Take the 2-minute tour ×
Sound Design Stack Exchange is a question and answer site for sound engineers, producers, editors, and enthusiasts. It's 100% free, no registration required.

In the picture you see the frequency response of two microphones. The black line is my source mic and the red line is the desired frequency.

enter image description here

I know there are so called mic modeller plugins but I only like to change the frequency via equalizer to match another frequency. So what would be the best workflow and how can I controll the results of the equalizer tweaking?

By the way: I work with Cubase 5. So if there is a (free) VST-Plugin it can help me, too. But I'm more interested in a general way of EQ tweaking.

share|improve this question

migrated from avp.stackexchange.com Jan 24 at 20:04

This question came from our site for engineers, producers, editors, and enthusiasts spanning the fields of video, and media creation.

3 Answers 3

up vote 2 down vote accepted

The first question is always: what's your application? Why do you need such close EQ curve matching, and is it possible there's a better way to accomplish your objective than using software to manipulate your audio? (For example: if you're trying to match an existing mic in a recording setup, would it make more sense to replace them both?)

Personally, I would use a multiband parametric EQ or an FIR filter. Start by recording clean audio, then use the EQ to add whatever control you need.

Without getting into the specifics of how to set up and operate a filter, here are some general tips:

  • Start by plotting a graph of the difference between these two curves. Determine how precise you need to be, and plot out a series of filters with center points and Q values that match the adjustments you need. Your final graph will show peaks where the black and red lines differ the most, especially at the left and right side of the curve.
  • Use a parametric equalizer VST plugin that allows for several bands in one instance of the plugin.
  • Find one that allows for using a graph to control the plugin: some plugins just have virtual knobs on the interface. What you want is one that lets you actually draw points on a line graph and shows you the resulting curve. Your final curve should match the plot you created.

The key here is in knowing the difference between your two graphs. Then you can plot the curve you need and implement it in your filter of choice.

I haven't evaluated any VST's recently, but here's a site with both an FIR and parametric EQ plugin. Both let you visualize your filters on a graph: http://www.reaper.fm/reaplugs/

Here's an article that describes using a parametric EQ: http://music.tutsplus.com/tutorials/how-to-use-a-parametric-equalizer--audio-2301

You can also get hardware devices that do the same job. I use a Berhinger Feedback Destroyer in my bass amp stack, and digital parametric equalizers can be had for around $300 online.

share|improve this answer
    
Parametric EQs are ideal for designing frequency responses yourself, not to match some a-priori given specification but to push a given signal in some direction. But if the desired response is already known, FIRs are much more natural and the obvious right tool if you're using digital implementations anyway (and aren't concerned too much with latency). –  leftaroundabout Dec 17 '13 at 0:53
    
I see your point... I'm looking at ReaFIR right now, and it seems to be more tailored around building your response curve from the graph, rather than designing virtual filters to adjust frequency response. In the end, it appears they do basically the same job, just with slightly different models. Also, it might help to explain what FIR is. As far as I can tell from the Internet, FIR is a tree. ;) –  TomXP411 Dec 17 '13 at 0:58
    
For simple tasks, IIR and FIR can certainly be used interchangeably. But they do differ in many regards. For instance, IIR becomes more expensive the more complicated the desired response is, whereas FIR has always the same processor load depending only on the resolution in the low-frequency end. For responses like these here – simple bass-rolloff but complicated high-frequency behaviour – this makes FIR a clear winner. (More important than processor: high-order IIR becomes unstable, but that's not very intuitive to understand.) — Didn't I explain what FIR is in my answer? –  leftaroundabout Dec 17 '13 at 10:25
    
Thanks for the detailed answer. I'll try the FIR filter. –  redreggae Dec 17 '13 at 14:26

When trying to digitally emulate some given frequency response, it's generally more effective to use not traditional EQing approaches (IIR filters) but FIR. How this basically works is quite easy to understand: you create a spectrum of the signal you want to modify, actually modify that spectrum in the obvious way (basically, scale each ω-point by the difference between source and target response) and transform that modified spectrum back into an audible signal. Yup, that's possible! Well, not exactly the spectrums: those are Fourier transforms, with phase disregarded. If you properly consider the phase, then the transformation is fully inversible (in fact, it's basically self-inverse, i.e. applying it twice will yield the original signal).

This has become a very common technique, dedicated Plugins include for instance ReaFIR. That, unlike parametric EQs (not to speak of graphic ones), allows you to specify a complete frequency response curve as fine-grained as your spectrums are.

share|improve this answer
    
thanks, very interesting. I think I have to read more about the theory. –  redreggae Dec 17 '13 at 14:36

Given the shape of the desired frequency response I'd say it might be a little tricky (around 8 kHz) using anything other than a graphic equalizer such as a 3rd octave type. You can get these as plug-ins.

Around 8 kHz it does look like it might be too much to expect from a multiband parametric equalizer (of the type that comes with Cubase) - the dip is a problem.

I'd also point out that room acoustics can dramatically alter the frequency response of a microphone and you may not get entirely what you expect from a GE however, this would tend to be more marked below 1kHz where sound relections are at there most annoying BUT don't rule out this being a problem.

It's not an exact science really.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.