Analog modelling

[Richard_Synapse]

Ah the final frontier. DSP sounding more analog than analog.

That's not a joke to me. I believe that's possible and I hope to prove it.

That's not possible, because "analog sound" is a subjective term. Of course you could produce a digital sound, ask people to compare that to other analog sounds and your sound is perceived by people as "most analog". This doesn't prove anything beyond that single example though (furthermore it's very unlikely for all people to agree on something).

Richard

[Architeuthis]

That's not possible, because "analog sound" is a subjective term.

I have come to believe there are measurable differences between a good sound and a bad sound (edit: actually good/bad are subjective terms. I should say effective/ineffective). I don't discuss these things subjectively. I can prove with large measurable differences between something that sounds analog and something that doesn't. Of course, I only care about a specific small set of analog features which happen to be measurable and which happen to be obvious. I am not at all interested in opinion.

You can start to believe anything if you redefine what words mean and make them meaningless.

Who is redefining the word? It seems like the words in question are vaguely defined in the first place.

Let's say I create a synthesizer that uses very simple algorithms that I invented without knowing anything about analog synthesizers (but my goal was to have something sound "analog"), and it happens to sound very analog. The filters scream and growl, everything has feedback capability, I have anything-to-anything modulation including self-modulation, cross-modulation, basically it has the same sound and functionality as any good analog modular synth, maybe even drifting oscillators if the user turns that option on.

Analog Modelling?
Analog Emulation?
Analog Simulation?
Analog Behavioral Synthesis?
Analog-Inspired?

What word do I use? I would like to know.

[Richard_Synapse]

have come to believe there are measurable differences between a good sound and a bad sound (edit: actually good/bad are subjective terms. I should say effective/ineffective). I don't discuss these things subjectively. I can prove with large measurable differences between something that sounds analog and something that doesn't. Of course, I only care about a specific small set of analog features which happen to be measurable and which happen to be obvious. I am not at all interested in opinion

Sure, but this is about measurements, not sound. You can have substantial differences in measurements, does not necessarily prove or disprove anything about the sound. You can even play back the same sound twice, and some people may hear two different sounds - especially if you tell them the second sound is "analog" or something

Richard

[Architeuthis]

Actually, Richard_Synapse, I could respond to what you just said 10 different ways, but at the end of the day, after you really boil down all the arguments and opinions and beliefs, for me, it comes down to how many people agree with the notion that a particular digital synth sounds good enough to replace an analog synths that make similar sounds. If 60% of people surveyed (who have experience with analog synths) feel this way, then it's safe to say you've achieved something better than an analog synth.

I don't see any other end to this discussion. I can measure things all day, but it still goes back to opinion. I'm arguing that "majority opinion" is the final word. Maybe I'm contradicting my earlier statements now, but whatever!

[Architeuthis]

Also, this discussion could easily be seen as pointless and a waste of time. I think it is important only because it is my opinion that analog synths should be replaced by digital synths, and until we have a "landline vs cellphone" kind of replacement of analog synths, this discussion will continue, as it should. We are not there yet, hence the discussion. I believe these things are inevitable.

edit: Err... oh wait, analog synths are owned by few in the first place, ok landline/cellphone not a good metaphor... pretend "number of people who want analog sound and think digital cannot compete" = landline.

[mystran]

Stop the non-sense please. There is no such thing as "the analog sound" as there is analog hardware that sounds good and there's analog hardware that sounds like garbage. When you model something you usually do so in order to replicate the character of a specific good sounding circuit, not some generic "analog sound" which doesn't mean the slightest thing beyond slightly higher noise floor (usually more so in the circuits that sound like garbage).

[IncarnateX]

If 60% of people surveyed (who have experience with analog synths) feel this way, then it's safe to say you've achieved something better than an analog synth.

