes played with such instrument.
Dla mnie z muzą z Sida jest jak z azjatami. Wszyscy wyglądają dla mnie tak samo :D Oczywiście jestem fanem Pokeya ale też bardzo chętnie czasem posłucham Sida. A mój ulubiony kawałek to Gordian Tomb. Szczególnie ten - kocham go :)
@wieczór: yes I think I know what you mean wit RMT.
What I was saying to emkay was that if we want to change way of doing music with POKEY (to be more synthesizer alike, and this has obvious advantages soundwise) then this has to come with the price of extra voice per instrument. And while two POKEYs like it sounds good in mono (or as I prefer 25-30% separation) but puts the stereo output in question due to the obvious separation, which is currently being avoided using echoed instruments, but their sound is nearly always the same as they use standard single voice hardware waveforms.
@Eagle, dla mnie np. muzyka z SIDa Roba Hubbard'a dajmy z 'Commando' a Linusa dajmy 'Progressive.data' brzmi bardzo różnie :) I mimo, że co często powtarzam, nie kryje się ze swoją głuchotą, to chyba coś je różni :D
:) niektóre instrumenty brzmią tak organicznie że w AY czy POKEY trzeba by użyć sample do naśladowania : ->link<-
a jeśli ktoś nie wierzy że to jest z C64 to proszę sobie choć na Sidplay2.5 ustawić wersję sid 6581R4_1985S i posłuchać: ->link<- lub lepiej odtworzyć na prawdziwym C64 wyposażonym w SID 6581AR4 (będzie lepsza jakość)
@booker: No pewnie że się różnią. Ja jestem muzycznie upośledzony a dostrzegam różnicę. Niestety jak słucham muzykę na C64 to mam wrażenie jakby to był remix dopiero co słuchanej innej muzyki. Tak jakby wszystko było na jedno kopyto. Muzykiem nie jestem więc trudno mi określić dlaczego ulegam takiemu wrażeniu. A rodzynków na C64 jest sporo które bardzo chętnie słucham i które mają nawet dla mnie jakąś wartość sentymentalną. No i do tego moja żona bardziej akceptuje puszczanie muzyki 8bit w domu z C64 niż z JIL :( (ona jest z obozu C64) Już mój taki pech że musiałem z wrogim mi "ideologicznie" osobnikiem się ożenić :):):)
A ten kawałek jest zdecydowanie inny niż wszystkie:
