Getting the basics - some help for someone new to mixing!

P

PhilLondon

New member
Hey all!

I've always just been happy with recording via VSTs just for the fun of it, and to be able to listen back and use DAWs as loopers and for basic demos and stuff, but I got a Kemper about a year ago aside from using it live I've started using it to record my stuff, and it really is a great tool so I want to start to make my recordings sound more professional and learn more in general.

I'm happy with the tone I've been recording, and I've been making sure my recorded level is at -12 to -18dB to ensure I've got enough headroom, and I feel that they are starting to sound somewhat balanced in relation to the instruments around them, and now I feel that I want to understand how to EQ, as I'm very new to it, and how to make my mixes louder and more polished sounding. I have not been concerned with volume, and have just been turning my monitors up, but now I want to be able to use the compression and limiter plugins to make the mix a louder and rounded, and learn how to EQ each instrument/part so that it sits in the mix better, but I really have no idea about it all.

I get that compression is meant to make everything the same volume, to stop peaks, but I don't get the parameters or how and when to implement them (on the output bus, on the input bus?), and the same goes for limiters (brick wall? I can't get it to do that), and that EQing basically allows you to make sure that each instrument occupies the space that it is dominant in, and cuts out the bits that don't really contribute to the sound and could cause other instruments to be swallowed up by, but that's really the extent of it.

Can you point me to some resources that can explain how to EQ and how to use compressors, limiters etc. Or anything else that I should know.

I'm using Logic Pro X with a Scarlett 18i8 via SPDIF. I've got a 15 inch Macbook Pro 2.5GHz (2014).

Thanks.
 
You have a fair learning curve ahead of you if you intend to understand every parameter of every device at the same time. My advice would be to disregard everything else and concentrate on adjusting one parameter at a time...and at least try to establish a basic idea of what it does. I'd start with the compressor, understanding that it's job isn't to make things as loud as possible....but rather to control the dynamics of a track so it's "easier" to place it into a mix. It may seem like subtle distinction but the the way you phrased it means to me that you'll be chasing levels rather than trying to create "space" in a mix. If you always try have everything at the same high level...it will reduce space in your mixes.

Some plug-ins are a LOT better than others too. Some give a very graphical representation of your signal and the way the device processes it...some leave you to wonder where you are at and make you guess at what you're doing. This clip ( https://www.youtube.com/watch?v=9GM9hWRru-Y ) shows the Sonitus compressor plugin that comes with Cakewalk Sonar DAW's. Note at about 3:00 when he shows you the interface with music playing. The moving dot represents the signal level and the adjustable curve represents your inputs to adjust the compressor. It doesn't take long with this kind of interface to begin to understand the parameters and what they do. If you don't have something like this I'd recommend looking at purchasing a product suite that does do this. The Sonitus suite only comes with Cakewalk but there are other packages out there that have good, graphical interface...you need that.

With EQ you also need a graphical interface to help train yourself to identify which frequencies are causing problems. That would be a spectrum analyzer. This is the best free plug-in I've ever found on the internet.... Real-time audio spectrum analyzer plugin (AU, VST) - Voxengo SPAN - Voxengo This will show your waveform in real time. You can learn a lot just watching it work. Used in conjunction with your EQ (the signal chain being 1. Wave 2. EQ 3. Spectrum analyzer) you'll be able to see your EQ adjustments reflected in the live waveform. After enough practice you'll memorize which frequencies belong to which problems...and how much to cut or boost of a particular frequency in order to fix it.

As far as where to stick an effect...it doesn't really matter but a good thing to keep in mind is that you can put an effect on a single track in a simple mix. However, if you have 38 tracks of guitar layers it's going to be MUCH simpler to create a bus named "guitars" and try to process them on the bus all at the same time. Your computer might be able to handle that much processing but your brain won't. :)
 
Last edited:
Thanks for your reply. I think that learning about compression is a good idea, as it seems like a vital tool that is applied to all mixes, and I'll also look at EQing. I've found a sticky for compression and a good video for EQing so I'm gonna look at them. Maybe just look at EQ and compression for the guitars only.

As for using. Another program, I feel fairly comfortable with Logic and I love the drummer program, and it cost me a fair amount so I think I'll work with it. Any other input would be amazing!
 