Really? As if any analog synth experienced by anyone per se sounds better than any digital? I got an analog synth, a Volca Bass. Sounds pretty shitty to me compared to my Fantom G, which is a rompler. I also got a Minilogue. I like that but have to make my own pathces because I find the factory presets harsh. I also got a so called "natural VA synth" as part of my JD-Xi, which also have an analog mono synth onboard. Now claiming that my analog synths sounds better than my digital to everyone, for every purpose, in every musical context, with any patch, is laughable, ridicoulus and as far from any musical realism as can be. You are just deluding yourself into a myth that was created at a time where analog synths were hard to get at and no one really could remember how they sound. It is a pile of BS. My minilogue and my JD-Xi VA resemble each other much more than minilogue and Volca Bass or the JD-Xi monosynth. If I made a blind test, you would fail miserably, and in thr context of a tune, no one would give a flying f.u.c.k. about how the music was made.
If you want to be rational about all this, the answer is simple: Consider every synth unique in itself, whether analog or digital. With this attitude you loose nothing if a synth is made to sound like another synth and actually fails this mission.

[stratum]

There is no such thing as "the analog sound"

On the other hand, there is something called "the digital sound" which means poorly designed algorithms running on 90's era cheap hardware (some of which was actually poor quality analog gear with digital displays only - but that does not alter the customer perception), so naturally there is a group of people who define "analog sound" by what it is not. Understandably this is not what is meant by modelling a specific circuit. But by the time modelling specific circuits became feasible, low quality gear was mostly past history, so the two senses of "analog sound" coincide in timing - I guess this is the reason behind the confusion.

[Architeuthis]

If you want to be rational about all this, the answer is simple: Consider every synth unique in itself, whether analog or digital. With this attitude you loose nothing if a synth is made to sound like another synth and actually fails this mission.

You do lose something.

the goal is to make it easier to compose music. The goal is to phase out analog synths like we should phase out humans from having a job (because we want AI to take over so we can have more time for creative work). Analog synths make music more difficult. They are expensive, first of all. You can't save your patches(in the case of modulars). I've owned, what I think is the best analog have to offer, and it inspired me to recreate those sounds for a digital synth, because an analog modular is too inconvenient. I've been selling off the pieces slowly as I'm already convinced that what I am developing is better, *because* it's digital, because it's convenient, and because it offers the same "effective" punchy snappy growly whatever you want sound... well I'm still in the process of making that happen. But my frustration with synthesizers today is that they don't pay enough attention to the details that matter. We need better envelopes, we need better filters, we need better pitch modulation, we need better everything. There's a lot of room for improvement and people don't see it, they think what we have is all there is.

Analog synths sound good because they are limited to the physical, natural world, unlike digital synths which can easily be made to do things physically impossible and therefore sound unnatural. But digital can take on those limitations as well. That's what I intend to... do...

[antto]

Architeuthis: so, did any of the analog synths you had have a sawtooth like on these pictures?
or do you know which analog synths do?

[Architeuthis]

No I don't know what analog SYNTH makes those waveforms, and that also assumes I've never owned one. HOW EVER, I did very much enjoy the Doepfer A-106-1 Xtreme FILTER, and I've already discovered how to recreate that sound, and the resonance looks very much like your top image. It will go sinusoidal to extreme spikey version of your top image. Easily done in digital with simple algorithms... although I'd say everything analog is easily done in digital.

Edit: A-106-1



Edit: Hmm, not exactly the same as your top image, yours looks more like a highpass saw inverted.

Edit: I'll take a wild guess. Bottom image doesn't look analog in my experience, but I am confident that not knowing doesn't really prove anything although my ego hopes I'm right. (It's not the kind of detail I'm talking about when I speak of important analog details)

Edit: I change my answer. I think the correct answer is both are analog. (both are physically possible)

[antto]

both images are from your own sawtooths which you "invented" in the other thread
there's some food for thought for you

[Architeuthis]

[antto]

it sounds to me like you have a kinda blurry romantic idea of "analog(ue)" sound, and you want to "nail" it digitally by random experiments
that's gonna take you a long time

in your position, i'd either drop the "analog" part and just experiment my way to create something briliant, no matter if an equivalent is possible with analog circuits... or select some actual analog synth and try to model/approximate it digitally in details

i've done both of these things.. you won't actually learn much about the analog synths while wearing the blurry pink glasses

[Architeuthis]

it sounds to me like you have a kinda blurry romantic idea of "analog(ue)" sound, and you want to "nail" it digitally by random experiments

I shall not be made a fool again!!! You're just encouraging me to be even more serious, "sharp", on target, with my work.

I may be blurry and I may be romantic with my concept of analog, but have already nailed what I am after. Hence my confidence. I am excited, not befuddled.

Ok a little befuddled.