Login / Register  0 items | $0.00 New
soundmodel
KVRian
 
515 posts since 28 May, 2010, from Finland

Postby soundmodel; Sat Sep 02, 2017 4:11 am Analog-modeling or analog-emulating?

So I've been wondering about the terminology used to market plug-ins and also describe certain kinds of DSP.

Is analog-modeling or analog-emulating more correct?
User avatar
BertKoor
KVRAF
 
10067 posts since 8 Mar, 2005, from Utrecht, Holland

Postby BertKoor; Sat Sep 02, 2017 4:17 am Re: Analog-modeling or analog-emulating?

It depends: is it conceived by mere modelling or by emulation?
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is back online!!
mystran
KVRAF
 
4597 posts since 11 Feb, 2006, from Helsinki, Finland

Postby mystran; Sat Sep 02, 2017 5:17 am Re: Analog-modeling or analog-emulating?

soundmodel wrote:Is analog-modeling or analog-emulating more correct?


"Analog modelling" traditionally means roughly "we looked at some analog gear and tried to make something that sounds similar" and doesn't really say anything about what was actually done, or in what sense it's supposed to be similar. [edit: this is not to say "analog modelling" is necessarily "bad" or "low quality" it's just a very vague term]

"Analog emulation" doesn't mean anything, because emulation is a purely software concept (making one CPU run a program that interprets code written for another different CPU).

The word you are probably looking for is "analog simulation" where you basically "model" stuff on a lower level (eg. individual components) then build a circuit out of them and run a numerical simulation on the result. In practice, so called "macro modelling" where you take some larger circuit block and build a simplified model for it is standard practice, so the actual complexity/quality of "simulation" can vary quite a bit... but like that's the basic idea.
Image <- plugins | forum
quikquak
KVRist
 
298 posts since 6 Aug, 2005, from England

Postby quikquak; Sat Sep 02, 2017 7:16 am Re: Analog-modeling or analog-emulating?

'Emulation' is not purely a software concept...
"to copy something achieved by someone else and try to do it as well as they have"
"to try to equal or excel; imitate with effort to equal or surpass"

The TV weather guy has a multi-million dollar cloud simulation running for him, but he still f**ks up my weekend's weather forecast.

To be honest they're all crappy marketing terms, that all say absolutely nothing. - Go Team GO!!!!!
soundmodel
KVRian
 
515 posts since 28 May, 2010, from Finland

Postby soundmodel; Sat Sep 02, 2017 7:49 am Re: Analog-modeling or analog-emulating?

quikquak wrote:'Emulation' is not purely a software concept...
"to copy something achieved by someone else and try to do it as well as they have"
"to try to equal or excel; imitate with effort to equal or surpass"


Yeah I too think that it should be analog-emulating, since that's what they try to do. They try to get the DSP to correspond to the analog.

Simulating to me is analogous to synthetic or synthesizing. It means that one generates or creates from scratch. It can also overlap with modeling in the sense that simulations can be used for modeling phenomena. But simulations themselves are "artificial" creations that replicate some aspect or the entirety of some phenomenon, mechanism etc.
User avatar
Rockatansky
KVRist
 
91 posts since 3 Jun, 2017, from the Thunderdome

Postby Rockatansky; Sat Sep 02, 2017 8:38 am Re: Analog-modeling or analog-emulating?

A simulation tries to mimic a process or a certain way of operation.
An emulation tries to mimic a certain behaviour or result.
There's a reason the thing is called a flight simulator and not a flight emulator.

"Analog modeling", as I understand it, refers to "mathematical models". In other words words: observe a real-world process and turn it into a mathematical formula. It does not automatically imply or require probing chip voltages.

Although, obviously, that would very much aid the process. But when I play with some of those so-called "analog modelled emulations", I highly doubt the creators even own a voltmeter.

It's all just marketing B/S to make you believe and (quite literally) buy into the illusion of purchasing expensive hardware to use on your computer for peanuts.
Confucamus.
User avatar
Zaphod (giancarlo)
KVRAF
 
2323 posts since 23 Jun, 2006

Postby Zaphod (giancarlo); Sun Sep 03, 2017 9:08 am Re: Analog-modeling or analog-emulating?

We started using the terminology "analog emulation" 12 years ago and really it doesn't mean anything, it's just a description of our process.
We used this terminology because our one was a close but not perfect simulation, with the basic idea hardware cannot be "simulated" completely. So I preferred the "emulation" word, as deformation of Italian language, where "emulazione" means you are trying to simulate something but not perfectly. Today we are very close sometimes, but who cares, the terminology is a sort of "signature". Other developers are not using it to my knowledge

Analog modeling/component modeling is something different, the term is commonly used by companies spicing the hardware gear. I preferred a different description because our process was a black approach, we don't "model" anything
User avatar
Zaphod (giancarlo)
KVRAF
 
2323 posts since 23 Jun, 2006

Postby Zaphod (giancarlo); Mon Sep 04, 2017 7:49 pm Re: Analog-modeling or analog-emulating?

