"alias free analog" modular environment

[Architeuthis]

i like saying "black hole theory" and "free energy". its not about being literally right, its about having fun while theorycrafting. im a kid underneath and im not about to give that up.

so, translated: when i kill all the aliasing, as in it no longer shows up on the analyzer at all, i like to say ive sent it into a black hole. hint: nobody actually believes theres a black hole in my audio chain. and the "free energy" is just: the same noise that dithers the oscillator and cleans up the sinusoidal distortion also works as an analog style noise source. one circuit, two uses, the second one is free. i havent broken thermodynamics, theyre just concepts that are fun and help me think.

bottom line: i use astronomy as inspiration.

[stoopicus]

i couldnt finish past projects because writing the code was the bottleneck, not: understanding the problem. now the bottleneck is gone and im shipping.

One thing you learn very early in a software development career is that coding itself is never the real bottleneck.

its a bit funny to be told ill "never know why things dont work out"

Please don't take generalized statements or truisms personally. Additionally, people pointing out potential pitfalls to you are doing you a huge favor, not insulting you.

[Architeuthis]

understood.

[DrGonzo]

Finally. Now we just have to wait for a proper plugin that emulates aliasing on a component level.

[JasonSpatola]

There are no sane people in this thread. Not one. Anyone who wasted their time postI---- ah, f**k.

[antto]

i cannot finish anything without this modular environment. i need an environment where i can fine tune things. the whole ending aliasing thing is a side quest to this.

absolutely understand you, back when i was starting in 2006 (or actually earlier), it was only because i found an environment for visual programming and/or "modular" for realtime DSP - you probably have heard it, it's called SynthEdit, if it wasn't for this - i would've never started.
but also, it was exactly in that early time when i had this new tool in my hand and i was finding out what i can do with it (but yet, i had zero knowlage of how things work, zero knowlage of DSP, zero knowlage of coding) when i ended up in a similar euphoria just like you - Huge ideas (my particular idea was to make a Monster Synth, better than anything existing, with All the features ever, i am embarassed to say this, but there are probably some corners of the internet which remember this period of mine, i do not bleach my old posts)
so then i quickly ended up needing custom modules for SynthEdit and started asking others to make them - no one did, when i got more impatient about this - i was slapped with reality - that's where i understood that i will have to begin learning how to write the code - and those same people who were telling me to "go learn" and who i was seeing as acting "bad" to me, they began helping me make my first steps into writing my first custom module, helping me set up the IDE and compiler, etc.. i was wrong about them, they weren't "bad", i just didn't see the picture clearly back then

jamcat, i couldnt finish my past stuff because i needed help coding. now that i have ai, it fills in the knowledge gaps so i can finally move forward.

this is the "trap" i mentioned
you created your Monster Synth, and i am 99% sure you haven't seen the code (it's mountains of code, and when you want to have some change in it - the chatbot rewrites that mountain of code quicker than you can eat a biscuit)

antto: if you call this broken, i dont know what to tell you. it is clearly a fully functional analog emulated filter that can scream, growl, self-resonant, ring, and generally filter

so, you see, you chalanged me to take you seriously and you threw some video at me, i watched it, i commented, and you are unable to answer
let me repeat my question: why does turning the Resonance up cause the resonant frequency to suddenly go down? is this a "feature" you explicitly wanted? is this modeling some filter which does it? or is it a bug (and if it is, how come you didn't notice it)?

i like saying "black hole theory" and "free energy". its not about being literally right, its about having fun while theorycrafting. im a kid underneath and im not about to give that up.

this
is IMO what leads you to form something like a sort of language barrier and an echo-chamber with the chatbot (the same thing can happen also when two real people talk)
the chatbot is also "trained" on literature (books, movie scripts, comedy, absurdism, fantasy, sci-fi), people use it to help them write stories

when you mix a highly technical topic (programming) with another one (DSP) things become difficult but the recent chatbots can sorta get surprisingly passable results. but then you also mix another technical topic (astronomy, which features many unknowns, including black holes), and another one which i'm not even sure if it's bullsh*tology (free energy, there's a lot of crap written with this phrase), add your brainstorms and the "language barrier" that i mentioned above, and now the chatbot's job is magnitudes harder, you say words like "black hole" or "free energy" or who knows what else, together into some DSP/coding discussion, for you those words may be metaphores and might not matter - but it matters a whole lot to the chatbot and it may cause it to "try to make some connections" (and when that doesn't make sense to it it might drift into fantasy and sci-fi or who knows where else).

