Digitaly Modding/Matching Microphones

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kingkorg

Well-known member
Joined
Apr 15, 2017
Messages
2,972
Location
Norway
The moment i heard about Voxengo CurveEQ i thiught this is something i have to try. I am blown away by the results. And this can be done in much better conditions and with a flat mic/room, cleaner preamps, better converters, speakers...
 
Last edited:
nice, izotope and fab filter also have EQ matching plugs... unfortunately theres a lot more in the sound of a mic than just the EQ curve.

Jess
 
As i mentioned in the video, and agree. This reveals all of the other differences when you get the eq part covered. What is the first other aspect in C1 you notice when they get matched eventualy in the last, or any other example in the video?
 
I don't believe in microphone modeling through software.
IMHO they only try to create a frequency response close the the desired microphone.
The developers of those plugins want to make us believe that you can make a Shure SM57 sound like a Telefunken ELA M251!
And we all know that this is not possible...
The quality of a microphone is much more than the frequency response!
 
The biggest difference i hear is in pattern response and distortion in hf peaks C1 has after matching. Even though transient response can be visible, i was never able to hear difference when two LDC are compared. SDC or dynamic to LDC is something else.

I think this is great tool if one wants to further match allready matched pair of mics. And in applying LF and HF attenuation in already well made U87 replicas without built in filters.
 
Nice one!
I'm rather not a fan of this kind of stuff, but i see some pluses.
For example - let say we record and mix some band album, and this band want to add other songs from a different sessions, different studio.
Always there's a problem to get some similarities in the music material due to different mikes (and microphone placement philosophy), different rooms, different drums, different backline etc. Knowing microphones used in previous session, like sm57 on a snare, sm57 on a toms, sm57 on a guitar - f**k, i hate sm57!!!!!!!! :D  We can easily get some similarities on instrumental tracks adding eq curves from new material.  For typical recording scenario i don't see any use of it, but for some trouble situations it can be usefull.

 
RuudNL said:
I don't believe in microphone modeling through software.
IMHO they only try to create a frequency response close the the desired microphone.
The developers of those plugins want to make us believe that you can make a Shure SM57 sound like a Telefunken ELA M251!
And we all know that this is not possible...
The quality of a microphone is much more than the frequency response!
Until now, I would have fully agreed with you. To my knowledge, there was no commercial solution that took care of all the other aspects, such as directivity and proximity effects, but now there is one:
https://townsendlabs.com/products/sphere-l22/
I don't know how well they are succeeding, though...
I would think it may encourage other companies to pursue this endeavour.
 
I haven't heard the most recent developments.
But some time ago, I was given a plugin where you could select your source microphone, and the microphone you would like it to sound like...
IMHO it was only a manipulation of the frequncy response.
(So you could actually as I mentioned, change a SM57 into an ELA M251!  ;D )
 
RuudNL said:
I haven't heard the most recent developments.
But some time ago, I was given a plugin where you could select your source microphone, and the microphone you would like it to sound like...
IMHO it was only a manipulation of the frequncy response.
Yes, that's what it's been until now, a toy.
Roland had that 15 years ago, in their VS880 "portastudio", supposed to make a small electret mic sound like Abbey Road's mic locker...
 
Theres a song on radio I recently produced, half vocal was recorded on u47 in Nashville and other half on Sony c800 in my room.

I used slates VSM to model the c800 into a u47 and it worked perfectly, smoothed off the top end and gave midrange forward thing enough to where no one could tell them apart in the mix.

another thing I forgot to mention is RMS. Loudness is a big factor in mic differences too.


Jesse
 
ln76d said:
For example - let say we record and mix some band album, and this band want to add other songs from a different sessions, different studio.
Always there's a problem to get some similarities in the music material due to different mikes (and microphone placement philosophy), different rooms, different drums, different backline etc. Knowing microphones used in previous session, like sm57 on a snare, sm57 on a toms, sm57 on a guitar - f**k, i hate sm57!!!!!!!! :D  We can easily get some similarities on instrumental tracks adding eq curves from new material.  For typical recording scenario i don't see any use of it, but for some trouble situations it can be usefull.
This is a standard feature in Samplitude since 2007! Works for guitars, keys, whatever, too.
For example, once I punched in a part that had been recorded through a Vox Tonelab with a lot of FX. I punched in through a UREI LA610, run FFT Analyser, that creates a comparative convolution file, and voila!
 
