pouët.net

Amiga - native Paula sound

category: general [glöplog]
Quote:
But listen to the funky bass notes! Cool, eh?


No.
added on the 2013-10-05 12:34:29 by grogon grogon
Yeah, I deserve that facepalm, I guess I was reaching and starting to make ridiculous claims in defence of Pokey (a general purpose sound chip) against both SID and Paula (later and superior, and designed for music specifically).
added on the 2013-10-05 16:05:38 by Foebane72 Foebane72
It's also sad that Pokey (and Paula, I believe) have to share IC space with the joystick / paddle / keyboard interface circuitry - in fact, that's what Pokey stands for: POtentiometer / KEYboard :(
added on the 2013-10-05 19:15:01 by Foebane72 Foebane72
AMIIIIIIIIIIIIIIIGAAAAAAAAAAAAAA!!!!!
added on the 2013-10-05 19:39:47 by djh0ffman djh0ffman
AAAAAA MIIIIIII GAAAAAAAAA !!!1

\o/
added on the 2013-10-05 21:01:36 by leGend leGend
Quote:
Mind you, the chip wasn't designed with mod music in mind (which didn't really exist at this point).


But 99% of the time it manages to play it so well.

I have experienced an exception: Nebulos by Audiomonster, played through a version of MED by Teijo Kinnunen (I don't think it was OctaMED). The tune mostly played fine, but there were the odd blips and squeaks, as if the Paula was being pushed to and almost past its limits by the complex MOD file. Or it could just have been MED not playing it very well. BTW, I never heard the tune in "Ice" by the Silents where it was used, until years later, but when I did it sounded fine.
added on the 2013-10-06 10:16:37 by Foebane72 Foebane72
Quote:
But 99% of the time it manages to play it so well.

Err, obviously, because people developed the format and play routines to match the Amiga architecture - but most certainly not the other way around. If a tune sounds different in MED than in ProTracker, it's most likely not pushing the Amiga to its limits, but rather pattern effects are interpreted in a different way, e.g. pitch slides might have a different range or so.
Quote:
Err, obviously, because people developed the format and play routines to match the Amiga architecture - but most certainly not the other way around. If a tune sounds different in MED than in ProTracker, it's most likely not pushing the Amiga to its limits, but rather pattern effects are interpreted in a different way, e.g. pitch slides might have a different range or so.


Yes, I realised that after I made the post, that MED would interpret things ever so slightly differently, like different trackers are likely to.

In fact my previous post was badly written. Sorry.
added on the 2013-10-06 19:51:03 by Foebane72 Foebane72
One thing never capitalised by ANY tracker on the amiga is real-time loop modification. There could have (or still could be) some serious fuckery and almost grain based shenanigans to be had with that!
added on the 2013-10-07 00:36:49 by djh0ffman djh0ffman
Completely agree. The SoundTracker/ProTracker system was "unfortunately" already so amazingly good and expressive compared to what there used to be that a whole generation of musicians had more than enough to satisfy their creative needs.

I can think of at least two things that would probably be completely doable on a plain vanilla OCS Amiga: loop mangling and Waldorf style wavetable synthesis.
added on the 2013-10-07 19:15:19 by yzi yzi
Loop modification as in funk repeat/invert loop effects?
I think the funk repeat etc. commands weren't supported and standardized enough to be widely used. And people didn't have an idea wtf they could be used for, because certain musical styles hadn't been invented yet. ;) And for full-blown modern sample mangling things, you'd really need more and better features than just the EFx stuff.
added on the 2013-10-07 19:45:57 by yzi yzi
There were many interesting things that "could be" but unfortunately... Amiga trackers freezed in time at some point with not much more than fixing bugs.


Quote:
The SoundTracker/ProTracker system was "unfortunately" already so amazingly good and expressive compared to what there used to be that a whole generation of musicians had more than enough to satisfy their creative needs.


Yes and no. Sometimes it was trying to go around the limitations with some unforseen ways. Like loading twice as long samples to Protracker by compressing it with PowerPacker...
added on the 2013-10-07 19:59:22 by grogon grogon
Using just a little larger, or a little more samples doesn't really allow any new forms of expression. Some small changes were made to trackers like the "chord builder" thing in ProTracker later on, but even that didn't really change the game at all. Anyone with a synthesizer was able to sample their own chords anyway.

Sample mangling (splitting samples to very small areas, changing loop start/end points on the fly, reversing sections, etc.) and wavetable synthesis could have given actual new forms of expression. Another thing that might have been possible is precalc-based dynamics effects like sidechain compression emulation. This could have been done by calculating volume curves for each sample, and using those values to e.g. attenuate other channels.
added on the 2013-10-07 20:32:19 by yzi yzi
Quote:
Using just a little larger...


