Mic Emulations?

[iomegaman]

So I have a couple of UAD cards, honestly many of the plug-ins are really close to actual hardware that I have, so I guess on one leve a plug-in that emulates a device "can" be pretty darn close...but I notice a few companys (Antelope/etc.) now making MIC emulations...and I gotta ask...is this even doable?

I think its one thing to process audio that you've already digitized in the box with a plug-in that is designed to emulate a circuit, but microphones seem like a different beast.

I have some decent amp sims (S-gear is my go-to)...and all DAWs seem to have some form of EQ/Comps that are modelled...I got not problem with that...but a microphone is a "capture" device...it seems to me you cannot emulate something thats not there in the first place...the other emulation software can treat the Digital signal with filters etc that can mimic to some degree the hardware (you'll never convince me about headroom and transformers, those things are either there or they're not)...

How can you possibly emulate a $5000.00 microphone that has specific capture characteristics with a cheaper less "capturing" device? (The Waves Kings mic  stuff comes to mind)

Is this snake oil with better technology or is it actually possible that some of these mic/software combos can get close like the UAD stuff?

Looking at the antelope stuff it appears you also need their interface...so for <$2k you can have >$14k of mics?

 

[PRR]

It is certainly possible to process a track from a $5,000 mike to sound something like a $50 mike.

Or a $5 mike: "telephone filter" was standard gear in movie soundtrack work.

 

[scott2000]

I think Antares had something back in the  early 2000s that was supposed to be something like this...

 

[john12ax7]

It is certainly possible to process a track from a $5,000 mike to sound something like a $50 mike.

I would say that is the doable part.  Claims of the reverse direction is more snake oil territory.

 

[pucho812]

Antares had a piece of software and a hardware unit at one point that did eq curves under the guise of mic modeling.  We all knew back then it was next to impossible to take mic A and make it sound like Mic B but it did allow for an extra eq in the plug in box.  So we used it and it was ok.

A few years ago at aes or Namm I went down this road with a well known mic company doing  mic emulations.  there are a couple on the scene.  both have their own mic that you can connect up and  then boom run their software and their mic now sounds like " insert model here". one company  has a mic with a cardioid pattern, their claim is with their mic and software it can sound like any famous mic under the horizon. They even claimed some ribbon mics as their famous mics you can get.  I questioned how they got the different polar patterns from a cardioid charged capsule, and how  does the mic knows where it is in the room? they had no answer for me.  Now I am far from any mic guru but I can't see how a single capsule cardioid condenser is ever going to sound like a ribbon mic doing figure 8 or anything else other then cardioid pickup.  I told them at best they have modeled distortion characteristics and eq curves.  they proved it when they asked me what I thought of the u47 emulation and when I told them they proceeded to be mad because I didn't know the numbers on a VF14 tube off the top my  head. they really started to scowl  when I agreed with them saying I don't have to, I am not trying to recreate that in digital land.  As a repair tech who sometimes works on mics, my end of the tube is does it work and is it low noise, does it sound good when in use and if not why.

Then There was another company doing mic modeling. They wrote a fascinating white paper on their tech and were extremely open as to how they do things. They  have a mic as well but unlike the first guys their mic has duel capsules which have individual outputs. you then take those individual outputs and run it with their software.  while it sounds nice, the duel capsule allows for close emulations of things. Currently things are at the point where it will be a difference, as always YMMV and what works for you may not work for me. proceed with caution.

 

[gyraf]

Agree, not possible as advertized.

And agree, the Slate was a pure marketing horror show, trying to sell the dreaded Antares microphone modeller once again.

Antelope has some pretty good ideas thrown in, hardware quality way above the other, which helps processing somewhat by providing better raw material. And looks good.

But if you set it up and close your eyes, you still can't even start to guess what emulation program it's set to  :

To sum up, the technology may be useful for doing small repairs quickly on non-critical material. Not for much else.

Jakob E.

 

[pucho812]

Agree, not possible as advertized.

And agree, the Slate was a pure marketing horror show, trying to sell the dreaded Antares microphone modeller once again.

Antelope has some pretty good ideas thrown in, hardware quality way above the other, which helps processing somewhat by providing better raw material. And looks good.

But if you set it up and close your eyes, you still can't even start to guess what emulation program it's set to  :

To sum up, the technology may be useful for doing small repairs quickly on non-critical material. Not for much else.

Jakob E.

Well the townsend  labs mic stuff looks to be a better way all around. I can't argue how close or how far off they are with their emulations and to me if I used it I wouldn't care as long as it was sounding good for the track, which as we know is subjective.
Plus I love how open they are about their tech.. Even have a patent...
https://townsendlabs.com/sphere-whitepaper/

 

[iomegaman]

Well the townsend  labs mic stuff looks to be a better way all around. I can't argue how close or how far off they are with their emulations and to me if I used it I wouldn't care as long as it was sounding good for the track, which as we know is subjective.
Plus I love how open they are about their tech.. Even have a patent...
https://townsendlabs.com/sphere-whitepaper/

Did not know about these guys(townsend), or if I did I forgot...I will say this...UAD generally does not get in bed with bad design...the idea of multiple patterns coupled with Townsends tech married to UAD MIGHT get past some of the smoke and mirrors...not sure they can give the tin man a heart, but maybe its the journey anyway...