These two concepts EQ and compression is difficult to just say do X and do Y. As a fellow noob, I will tell you this. First get your tracking down, mix everything first and get your levels set. Don't put anything on it, just mix with everything natural, don't worry about how loud it is.

Once you get the volumes correct and then maybe work with some panning to get everything to sit right. Then, when you start thinking you need something louder and you turn it up, and then it messes something else up, that is when you look at EQ. Look at your charts to see where your instruments you are mixing sit as a guide. Use the EQ to cut or add for your mix to get the sound to come through without chasing volume.

I would compress at first only to even out the volume (compression can also be an effect), I would try fader riding first, but that is the original intent of compression (old analog days they didn't have automation so compression was used to even out volume).

Use extreme settings to see what everything is doing. Cut out highs and run it all the way out, cut lows, compress the hell out of everything, then turn it off. What I am trying to say is, you have to experiment to get a feel for it. It is all subjective for the most part and at the end, it is your call.

Once you get something, throw it up here and let those with knowledge give it a review. Ask a question, get many answers, and that is really what is great about it. This recording and mixing is part craft, part science, part art.
 
DM60 said:
Use extreme settings to see what everything is doing. Cut out highs and run it all the way out, cut lows, compress the hell out of everything, then turn it off. What I am trying to say is, you have to experiment to get a feel for it. It is all subjective for the most part and at the end, it is your call.
Click to expand...

I think that's some great advice right there. If you really want to dive in and get a better sense of what compression and EQ does, just start by turning some settings up all the way so that the effect is exaggerated and obvious. This gets you in the ballpark of at least understanding what can be done. Then back off the settings until you are able to progressively hear more subtle changes. Just play around with it. It's non-destructive after all, so no harm done.

Compression is more subtle than EQ imo, and as a result, more mysterious.

With EQ, I tend to just use my ears and sweep a narrow but high db boost across the spectrum listening for anything in that track that, when boosted to the extreme, just sounds horrible in conjunction with everything else. When I find that frequency, I cut it, typically by a much smaller amount than the boost I did to find it, and fine tune the Q and whatnot. You just have to play around with it a lot and keep listening I think.

Conceptually, I guess this is kind of vague, but I liked what DM60 said there and wanted to expand upon it. I read a lot about compression and EQ, but nothing helped me personally understand how they work more than trying them on a variety of sounds over time. Start with extreme settings just to get a sense of what's going on, then dial it back as you are able to detect more subtlety.
 
heatmiser said:
..With EQ, I tend to just use my ears and sweep a narrow but high db boost across the spectrum listening for anything in that track that, when boosted to the extreme, just sounds horrible in conjunction with everything else. When I find that frequency, I cut it, typically by a much smaller amount ..
Click to expand...
I see this approach mentioned often, and well, I see it as both kind of extreme and maybe not the way to go.
Sweeping with a narrow Q and boost is certainly good for teaching how the various freq bands relate to their tones, and you no doubt will find sometimes bizarre tone abortions in tracks. But it can be kind of a backwards way of trouble shooting. Similar to the logic of "don't eq in solo" - a rule I break but I'll explain.. I believe it sets sort of an unnatural condition for hearing what is needed.

Rather, I'd say listen to your tracks in the mix - and you'd likely be doing this in several phases focusing in on various things or combinations at different times, but try to identify where and what is out of balance.
In other words, if something is muddy, boomy', resonating (like an acoustic guitar track is a great candidate for that), the question becomes where (approximately in the frequency range), and which track - or tracks. Sometimes it can turn out to be a combination of a few tracks piling up.
If it's a narrow problem, 'sweeping can be just the ticket. Do it in the mix, or if you really need to get a clearer view' of the source- of a pair of 'suspects' if you like, do them in solo till you sus it out. Then zero the eq out, drop them back in to mix and let you ears readjust to this unnatural condition you're just 'whacked your perception with!
;)
Now go after your eq'ing with that in context.

For actually broader tone issues, sweeping' probably won't cut it.
 
Yes, I agree that trial and error is a great way to achieve understanding.

If I'm cutting a frequency, say 100hz, by 10db, does that mean I'm quieting it, but not silencing it? Why wouldn't I just go all the way down and silence it? And when would I?

