cclarry
...you can't model air molecules moving through the air and
hitting the spluttered diaphragm of a Microphone.

 
Engineers do this every day with Computational Fluid Dynamics. You could add adjustments for humidity and temperature. Mic placement, humidity, temperature, diaphragm materials could all be modeled.  Just need $$ and computational power. The latency of a an amp sim using CFD would be excessive today. Tomorrow, who knows. 
 

When I first heard about modelling - I think it was a Line 6 device - I too thought "NO WAY, it will never happen". That was the tone snob in me, I think. :)
 
But the developer in me (I've written over a dozen books on software dev, so no layman here either) totally sees it. I agree with you 100% - we are not there today and we may not be there soon. But think how you used to need a UAD card because the PC couldn't keep up with native plugins. Now, a good machine can run a huge number of powerful native plugins, and with each iteration they get better and better and better.
 
The technology will get there - even the randomness of chaos can be modeled with enough time, ingenuity, and horsepower. We may not have even invented the tools that will do it yet, but it will happen. I actually think perfect modelling of analog gear is no that far off. My guess is five to ten years at the rate we're going. And then that changes things for hardware guys. Who would invest the time and energy to build an expensive hardware unit that could be cloned with a few months work and resold for a few hundreds bucks? We could see a day where there are no more outboard hardware builders, or just a very few boutique guys. Again, not tomorrow, but 20 years?
 
When someone truly nails analog circuit modelling, why not have them license the technology so anyone can build a schematic visually, choose the virtual brand and material for each component, and instantly have a working software model? Sort of like how many game developers use the same underlying game engine they license from a 3rd party. Now picture your future cell phone being able to run it all and just "future bluetooth" to a set of real stomp switches on the floor or a touch panel monitor.
 
This is all WAY more easy to believe than going back 200 years and envisioning taking moving color pictures and transmitting them with sound across the entire globe in real-time, or even just capturing video on your cell phone. :)
 

Ok...ear test....2 files...one WITH the 432c on the "Fresh Air" setting, and one without...tell me which is which...

https://soundcloud.com/cclarry/vital-instruments-1
 
https://soundcloud.com/cclarry/vital-instruments-2
 
 

Bill is basically describing what Nebula tries to do -- capture the sound of a piece of hardware by sampling the output given various inputs.  As has been mentioned, this can work very well.  At least, it can get closer than most circuit modeling techniques have achieved, from what I've heard -- but this applies mainly to steady-state characteristics as heard from EQs, preamps, tape machines, and the like.  Nebula excels at these, but I've yet to hear a convincing dynamic saturation sound or compression action from Nebula.  Maybe soon that'll change ...
 
Cheers,
Eddie
 

Cookie Jarvis
I've been following the discussion here and thought I would throw out my own .02
 
First of all, the whole secret to "capturing" the characteristics of a hardware unit for software creation should have nothing to do with individual components. The important thing is output, not the individual elements used to create the output. Sound goes in(Input A) and then sound comes out(Output A). If you do a precision analysis of the difference between Input A and Output A you have the "characteristic" of the hardware unit. Now the hard part is to engineer the software that will result in the same difference. You also have to handle the various permutations of button and switch positions to completely model the hardware unit. This process has a lot in common with capturing Impulse Responses for reverb units.
 
I don't know about you but I'd rather have a software copy that creates the same sound manipulations than one that has modeled transistors and diodes
 
Bill

 
Yes and no in fact , since it's a dynamic thing , it varies with time and level most of time , this non lineraty is what we miss and love (our ears ) and it's the biggest thing to emulate , and to do so , we have to knows witch componenent impact the sound and how ...
 
what you're descibing here is the use of convolution mainly ....witch is great by the way ..but not a deep recreation of analog ..
 
Let me just had another cent to my opinion : few weeks ago i helped a friend to do a "for theatre" music ...and i was all Rock and seltic music ...all recorded live ..
 
As you might know (or not) i'm doing mainly electro based music (hip hop , electro ect) and when i do acoustic(like abbey road kinda stuff) , it's plugin based ....
 
On this project , i didn't had to used one emulation plugins FX for the mojo it add ...not one (or maybe a tape for one or 2 bus) , bus that imperfection , that analog feel was there on takes (via Neve consoles , preamps , pedals ect ...)
 
So the end of the story , is that since we can have great gear for low prices nowdays (500 formats is one options) , if you guyz doing live based music , all those debate are WAY less important that All ITB guyz like me ....** since that mojo can be there from start ....
 
 

I'm listening on crappy computer speakers at work, but my son and I like 2nd. The difference is very very subtle to us.

cclarry
So I guess no one was up for the challenge...
 
Ok...ear test....2 files...one WITH the 432c on the "Fresh Air" setting, and one without...tell me which is which...

https://soundcloud.com/cclarry/vital-instruments-1

https://soundcloud.com/cclarry/vital-instruments-2

2 is the one here crappy in ear headphones , the problem is to switc from one to another ....anyway to much nothing for me