That was just an example.
added on the 2013-10-07 21:01:54 by grogon grogon
The point is, there were expression-enabling technical possibilities well within the capabilities of the OCS chipset, but these were never explored, because no-one happened to think of them. And there was little motivation to think of anything radically different, because the ProTracker system was already good enough. There was an abundance of quality music and player routines using that system and everyone got into the paradigm. Your example just proves the point. ProTracker had set the way of thinking in stone, and everyone just wanted more of the same thing. More samples, longer samples, more bits, more channels. The expansion from only a few channels to dozens of channels was a substantial change, similarly as moving to mp3 soundtracks in demos and soft synths in small intros.
added on the 2013-10-07 22:11:50 by yzi yzi
Quote:
One thing never capitalised by ANY tracker on the amiga is real-time loop modification. There could have (or still could be) some serious fuckery and almost grain based shenanigans to be had with that!


erm, it was. a lot of replays devised in the mid to late 80ies did that ! (mostly to emulate sid sound, e.g. SIDMon) ! + Listen to what e.g. Felix Schmidt did back then (e.g. the famous Paranoimia intro tune and some other ones). I myself wrote a tracker (Syntracker) in 1991 that used similar realtime filter code.

Quote:
Sample mangling (splitting samples to very small areas, changing loop start/end points on the fly, reversing sections, etc.) and wavetable synthesis could have given actual new forms of expression


hey, my tracker (AON) did that in 1993 and it was not the first replay to do so, the Maniacs of Noise replay (e.g. for one of the tunes in the game Unreal) did also use wavetables.
added on the 2013-10-07 22:57:12 by xyz xyz
My example proves only that the 'thinking part' always existed. But you are right - the music wasn't the main priority in demomaking process, so there was no pressure on making music tools better.

Correct if I'm wrong, but Renoise got this feature when... this year?

http://www.renoise.com/blog/Slice%20Markers%20Explained
added on the 2013-10-07 23:03:31 by grogon grogon
I'm not sure what kind of features you're talking about, but can you point to any impressive example tunes made with them? Beat slicing and loop mangling on the Amiga in the 1980s or something? Music that was way ahead of its time? I suppose you can't.

By the way, it was only after the Swedish company Propellerheads came up with their ReCycle software that one particular type of loop-slicing thing really took off. Even though some of that would have been possible already with, say, the Amiga in 1985. Nobody just thought of it before. Or at least nobody who also could have made impressive music with it to prove the concept.
added on the 2013-10-07 23:08:40 by yzi yzi
By the way, I seriously disagree about music not being the main priority. Every true classic demo has an awesome music. (if it doesn't, then I don't like it) It's just that the music that was being produced was already good enough from the coder and producer's point of view. There weren't that many bold visionary thinkers around. I mean, even if I had accidentally coded a loop-mangler system, I wouldn't have had the courage to make music with it, because I would have considered the sound just too weird.
added on the 2013-10-07 23:21:03 by yzi yzi
AON demo songs (not recorded by me) shows some wavetable based tunes done by Pink^Abyss in AON. you can find some more complex tunes in some of the Atari Falcon demos, e.g. Sonolumineszenz by Avena , tracked on Amiga in AON)

Unreal tune using MON replay

Paranoimia tune by Felix Schmidt 1989

@grogon: yep, the "thinking part" definitely existed. It is also a logical progression to the SID sound, at least the SID chip developer focused on wavetable synths in the early 90ies (google ensoniq ASR)

@yzi: are you kidding me ? musical styles ? back then there was no "glitch hop", "dubstep" and what kids listen to nowadays, hence the technology was used differently.
added on the 2013-10-07 23:30:31 by xyz xyz
I'm kind of saying, why didn't you invent glitch hop then. Just think of what could have been done with the Amiga already in the 80s... But these expression possibilities weren't explored. The world wasn't ready for it.
added on the 2013-10-07 23:37:52 by yzi yzi
Quote:
added on the 2013-10-07 23:08:40 by yzi
By the way, I seriously disagree about music not being the main priority. Every true classic demo has an awesome music. (if it doesn't, then I don't like it) It's just that the music that was being produced was already good enough from the coder and producer's point of view.


If I say - the process of creating music wasn't important for a guy who was setting up some vector cubes - will it make more clear?

The whole family of trackers was (and is) lacking in so many places that there was no idea where to start to fix it. For example sample offset (9xx) lacking in it's accuracy with longer samples, clicks it produced occasionally (without a way of avoid it) etc... it was a simplest way of beat mangling tool (which no one intended it at the beginning). It was bad and people did live with that.

Another example: 32 samples limit. Look at Travoltas "Condom Corruption" mod - sample #6, #7, #10 (just an example, there are many more and better ones) and think like a coder "ah, yes - I see you need more instruments in your tune - we will definitely provide you better music system next time!".
added on the 2013-10-07 23:46:14 by grogon grogon
Quote:
@grogon: yep, the "thinking part" definitely existed. It is also a logical progression to the SID sound, at least the SID chip developer focused on wavetable synths in the early 90ies (google ensoniq ASR)


True. But I mostly speak about Protracker (one to rule them all, you know ;>). It could evolve, but it didn't.
added on the 2013-10-07 23:55:38 by grogon grogon
1991, some nice sound modulations - has it's own flavour

Paul van der Valk in Turbo Imploder 4
added on the 2013-10-08 00:03:36 by grogon grogon

login