JessJackson said:
Theres a song on radio I recently produced, half vocal was recorded on u47 in Nashville and other half on Sony c800 in my room.

I used slates VSM to model the c800 into a u47 and it worked perfectly, smoothed off the top end and gave midrange forward thing enough to where no one could tell them apart in the mix.

another thing I forgot to mention is RMS. Loudness is a big factor in mic differences too.


Jesse

Exactly. I believe Adele Skyfall was recorded with two different mics as she was touring or something. I believe it was Røde and Neumann. This is exactly why i think my method here can be useful. However please know that i have over 20 condenser mics, and at least 10 are modded, or made from scratch.  So i would never start a project with Studio Projects and try to model Elam. I would rather use SM57 and let it be 57. I want to believe in capacitors and resistors, it's fun!

You can never make dynamic sound like condenser, as i stated in the video. This was not the point.

However if you gave me u87 and SP C1 i would be confident to use them as pair of overheads or room mics matched this way.

Keep in mind Shure KSM is edge terminated, and C1 is k67. No matter what i do i can still hear difference in natural resonances and  pattern response. 

I will just try as hard as i can to find u87  and match it to C1 or Behringer B2. The problem is that in Norway renting u87 is about 200 dollars/day, and i never wanted to own one as long i can have 10 used kickass Røde mics for that kind of money, so i dont think i'll ever buy one.

Wow, i just saw Sphere has free demo VST. Well i guess our task here is to recreate the mic :D

https://townsendlabs.com/sphere-plugin-demo/
 
kingkorg said:
Wow, i just saw Sphere has free demo VST.
I read their technology white-paper.
https://townsendlabs.com/sphere-whitepaper/
Very articulate and very honest, no marketing BS, it's refreshing.

Well i guess our task here is to recreate the mic :D
Well, from their own admission, the Sphere mic is not as good at simulating dynamic and ribbon mics than LDC's. That could be an opportunity...
 
I am 99% certain it's 797 audio k67. I analyzed correction curves of the plugin before it models anything else. There's setting where it just lets through clean mic signal, and then setting where it corrects mic to flat response. Measure that curve and flip eq curve and you get frequency response of the mic.  Looks exactly like C1 and B2 pro response i have measured. When i apply these to some recordings i have made with these and it makes perfect sense.
 
I have to admit that i was wrong. After measuring impulse responses of the plugin, and finding average value of correction curve plugin is applying to original mic before emulating, i came up with this curve. Which is exactly the same curve of my Microphone Parts RK47 capsule in one of my mics.

Now that i have real u87 to experiment with, i applied the plugin to my RK47 mic, and it matched sound spectacularly. Applying plugin to 797 capsule gave too bright results with c12 c800 and ELAM. As i didn't have real u87 i couldn't  be sure 100%. I will post audio later. Here is the average correction curve that matches my rk47 capsule. 
 

Attachments

  • Shape of townsend capsule.jpg
    Shape of townsend capsule.jpg
    348.8 KB · Views: 20
kingkorg said:
I have to admit that i was wrong. After measuring impulse responses of the plugin, and finding average value of correction curve plugin is applying to original mic before emulating, i came up with this curve. Which is exactly the same curve of my Microphone Parts RK47 capsule in one of my mics.

Now that i have real u87 to experiment with, i applied the plugin to my RK47 mic, and it matched sound spectacularly. Applying plugin to 797 capsule gave too bright results with c12 c800 and ELAM. As i didn't have real u87 i couldn't  be sure 100%. I will post audio later. Here is the average correction curve that matches my rk47 capsule.
No one had doubts about the capability of convolution to perfectly mimick a given frequency response at one on-axis location. Several softwares have done that since about 10 years now. The difficult (if not impossible) task is to mimick the directivity pattern of any mic. Only a microphone array can do that.
 
Back
Top