this is why i suspect, especially if the discussion has a long "memory" (i know the paid versions offer this - much longer "sessions" (note: my only experience with chatbots is from duckduckgo's free chatbot feature, the sessions are very short)) .. mixing DSP/coding with brainstorming, child-like dreaming and crazy ideas, and words or phrases or metaphores from completely different areas that have nothing to do with DSP/coding, and mixing things which may having something to do with DSP/coding but are too difficult to mix, and the conversation will drift into fantasy IMO.
and you also develop your own language, between you and your chatbot. one which no one understands really, you say some sentance thinking one thing, the chatbot reads it and "thinks" something else, replies to you (probably some crap) and you read it and think something completely else and as a bonus - it doesn't make sense to you but reminds you of something different so you get inspired and you ask a question in a different direction (you change the subject), and this is a recipe to steer the conversation into the world where nothing matters and nothing makes sense

so, translated: when i kill all the aliasing, as in it no longer shows up on the analyzer at all, i like to say ive sent it into a black hole. hint: nobody actually believes theres a black hole in my audio chain. and the "free energy" is just: the same noise that dithers the oscillator and cleans up the sinusoidal distortion also works as an analog style noise source. one circuit, two uses, the second one is free. i havent broken thermodynamics, theyre just concepts that are fun and help me think.

bottom line: i use astronomy as inspiration.

right
i understand, but when you do this sort of thing:
1) you confuse the real humans here, at least me.. i'm reading about "ending aliasing" and i see "black holes" - how dafuq am i supposed to make sense out of this and how am i supposed to not think you're drunk or something?
2) you confuse your chatbot and make its job harder

i guess, it's the same problem for both humans and chatbots, except:
- when things start to not make sense to the human - he will stop following or tell you you're just loco
- the chatbot _must_ follow you and generate a response, no matter what

[Architeuthis]

antto, you said: "so, you see, you challenged me to take you seriously and you threw some video at me, i watched it, i commented, and you are unable to answer"

my response: you have to understand how confused i am as to why i even should explain it. this is basic analog filter behavior. i didn't realize until now that, if you've never had hands on experience with an analog synth, yeah, you might think dipping in frequency, by increasing resonance, is broken. it's not, it's the opposite: it's a sign it's working with flying colors. anyway, i am not and will never be a mathematician nor technically minded, so here's an ai explanation:

CLAUDE CODE: <removed because it doesn't matter, go ask ai yourself>

TLDR VERSION: with nonlinearity you get pleasing distortion, with linearity you get ugly distortion

LASTLY: i need my analog modular environment, not synthedit, not reaktor, not max/msp, not puredata, not vcv rack, mine (my precious ), done how i want, exactly to my specification.

EDIT: clap plugin almost done
EDIT2: i understand ai doesn't code the best. i have plenty of time to go through the code and reduce and make it efficient. i plan on making the code easy to understand and edit. right now, it's very far away from that, and i am aware. the goal is that it's very easy to make your own modules.
EDIT3: i am working on adding a codeblock module.
EDIT4: i am taking feature requests

[9b0]

i am leaving this here as a note.

this topic was originally called “i know how to end aliasing”. the OP claimed to have a method that solves aliasing forever, but no concrete technical explanation was provided, only vague terminology and buzzwords. he also claimed that this technology would become the basis of humanity’s future, affecting every imaginable scientific field and industry.

his explanation changed repeatedly. first it sounded like the method was based on something similar to phase modulation self-feedback. later he claimed he was building an additive synth. later still, he posted code examples and claimed that his polyBLEP oscillators were almost alias-free, while the shown implementation appeared to be a simple 1-point polyBLEP derived/yoinked from Tale’s code, which is very far from being “almost alias-free” in a strict technical sense. he also claimed that aliasing can only be defeated through noise, which is demonstrably not true, and what he meant by this was never clarified, while it clearly turned out, he did not know, aliasing free waveform generation is not something humanity needs to solve.