 

[JohnRoberts]

Back in the mid 80's the guy who hired me to work at AMR (Peavey) planned to make black box  emulations of famous (read expensive) recording mics... It took some work but we eventually talked him down off that ledge. It is a great product idea, just difficult to execute.

Now a few decades later the DSP technology is much improved so one might have a better shot to mimic certain physical mic characteristics (mass, diaphragm size, etc).  Pattern still seems a hill too far, but you might obliquely mimic proximity effect with bass boost, etc. .

JR

 

[pucho812]

Did not know about these guys(townsend), or if I did I forgot...I will say this...UAD generally does not get in bed with bad design...the idea of multiple patterns coupled with Townsends tech married to UAD MIGHT get past some of the smoke and mirrors...not sure they can give the tin man a heart, but maybe its the journey anyway...

Well their latest version of their software is UAD and a version without UAD.

 

[pucho812]

Back in the mid 80's the guy who hired me to work at AMR (Peavey) planned to make black box  emulations of famous (read expensive) recording mics... It took some work but we eventually talked him down off that ledge. It is a great product idea, just difficult to execute.

Now a few decades later the DSP technology is much improved so one might have a better shot to mimic certain physical mic characteristics (mass, diaphragm size, etc).  Pattern still seems a hill too far, but you might obliquely mimic proximity effect with bass boost, etc. .

JR

well like I said, if  we could boil a mic down to frequency response and distortion characteristics, then everyone would be doing it.  I agree the idea is good one, but mics are mics and so much is dependent on  outside  forces that I can't imagine being able to accurately do that with current DSP tech.  but who knows what the future holds?

 

[pucho812]

Emulations at a new low

https://reverb.com/item/11138194-nux-solid-studio-ir-power-amp-speaker-cab-mic-simulator-guitar-effects-pedal?gclid=EAIaIQobChMIq66q8Oe15AIVjiCtBh2MEAKSEAQYAiABEgIKIfD_BwE&gclsrc=aw.ds&merchant_id=105622467&pla=1&utm_campaign=2058510199&utm_medium=cpc&utm_source=google

 

[iomegaman]

well like I said, if  we could boil a mic down to frequency response and distortion characteristics, then everyone would be doing it.  I agree the idea is good one, but mics are mics and so much is dependent on  outside  forces that I can't imagine being able to accurately do that with current DSP tech.  but who knows what the future holds?

I don't know...Nassim Taleb says that of you want to get close to what the future holds go back 50 years and remember the world before now and try to imagine how today would have looked to us back then...

I know a lot of advances have been made in DSP technology but frankly there is a limit on what you can manipulate with technology...the Nyquist theorem seems to be about recreating a set of frequencies but add to that room conditions such as humidity, the medium you are recording to, the altitude , the actual voltage supply coming out of the wall and a myriad of other conditions that are more than just sound waves...it seems like an infinite list of conditions and even with quantum computing infinity by definition is infinite.

Suppose we get quantum computing break throughs in the next few years, there's still code needing to be written that can take advantage of it etc...

I think if I was betting on recreating the environment I would project that instead of one device doing many things we will probably have a plethora of devices doing fewer things, but added together...maybe a 20 mic interface that allows better tweaking because you have more source material to dsp...

I dunno...the UAD stuff is quite good and often thats good enough.

 

[Zebra_PD]

Got a Slate WMS in the studio and its actually a quite good mic without the emulation. It holds it's own with most 10k microphones.

So for emulating transformer based mics like a U87 (A/B with both a original Neumann and a D87) or the character of a U67 or M49 it does actually very well. But compared to for example our D49  that we built with only quality parts it's not really there. But that is probably also because tube emulation still is not 100% in any plugin I ever used so far.

I use it as a simple way of trying out what mic will work for a voice. Then I put up one of our handmade clones and keep the Slate mic as a second mic so I have the option  of using a different mic of choice later on in the mixing process. Usually taking the second best take and using it for dubbing duties. Saves a lot of time and hassle.

 

[abbey road d enfer]

is this even doable?

In the principle, it is, but emulating a mic's response is a 3D matter. It is necessary to re-create the original "soundfield" and apply the appropriate convolution.
Most of the available mic emulation systems recreate a 2D image of the soundfield, one that is valid for one single position.
It is possible to recreate a 2.5D image of the soundfield with two mics, that's what the Townsend system does.
Real 3D modelling takes a 3D microphone such as an Ambisonics.
I don't know of any commercial software capable of applying convolution to an Ambisonics signal.
Dr Angelo Farina has created X-volver, a software that allows doing that. One has to divide the space around the mic in a mesh topology and capture the Impulse Responses on a 3D balloon. This takes time, but it's probably not more difficult than sampling a piano.