Also, I was watching a video that talked about using a plugin that identifies that pitch and frequency of a tone. So if I wanted to find out the frequency that a snare hit lives in, I use this plugin and I kind of tune it to the snare hit, which will then give me the frequency. I can then dial out everything except the frequency that is identified. Does anyone know what this is called in Logic? I can't find the video, and is this useful?

Thanks.
 
Last edited:
PhilLondon said:
As for using. Another program, I feel fairly comfortable with Logic and I love the drummer program, and it cost me a fair amount so I think I'll work with it. Any other input would be amazing!
Click to expand...

I wasn't suggesting to change your DAW...only consider an effects plug-in suite for your DAW that had a more graphical interface.

I can't be sure the effects described on the Logic Pro web site are the ones you actually have...but it looks like that EQ has an analyzer built into it. So that's good. But the compressor looks like it would be hard to learn on (lack of real time graphics and attenuation curves).
 
StuGort said:
I wasn't suggesting to change your DAW...only consider an effects plug-in suite for your DAW that had a more graphical interface.

I can't be sure the effects described on the Logic Pro web site are the ones you actually have...but it looks like that EQ has an analyzer built into it. So that's good. But the compressor looks like it would be hard to learn on (lack of real time graphics and attenuation curves).
Click to expand...

Ahh I see.

I am confused by the compressor a little. I'm sure that this is embarrassingly stupid and obvious, but I'm going to ask anyway! I understand what all the parameters do, but I have a few questions. The first picture is nothing done to it, and the second is the compression I've added.

1.png

I have the drummer track, and the kick and snare hits are causing it to peak quite high, so I want to use the compressor to squeeze it a bit, but I don't get what the threshold parameter is relative to. So I'm thinking that if (based on the picture) I've set the threshold to -22dB, the ratio to 5.6:1, the attack is set to 17.5ms, the release at 660ms and the gain is at 3.0., it should reduce the volume by 5.7db to 23db when the meter hits -27.6dB, is that right? Although it's not reducing it to -17, -6.5 is where it's peaking.

For reference:
threshold to -22dB
ratio to 5.6:1
Attack is set to 17.5ms,
Release at 660ms
Gain is at 3.0
Knee is at 0.5

Screen Shot 2015-03-24 at 21.43.54.png

I've managed to compress the drums with these settings with a little reduction from -4.0 to -6.5db at it's peak, and it sounds pretty much the same. How does that look?

I understand what it's meant to do, but I don't get how it works in relation to the numbers and I don't get how to read the graph.Is the input gain on the left the pre compressed signal and the output gain the post compressed signal? Any help?
 
Last edited:
When it comes to EQ: It also helps to know where to look for "core frequencies" in a given sound. For instance ... in a kick drum, the meat of the whumph is found in the 70-100 hz range, the mud in the 300 to 800 range ... the click of the beater in the 2-4khz range ... all sound dependent. The snare has different core frequencies. A piano is all over the place. Acoustic and electric guitars are completely different.

But it really helps if you know exactly what frequencies in what instrument constitute the key ranges that make the instrument ... well ... sound like what it does.
 
PhilLondon said:
I have the drummer track, and the kick and snare hits are causing it to peak quite high, so I want to use the compressor to squeeze it a bit, but I don't get what the threshold parameter is relative to.
Click to expand...

Threshold is relative to 0dBFS, the absolute maximum digital level.

If you want to lower the peaks, don't use any make up gain. In fact, the limiter portion of your compressor may be the better option.

The graph with the diagonal-ish curve indicates the difference between the input (on the horizontal axis) and the output (on the vertical axis). The higher the ratio the flatter the top of the curve gets. A 1:1 ratio would show as a straight diagonal line from corner to corner. An ∞:1 ratio would have an absolutely horizontal top part of the curve. The threshold is where the curve starts to flatten out from 45°. A hard knee is when that curve transitions immediately from one angle to the other, while a soft knee has a smoother transition across the threshold.

The ratio only applies to the signal that goes above the threshold: with a 5:1 ratio any signal that goes over the threshold by 5dB at the input will be reduced to 1dB over the threshold for 4dB of reduction. With the same settings, a signal that goes over by 10dB at the input will be 2dB over at the output, a reduction of 8dB.

Attack and release control how fast reduction is applied and removed. A limiter is, for the most part, just a very fast compressor with a very high ratio.
 