the topic was not his analogue modeling environment. as a dev, i am not interested in it (just like most of the devs), and since i have competing products on the market, i especially do not want to be associated with bashing another developer’s product, or harming his business. my posts were directed at the OP’s debatable technical claims while the topic was called 'i know how to solve aliasing'.

however, since the OP changed the subject and deleted or edited earlier posts, the discussion has become unreadable and misleading. that is why i am leaving this summary here.

in my opinion, this behaviour is unacceptable on a technical forum. repeatedly making extraordinary technical claims, refusing to provide evidence, changing the subject, and then editing the history of the discussion makes constructive conversation impossible.

[whyterabbyt]

however, since the OP changed the subject and deleted or edited earlier posts, the discussion has become unreadable and misleading.

That's been their consistent MO for 20 years.

[Architeuthis]

i am leaving this here as a note.

this topic was originally called “i know how to end aliasing”. the OP claimed to have a method that solves aliasing forever, but no concrete technical explanation was provided, only vague terminology and buzzwords. he also claimed that this technology would become the basis of humanity’s future, affecting every imaginable scientific field and industry.

his explanation changed repeatedly. first it sounded like the method was based on something similar to phase modulation self-feedback. later he claimed he was building an additive synth. later still, he posted code examples and claimed that his polyBLEP oscillators were almost alias-free, while the shown implementation appeared to be a simple 1-point polyBLEP derived/yoinked from Tale’s code, which is very far from being “almost alias-free” in a strict technical sense. he also claimed that aliasing can only be defeated through noise, which is demonstrably not true, and what he meant by this was never clarified, while it clearly turned out, he did not know, aliasing free waveform generation is not something humanity needs to solve.

the topic was not his analogue modeling environment. as a dev, i am not interested in it (just like most of the devs), and since i have competing products on the market, i especially do not want to be associated with bashing another developer’s product, or harming his business. my posts were directed at the OP’s debatable technical claims while the topic was called 'i know how to solve aliasing'.

however, since the OP changed the subject and deleted or edited earlier posts, the discussion has become unreadable and misleading. that is why i am leaving this summary here.

in my opinion, this behaviour is unacceptable on a technical forum. repeatedly making extraordinary technical claims, refusing to provide evidence, changing the subject, and then editing the history of the discussion makes constructive conversation impossible.

yeah, agreed. bandlimited alias free saw, square and triangle have been solved for decades. blit, blep, minblep, polyblep, wavetable, additive, i know all of it. thats not what i was claiming and its not the point. the alias free oscillators on the site are building blocks i use, not a discovery im announcing. that was scaffolding, and i can see it read as the headline. fair, thats on me for framing it that way.

the actual claim: you can reach any bandlimited target spectrum with recursive deterministic generators, building the sound in band by construction instead of generating out of band and correcting after. chaotic maps are just the nonlinear end of that same recursion. alias free isnt the interesting part, its a free consequence of generating directly in band, same reason additive is clean. the interesting part is control, steering a recursive generator to a specific timbre by formula with no stored tables.

so if you want to argue the real thing, argue that. everything you listed is prior art i already use, not something i think i invented. its on soundemote.io if you want to hear it. believe it or dont.

[Architeuthis]

honestly i didnt walk in here thinking this was a peer review panel. i came to theorycraft and have fun with sound. bold claims, wild framing, black holes, free energy, chaos, thats play for me. thats how i think and how i enjoy this. i say stuff like "alias free black hole" because its fun, not because im submitting it to a journal.

i get now this room doesnt run on that, and thats fine. no hard feelings, i just misread the room. i thought it was a place to riff about sound and it turns out its a place to defend every word like a dissertation. thats a mismatch on me, not a crime on anybody.

so ill take the playful stuff where playful is welcome and keep building either way. works on soundemote.io. anyone who actually wants to mess around with sound and ideas, im easy to find.