View Profile Blackhole12
Applied Mathematician and Software Architect who occasionally writes music.

Erik McClure @Blackhole12

29, Male

Software Engineer

University of Washington

Black Sphere Studios

Joined on 9/21/06

Exp Points:
2,232 / 2,500
Exp Rank:
Vote Power:
5.67 votes
Police Officer
Global Rank:
B/P Bonus:

EQing with FL Spectroman

Posted by Blackhole12 - August 22nd, 2009

See the following tutorial with REAL IMAGES! --> [link]

What is EQing?

EQing (EQ stands for Equalize) is part of the mixing process - while similar tools are used in mastering, mastering is something that pretty much no one on NG actually does (save for B0UNC3 and a couple of other guys). What I am going to be teaching you here is how to properly EQ your instruments in the mixing process so they don't run over each other and rape your speakers. This often allows you to increase the volume of your track, which is why its so often mistaken for mastering.

I will first give you a short rundown of the parametric EQ 2, (a more extended tutorial can be found [here]), explain why clashing frequencies cause problems, demonstrate how to remove them, and then show you how the FL Spectroman can be used to make this process far easier.

The Parametric Equalizer 2

This is a tool that was added in FL studio 7 and is ridiculously helpful when doing EQing on your instruments, since it shows you areas of loud frequencies on the EQ bar itself. Note that all of our actual EQing work is being done on the instrument mixer channel. This means that you need a mixer channel for all of your instruments, and you need a parametric EQ 2 on all of those channels. Thus, as you are playing around with the EQ you can see what it is that your actually doing. It is important to realize, however, that the parametric EQ is far more powerful then it looks at first glance. The basics you should already know - there's 7 control points, and you can adjust their effect by either moving them up and down in the graphical representation (called a token) or by moving the bar on the right that corresponds to that control point up and down (see figure 1).


Below there, there are two dials. One changes the control point's frequency (labeled FREQ), and the other controls how wide its area of effect is (labeled BW). This knob is your friend. You absolutely must utilize how wide a given control point is if you want to do any sort of meaningful equalization on your instruments without ruining how they sound. [image]

Now, if you look above that bar, there are two more configuration options. One is a dot - this dot can be clicked by your mouse and dragged up or down. It will either change into up to four solid dots, or up to four hollow dots. These signify the eight possible slope values that the control point can have. As you can see in figure 4, these can have a huge effect on the overall shape of the control point. Manipulating the shapes and intensities of control points is key to getting an instrument to both sound how you want and prevent it from running over the other instruments. [image]

Above the dot is the symbol for your band type. Right now its just a basic -<>-, or possibly a -< if its on the end. This is not the only available band type - you can click and drag this value as well to browse through all the available band type shapes. [image] Here is a list of them:

[image] OFF - This turns the control point off and removes the token from the graphical representation of the EQ. USE THIS OPTION. If you aren't using a control point, turn it off, otherwise it will consume CPU. Optimizing your EQ can be helpful in getting a few clock cycles back.

[image] Low Pass - This filter will kill all the frequencies to its right-hand side, hence why it is called a low pass filter. We'll be using this shortly for frequency edge control

[image] Band Pass - This is a combination of Low pass and High pass. Frequencies to the right will slowly fade out until being cut off altogether, as will frequencies on the left.

[image] High Pass - The opposite of a low pass, removes all frequencies to the left.

[image] Notch - This kills off whatever frequency your control point is currently on and any that are close by. We'll go into where this is useful later, but it's not actually used as much as you'd think.

[image] Low Shelf - Default of the rightmost control point. This creates a shelf effect, lowering all frequencies to the left of it to a given value.

[image] Peaking - Default of the 5 center control points. This is the normal behavoir of the control points as I'm sure your used to.

[image] High Shelf - Opposite of Low Shelf.

Note that you can right-click a control point's graphical token to display a menu for all its possible band types and slopes, as well as an option to reset it. [image]

Now that we know how to use the parametric EQ 2, we need to know why we're using it. If you look at figure 14, you can see that at 369 Hz (hover your mouse over the window to get the exact value) we have a very loud spike (called a choke-point) that's causing clipping issues. One might first think that we simply need to lower the volume, but in fact that spike is whats causing the clipping, and if we are smart, we can remove it without damaging the rest of the instrument's sound.


Now, I'm sure the thought of "NOTCH IT OUT!" has occurred to you, but that is bad. Very bad. If you try to notch out that frequency, even if its with a very steep, thin notch, you will ruin the instrument. If we want to remove our static, we must be more precise then that. This calls for one of our more surgical tools - the peaking filter with a Steep 8 setting at the lowest bandwidth possible. [image] Using this, we can reduce the volume of only that frequency, leaving the frequency intact, with the clipping issues gone. Figure 15 shows that I reduced the volume by about 8 dB. This is the absolute minimum required to remove the clipping. Any more and we start to infringe on the note itself and ruin the instrument. This leads us to the primary lesson:

You are not a ninja. Your job is not to attack the frequencies like they're coming after your mother, your job is to delicately push them around until each one of them is at the perfect volume so as not to disturb the rest. In some cases, you will have a giant blob of frequencies that you need to take down (this is especially true when it isn't a single note or frequency that is the problem, but rather a certain range of notes). If its a whole lot of really nasty frequencies, you'll want a gentle slope so you can lower all of them.

Sometimes, however, it isn't quite as obvious where the problem is. I located this problem by finding the brightest band, but if there are a lot of brighter bands (especially in the higher frequencies), it's can be difficult to figure out which one is actually causing the problem. Your problem areas tend to be the lower middle to the upper middle, so chances are its somewhere in there, but this is where the Spectroman can come in handy. Figure 16 shows a situation that has a significant clipping problem, but just by looking at the frequency bands you'd never be able to figure that out. Trying to fix this would be a large amount of trial and error... but we have the Spectroman.