Okay, of course, a ratio... doh.

Okay. So in relation to the chart. The peaks on the graph seem to be going to a maximum of -7, approx. The threshold is -22 and the ratio is 5.6:1.

So, when the peak hits -4.7dB on the input, which is approx 17 difference, the compressor kicks in and should limit it to approx -13, but it's only going to -10. That's with the gain off. Why is that?

Also, is the idea not that you can then increase the volume of that track, but that it's made all the quieter bits louder, which results in a perceived change in volume? Or does it facilitate the ability to do both?

Thanks again for your time.
 
PhilLondon said:
Okay, of course, a ratio... doh.

Okay. So in relation to the chart. The peaks on the graph seem to be going to a maximum of -7, approx. The threshold is -22 and the ratio is 5.6:1.

So, when the peak hits -4.7dB on the input, which is approx 17 difference, the compressor kicks in and should limit it to approx -13, but it's only going to -10. That's with the gain off. Why is that?
Click to expand...

Because your attack is set too slow to catch the very short transients of a snare or kick. Before the compressor can apply enough gain reduction the signal has dropped below the threshold.

PhilLondon said:
Also, is the idea not that you can then increase the volume of that track, but that it's made all the quieter bits louder, which results in a perceived change in volume? Or does it facilitate the ability to do both?
Click to expand...

Compression isn't there to solve lack of headroom or clipping, it's there to make the track sound better and/or fit better in the mix. If you're out of headroom then you need to turn everything down.

The simple description of a compressor's job is to bring loud and soft closer together, but it's way more complicated than that. It's quite possible to do the opposite with certain settings. Sure, if you compress the loud parts down and add some make up gain you'll get a perceived increase in volume, but there's nothing inherently better about that. Ultimately what matters is how it actually sounds in the mix.
 
Okay, I think I'm getting it now. Although it does seem to mess with the dynamics when you start to compress more and more. I'll post back tomorrow. It's bed time here.

Again, thanks a lot. :)
 
So, just to kinda clarify. If I have a recorded guitar part, and some parts are kinda quiet or I have a drum track and the snare hit is causing spikes etc, I use a compressor to make the dynamics a little more equal, so that it's a bit more level in the mix and so that certain parts don't get drowned out? And also to bring the peaks down in general if they are causing my mix to be a bit too hot?

Thanks.
 
PhilLondon said:
So, just to kinda clarify. If I have a recorded guitar part, and some parts are kinda quiet or too low, causing spikes etc, I use a compressor to make the dynamics a little more equal, so that it's a bit more level in the mix and so that certain parts don't get drowned out?
Click to expand...

That's the basics of it. But inevitably compressors get used for the aesthetic effect, the feel they impart to a track. It can get hard to separate practical dynamics control and creative effect. I find myself doing a lot of volume adjustment by editing clip gain and use a compressor for the final control and "vibe".
 
Okay, I think I'm getting it. I'm tracking a song with my friend, and I'm going to use it to work on and experiment. I'm using Logics Drummer program as we don't have live drums and I actually prefer it to my EZ Drummer plugin. I'm going to work on looking at what tracks need it in order to make it sit better in the mix, as a starting point.

I'll probably be back at some point asking silly questions.

Thanks for your help so far, I appreciate your time. :)
 
So, if I understand correctly, compression lowers the higher volume dynamic in order to balance the overall sound out, so that the lower dynamics are perceived as louder, therefore compressing the low and high dynamics of a performance closer together?

Someone mentioned automation, so I can just bring the volume up manually if it's required. I have some questions relating to this.

Why would I use compression if I can do this?

Why would I use automation if I can use compression?

And, when would I use both?

Thank you. :)
 
There are others who know much more than me, but a compressor is a generic setting for the track. Automation or even clip editing (I use Ableton's clip editing to pull down the highs so I can set my compressor less aggressive) you can remove the high peaks and not have to deal with them in the compressor. Example, take my peaks and reduce them manually (automation) when I set my compressor, I don't have to deal with those peaks and can adjust to the average more than to the extremes.

Not sure that makes sense, but in summary, by getting rid of the extremes, your compressor settings can be less aggressive leaving you more dynamics.
 
You must log in or register to reply here.
Back
Top