I see other companies are "copying" our crappy definition looool
Thank you for the hands up, it suggested our next product launch. I'm curious to see if other developers will follow us in our next terminology lol

Like I said before, this is crazy. We used "emulation" because the "simulation" was not perfect. That made my day
User avatar
FabienTDR
KVRian
 
921 posts since 23 Feb, 2012

Postby FabienTDR; Mon Sep 04, 2017 8:53 pm Re: Analog-modeling or analog-emulating?

soundmodel wrote:Is analog-modeling or analog-emulating more correct?


It's both BS! Most of these "let's press 50 year concepts into digital" concepts are in fact great examples of bad digital processing.

Ironically, it's right there were you'll find most nyquist warping, quantization and aliasing distortion on the plugin market (Giancarlo's impressive black box approach aside, which has excellent antialiasing by design).

I understand that some laymen have romantic ideas of the various marketing terms used in ads, but all use multiplication, addition and branches. There is no difference. It's all simply math.

These terms are made to fool, and they do it well as it seems. ;)
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!
User avatar
Zaphod (giancarlo)
KVRAF
 
2323 posts since 23 Jun, 2006

Postby Zaphod (giancarlo); Mon Sep 04, 2017 10:05 pm Re: Analog-modeling or analog-emulating?

Eheh thank you fabien, I'm an huge fan of your compression
User avatar
Urs
u-he
 
21485 posts since 7 Aug, 2002, from Berlin

Postby Urs; Tue Sep 05, 2017 9:06 am Re: Analog-modeling or analog-emulating?

FabienTDR wrote:These terms are made to fool, and they do it well as it seems. ;)

Let's say a company wants to point out that they used Nodal Analysis for filter equations instead of RBJ's Audio EQ Cookbook. What terminology would you suggest which isn't on your "made to fool" list?
Oden
KVRist
 
214 posts since 30 Oct, 2010

Postby Oden; Tue Sep 05, 2017 10:45 am Re: Analog-modeling or analog-emulating?

I think a lot of the responses are misunderstanding the question. It's not about what's accurate, it's what sells.

That means you use whichever term happens to be "hot" at the time. :D
quikquak
KVRist
 
298 posts since 6 Aug, 2005, from England

Postby quikquak; Tue Sep 05, 2017 12:40 pm Re: Analog-modeling or analog-emulating?

Tell 'em it uses unidirectional copper wiring in the plug-in! - See how far you get. :P
Don't forget the Magic rocks - people will believe anything: Read this - http://www.machinadynamica.com/machina31.htm
User avatar
FabienTDR
KVRian
 
921 posts since 23 Feb, 2012

Postby FabienTDR; Tue Sep 05, 2017 2:00 pm Re: Analog-modeling or analog-emulating?

Urs wrote:
FabienTDR wrote:These terms are made to fool, and they do it well as it seems. ;)

Let's say a company wants to point out that they used Nodal Analysis for filter equations instead of RBJ's Audio EQ Cookbook. What terminology would you suggest which isn't on your "made to fool" list?


"Nodal Analysis based design" maybe? :)

Most of this stuff is digital business and has little to do with the challenges of analogue circuit design (beside circuit analysis of course if your approach asks for it, but the implementation is digital business). At the end of the day, we have digital wizardry. Not more, not less.

There are many ways to approach the problems of sound synthesis and processing, and it's great when devs explain what they are doing.

But what does all this look like "in the stomach" after the compiler optimized the code? (i.e. on the CPU)

It's hard to even define the term "Analogue" properly. So trying to differ between Analogue modelling or Analogue emulation is somewhat pointless imho. I didn't mean to criticize marketing talk in general either. Just the silly technically irrelevant blabla. Your communication clearly has "Hand und Fuß" ;)
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!
User avatar
Urs
u-he
 
21485 posts since 7 Aug, 2002, from Berlin

Postby Urs; Tue Sep 05, 2017 11:26 pm Re: Analog-modeling or analog-emulating?

I always liked the term "Component Modelling", as coined by Korg. I avoided using it because I think they slapped a TM over it.

Unfortunately "Analogue Emulation" is the term that has been established since the early days of VST. While it's a bad misnomer, it has simply gained great clarity for the people out there: An "analogue emulation plug-in" strives to look, sound and feel like a certain hardware device.

We could go on using that term, but then we also want to draw attention to the idea that things have advanced since the early days of VST. So we seek terms that resonate with the established terminology, yet is different enough to make clear that it isn't "yet another analogue emulation". So here we go with terms like Analogue Modelling, Circuit Simulation, Topology Preserving Transform, Zero Delay Feedback and so on.

Out of all these, "Analogue Modelling" rolls down easiest to me, and it also points at the difference between early days and what we have now. Because, unlike what "Component Modelling" suggests, we do not necessarily have code that describe individual components, but we base our equations on the properties of the components in an analogue circuit, and their interaction. What we have is indeed a "model", and we use that model to run a simulation in realtime.

The question is linguistic then, can we take the process of "creating a mathematical model of an analogue circuit and implementing it as a software algorithm" and call it "Analogue Modelling"? Seems like wrong way round, but who cares as long as we can express what we mean by it? We can't possibly do like Night's Watch and call it Analog's Model.
Next

Moderator: Moderators (Main)

Return to DSP and Plug-in Development