FL Spectroman

The Spectroman, when you simply add it on to your mixer track at first, is a dot-matrix of frequencies. This can be handy for other reasons, but for our purposes it is rather useless. In the upper right there are two small buttons that control output mode. [image] Click on the gray one, and you'll get a much more useful output. Now we are shown a bar-graph of the frequency output, along with the peaks. Unfortunately, this isn't particularly helpful right now because its just telling us the same thing the parametric EQ was telling us. However, that's where the Stereo button comes in.


Clicking this button splits the output into red (left) and blue (right). Using this, we can instantly tell that we have 2 frequency issues at 1300 and 850. However, if you look closely, you'll see that the 850 peak is misleading, because if you were listening to the clip like I was, you would have noticed that the 850 peak happened before the 1300 peak, and had no clipping associated with it. The peak seen on the spectrometer is simply an afterimage. It isn't the cause of the clipping, so all we need to worry about is the 1300. As you can see in figure 20, this was a much wider frequency issue then the last one, so I've tweaked the bandwidth accordingly. This demonstrates how the FL Spectroman makes it a heck of a lot easier to pinpoint what frequency is actually causing the problem - more often then not, it isn't what you thought it was. Be careful about red herrings, like the 850 one here. Even though it was obviously a high peak then 1300, it wasn't causing the problem. [image]

Things can get more complicated when you have multiple instruments that, when played alone, do not conflict with each other, but when combined, cause clipping issues. This is always the result of a residual frequency peak that while not significant enough to be noticed during solo play, combines with a corresponding residual peak in another instrument that creates a problem area. Usually a good thing to do is to stick a parametric EQ on your master channel (not a Spectroman!) so you can get a feel for what the frequencies look like at the trouble spot. [image]

The reason you don't want a Spectroman is because the parametric EQ 2 offers much higher resolution and is, in general, just a lot easier to use to get a feel for the troublespots. The Spectroman is designed to allow you to pinpoint very specific problem areas, not for general EQing. Use the parametric EQ's graphical display for that.


As you can see in figure 21, there is a problem in the 300-500 range. This is fairly common, because most of the audio information is contained between the 200 and 2000 Hz, so most frequency clashing will occur there (with the exception of bass, because everyone artificially increases bass to completely ridiculous levels, inevitably causing distortion. This practice is so common there is an entire set of specialized techniques just for maximizing bass levels which involves compression and spreading out the frequencies. It's disgusting, really).

The problem is that because most of the audio information is in this range, aggressive reduction of frequencies will murder the presence of your instruments and destroy your song. Thus, it's something of a perfect storm. You need all the information in the 200-2000 Hz range, but it also causes clipping. Lower the volume too much, and you'll only hear the frequencies in this range and lose your bass and upper frequencies. This turns the entire thing into a delicate balancing act, trying to lower the frequencies or enhance others just enough to eliminate clipping problems without nailing the feel of the instrument.

A tip - In general, you shouldn't need a compressor on your instrument unless it is very close to a pure sine wave. Properly EQed instruments usually can brickwall against the 0 dB barrier without actually causing clipping. You'll still want to put a limiter on there eventually, but you should be able to at least get away without having one. There are, of course, may exceptions to this. One of course being the confounded bass - you'll want to compress that. When seeking a good bass response you want to isolate the frequencies that produce a better response.

Going back to our problem of multiple instruments, once you've found the issue in the parametric EQ you'll want to open up the trouble spot, stick it on loop, and pull up either a parametric EQ 2 or a Spectroman for both of them at the same time (which one you use is personal preference). What your looking for is corresponding peaks, NOT the highest peaks. These are usually mid-ranged peaks that look fairly mundane until you realize both instruments are peaking on the exact same frequency. If you are using the Spectroman, this is a good situation to use your AMP knob - it will increase the amplitude of the graph and allow you to look at the smaller peaks in greater detail while letting the larger ones run off the top.


Once you've found the problem frequency, tweak both instruments (usually there's on instrument that's causing most of the problem) until they start cooperating with each other. Note that while doing precision EQing, you can always increase the volume of the instrument temporarily to exaggerate the clipping problems. Consequently, you should never have any sort of compressor on a mixer channel-chain that you are EQing. It has to reach the speakers completely untouched or you're not EQing the instrument, your EQing the compressor.

EQing, of course, can be used for more then just eliminating problem frequencies, you can also use it to enhance how your instruments feel. Increasing the high frequencies (7000-15000 range) can help remove the feeling of an instrument being played "behind a wall." Increasing 200-500 range can help for pads that should have a strong presence and overall "warm" feel. The best advice I can give you is to experiment around with them. For example, I tend to "cap" most of my EQs with a lowpass filter on the high end to exaggerate some of the 8-10k frequencies while removing the unnecessary 15k+ ones.

A similar approach can be done for bass, where completely inaudible frequencies can choke an instruments frequency range. Eliminating those lower tones can help a track breath. Some synths also exhibit bizarre glitches that can cause either extremely low or extremely high frequencies at exceedingly loud volumes. These can't even be physically played by the speakers, let alone be heard by you, but they'll confuse compressors and anything that's dealing with the actual wave because of how deformed it becomes. Using an EQ, you can easily eliminate these problems.

And with that, I conclude this tutorial. Happy EQing!

Comments (3)

great stuff! very handy/

I'm glad you found it useful 8D

Wow, that was quick... Thanks ^^. I guess I'll finally know what EQing is :D.

your welcome :D

Thank you so much for this tutorial! You really really helped me out so much! I can't thank you enough my man!