Obtaining Difference In Frequency Response Between Two Mics

M

mario_c

New member
I've done a search and cannot seem to find a post or reply which answers my question. I am wondering how I can go about using a program such as Room EQ Wizard to find the difference in EQ between two microphones and export the eq data file to be imported into a plugin within a DAW such as Cubase 8. Is it possible to calibrate REW with one mic (such as an AKG c414b-uls) instead of calibrating with the measurement mic (flat frequency response mic, such as the ecm8000) and then run a measurement of a second mic (like a c414 xls) to determine what the difference in frequency response is? Can I export the eq data to import into a plugin in Cubase?

I have used q-capture and q-clone to obtain the frequency response of my preamps to see how they are impacting my sound. I am able to use those frequency responses to Ed a track now if I want which is nice (of course it does not allow me measure the harmonic distortion of the preamp or to accurately measure the difference of a mic).

I use PC so I cannot use Fuzzmeasure.

To summarize:

1) How can I obtain EQ data that reflects the difference between my microphones which can be used as an EQ setting?

2) How do I export it? Which plugin can I import it with?

3) Is there any other data which may be useful (that can be found in REW or the likes) when comparing two mics such as octave distortions that I can also import into a plugin?
 
Not that this answers your question directly, but Antares used to make a plugin that did exactly what you are trying to do. You would tell it which mic you recorded something with and which mic you wish you recorded it with and it would do the conversion.

It wasn't terribly popular and I think it was based on impulse responses. (Which would include distortion and such)

I think it was call mic modeller, or something like that.
 
Thanks a bunch! That sounds great. I'll have to take a look. Isotope has a quick EQ matching option in their EQ plugin. Obtaining the harmonic distortion sounds great. REW allows you to get this info but I am unsure how to use another mic as the reference mic to obtain the difference from. Also not sure if exporting the impulse response and opening it in waves IR1 would be correct. Thanks again!
 
You could open the impulse response in waves ir1, but you would be applying the new impulse over the frequency response of the original mic.

The Antares plugin would strip the audio of the original mics response before adding the response of the mic you choose.
 
Impulse responses can't really reproduce non-linear aspects like distortion. The reason why is kind of obvious but not easy to explain, so I'll let you google for it.

So you've got what? A pair of wav files where you ran noise through your pres and recorded the output?

I think that if you deconvolve one with the other, you should end up with an IR file that you can load into whatever convolution plug you want. Pretty sure ReaVerb (free with Reaper) can do both the convolution and the deconvolution. My only real question is how to prepare the files first. I tend to think that you'd want both files exactly the same length, starting at exactly the same time.

I think you also need to have started with the same noise, though. When you recorded the original files, was it driven by a file containing a sample of noise or a random noise generator? You'll need to do it with the file. IDK how much difference it will actually make, but it can't be none. :)

ReaFIR in subtraction mode would be able to do something like this somehow also, but I think deconvolution is the right way to go.
 
mario_c said:
Obtaining Difference In Frequency Response Between Two Mics
Click to expand...


Why don't you just look at each mic's frequency response chart, which is usually pretty easy to get off the manufacturer's website...? Then you can then see what their differences are...and also use that to set your EQ if you think that's going to really help you more than simply using your ears.

Here's one for the AKG c414b uls


What's the purpose of taking that info and "importing into a plugin"...?
Do you think that by using the frequency response difference from a flat response mic and another mic as some sort of "plug in", you can then apply that to tracks to what...?...make them sound better...?...flatter in their responses...???

Just curious about your reasons and goals. :)
 
Oh shit, I read preamps and ran with it! I guess this is all about mics?

I started to say that's more difficult, but I think now maybe not, at least not if you've got a decent speaker to generate the source sound. I'd go with a sine sweep. You'd think that the speaker and the room would color the whole thing, but the nature of the deconvolve process should remove all of that. As long as the mics are as close as possible to exactly the same spot, and nothing else changes, it should get you as close as you're going to get. The speaker's frequency limits will probably impact the results, but bumps and troughs in the response shouldn't.
 
Seeing how close I can match the two for stereo recording from a distance. I know it isn't the best way but I like to experiment and try things out.
 
How exactly would one deconvolve?

I was obtaining impulse data in REW using a log sweep. I put the microphones as close to the same position as possible so, as you have said, remove the variables of the room. In REW you can use mic calibration files, meant to obtain room acoustic impulse responses using a measurement mic like an ecm8000. It makes that reference mic the "0" so it knows any changes in EQ can be credited to the room.

Now, trying to get the difference between the two mics, or even two mic pres is done by measuring one and then the other. Subsequently, use the arithmetic function of A/B to get the difference since they are logarithmic and use db. The only problem with this is the EQ curve you get or the harmonic distortion data isn't in a form that I can put into a plugin, not even text data that represents that EQ. So how would one export it? The only way I could think of was going to the EQ tab and generating "filters" which are meant to be the inverse of the EQ so as to noise cancel it. But these are not perfect and of course are the inverse so I would still need to manually put in the EQ data and put a negative gain attribute as opposed to a positive.
 
mario_c said:
Seeing how close I can match the two for stereo recording from a distance. I know it isn't the best way but I like to experiment and try things out.
Click to expand...

So...you want to use two different mics....but then try and "make them the same" by applying the EQ frequency response difference that you've obtained and imported into some sort of DAW plugin, and then applied that to the track that used that "different" mic, in order to make them more even for stereo use....?

OK...seems like way more involved than it needs to be...but have fun. :)

I think you could probably just use your ears to adjust the two mic tracks post-recording to get them balanced...
...and with distant mics, it really might not even matter all that much, unlike a tight stereo pair....IMO.
IOW...you're going to get differences anyway with distant mics, even if they are the same two mics....unless you were in a perfectly balanced room with mics positioned symmetrically/identically.
Even a slight difference in position of two mics in a perfectly balanced room, could yield source frequency response differences with distant mics.
 
That is how I understood stereo pairs regarding distant miking. I appreciate EQ afterwards to get as close to a similar centered image is easier (what I was doing a couple of days ago), but I am very interested in getting more involved and understand what exactly the mic is doing to the sound and seeing if I can use this information to make the other mic sound even closer than just EQing by ear. I enjoy doing this, which may sound odd haha.

I will even look to using it for mic pres. I have a few Neve 1073's which I can notice a difference from the ISA 110's on my isa428, but only when I AB them and not after post production stages. So I wanted to understand the differences. I looked in REW using a log sweep directly in and noticed that not only was there a difference between the Focusrite and the Neve in EQ and harmonic distortion but also between the individual Focusrite pres, especially in the lower frequencies. I want to figure out how to get this data and use the arithmetic function of A/B against another mic pre (or if I'm comparing mics and then compare one mic to another) to get the difference (in EQ and harmonics) and then have it in an exportable form to use in a plugin on a channel.
 
This won't work. Two different mics with virtually identical frequency responses still don't sound the same, so when used for stereo in a technique requiring relative volume and often relative volume and a timing difference too - the stereo field will be less precise. The differences in microphones is not just level at specific frequencies. It has a component that relates to how the diaphragm reacts to impulse sound - so large and small diaphragms react differently. If you can find any of those ultra slow mo videos of loudspeaker cones moving as a tone sweep takes place, you can see how different speakers react very weirdly and unexpectedly. This applies to mics too. Measurement mics rarely sound good for recordings, but look good on paper. I have a pile of Line 6 handheld radio mics that have mic modelling - so you can select from a number of classic mic 'sounds'. They're actually quite realistic but there's a problem. With the real mics, the polar patterns also change dependent on frequency, and this can't happen - all the simulations have identical polar patterns. This will apply to your idea to tweak a sound to be identical. In practice I doubt it will be that obvious unless you are recording something with very precise placement - like an orchestra. The effect is subtle, but manifests itself as a movement in position as instruments play through their range. Worse on the wider range instruments - harp, piano and cellos where they can play quite low, or very high in one pice of music. Sounds to me like the players suddenly slide their chair across a foot or two. Actually it's also very similar to when you have one of those very 'active' soloists - like the clarinets who thrash around from the waist - and when you listen in the audience with your eyes closed it makes you slightly nauseous?
 
Haha great explanation. I understand different diaphragms will react differently but as you pointed out, I am not trying to track an orchestra but rather just using them as drum room mics. Further, I am simply trying to learn about the frequency and harmonic distortion of mics and mic pres to apply these characteristics (more accurately- the difference between the two) to another.

So does anyone have any experience with Room EQ Wizard (REW)? How do you use it to compare mics? If you use REW for room correction, explain what you're doing to export the impulse response or EQ. This may help me figure out a way to apply it to comparing mics.
 
mario_c said:
I am simply trying to learn about the frequency and harmonic distortion of mics and mic pres to apply these characteristics (more accurately- the difference between the two) to another.
Click to expand...

The point rob was making is that you *can't* simply apply the EQ differences to another.
There's more to it than just EQ'ing one mic to sound/perform like another.

mario_c said:
So does anyone have any experience with Room EQ Wizard (REW)? How do you use it to compare mics?
Click to expand...

Well...REW is used for measuring room acoustics...not to compare mics. Mic frequency responses are usually measured in an anechoic chamber, and you can already get that information, which then you can use to compare the mic differences.
Of course, I get it...you want to somehow take the mic readings from REW, and turn that into data that lets you subtract the similarities of the twoi mics and leave you with data containing only their differences.

Why don't you ask the REW people directly, or on their user forum...maybe they can figure it out for you, or give you some ideas how to get what you want...?
 
mario_c said:
How exactly would one deconvolve?
Click to expand...
Run it through a deconvolver. ;)

I own the paid software that Voxengo had in part of a bundle, but like I said I'm pretty sure ReaVerb can do it, and I'm pretty sure there are free options. You have .wavs that you recorded? Make sure they started from the same source, start from the same time, and are the same length. Stick them in the program and tell it to do its thing. It'll spit out an IR that represents the difference between the two mics and you can jam that in any "IR loader" that you prefer.

Personally I don't care why. "Because I'm curious and have the time and gear to do it" is a good enough excuse for me. :)
 
I figured out how to use the voxengo deconvolver for EQ but does it also deconvolve the impulse responses of the room (reverb and delay)?
 
If you take one of the mic recordings and deconvolve it against the original sine sweep, you'll get an IR that represents the EQ of the speaker, the reverb of the room, and the EQ of the mic. It really pretty much is the difference between the two signals that encoded.

OTOH - if you deconvolve one of the mics with the other mic, the speaker and room are (hopefully) NOT different between the two, so should not be included in the the IR. They "cancel out", leaving only the difference between the two mics.
 
You must log in or register to reply here.
Back
Top