Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
If you’re mixing say a vocal, or a guitar, and you notice that you’re adding more than 5 processors to accomplish what you’re trying to, odds are you have conflicting processing.
For example, your vocal sounds muddy, so you add an EQ and attenuate some low mids.
Then you add a saturator, and it sounds muddy again. So you add another EQ or a multi-band compressor to cut out more of that range.
In short, the additive processor, the saturator is undoing the processing that came before it.
The first EQ you used attenuated low-order overtones. Then the saturator added them right back.
This can happen with reverb, delay, doubling, and other forms of additive processing.
Instead of doing what most engineers do and add more processors to fix a problem, address it where it’s happening.
If the saturator is adding back frequencies you don’t want, try a different type of saturation, or maybe don’t use it at all.
If you have this problem with reverb, try a different reverb with reflections that don’t accentuate that range, or pick one with an internal EQ that lets you shape the reflections.
Each processor you introduce inevitably degrades the audio. Granted very mildly, but noticeably after enough processors are added.
So, know what you want from an instrument, and listen intently as you introduce each effect to determine if they truly accomplish the desired effect.
This is similar to the last part of the video, but creates a different problem.
Some processors will inevitably make mixing significantly harder if they’re routed improperly.
For example, if I start a vocal chain with a reverb, consider all of the issues this is going to cause.
If I add compression next, then I’m not compressing just the vocal, but instead compressing the vocal and all of the reverb reflections. Not only does this make the compressor incapable of properly controlling the vocal, it makes the reverb sound completely different than it did originally.
If I add saturation next, then I’m distorting not just the vocal but the reverb reflections.
The same thing would happen if I saturated delay, a chorus effect, and so on. This is an obvious example, but take a look at your chains and see if there’s any routing that could be causing these types of issues.
Also, here’s a general chain that avoids these types of problems.
Subtractive EQ into Compression, then saturation, followed by your temporal effects like reverb and delay, with additive EQ at the end.
If you’ve ever used reverb as a send and noticed the mix sounds better than when it’s in the channel strips chain, that’s because the send is routed after all of the processors on the channel strip, so it’s the routing that’s improved the sound.
Once you understand the flow of the signal from one processor to the next you can decide if you want to follow what I recommended here, or deviate from it while still avoiding these issues.
This is a new one, mainly due to the popularity of resonance reducing plugins like Soothe 2 and now FabFilter’s spectral function in the Pro Q4.
I think this comes from a misunderstanding of what resonance reduction plugins are attenuating. In short, they’re likely just cutting higher order harmonics or overtones.
These plugins have no way to determine what is and what isn’t musically related to the song - in other words they can’t differentiate between a harmonic and a resonance that’s not musically related.
It just attenuates based on amplitude. Which means, if a harmonic is high in amplitude, which it usually is compared to other frequencies in the spectrum because it’s naturally part of the instrument’s timbre, then it’ll be cut.
So, the majority of the time using a plugin like this just means that musical aspects of the performance are being turned down.
If you attempt to do this with bell filters, that’s more reasonable - at least you can control the exact frequency that’s being reduced.
But if you’re notch filtering this, and using narrow bells on that, and you’re doing that consistently, I’d really recommend relaxing a bit.
Not only will it slow your mix process down, it’s not doing what you think it’s doing. You’re not removing problem frequencies, you’re focusing on a frequency and becoming acutely aware of it, which makes it easier for you to hear.
It’s called cross-modal facilitation. In short, when we combine visual cues and auditory cues of a corresponding phenomenon, it greatly increases perception.
When test subjects were presented with a 4kHz tone, the threshold for their hearing was reduced by 15dB if a visual cue was presented in tandem with the auditory cue.
In other words, you can hear the resonance better because you’re expecting to hear it, and visually you’re being cued in on its presence by an analyzer.
It’s not because it’s too loud to other listeners, its too loud to you because you’ve being primed to hear it.
So, instead of focusing on resonances just use gradual EQ curves. I guarantee you that after a couple mix sessions, your mixes will sound better and you’ll be able to actually enjoy the process of mixing.
Speaking of gradual frequency changes, there are 2 ranges that I often argue to be the most important.
Granted, all frequency ranges are important, but these 2 have the biggest impact on our perception of a recording.
That is, the low mids and the high mids.
Let me give you a good reason behind this.
The low mids, or roughly 150Hz to 400Hz. contains the most information in a song.
Starting from low frequency instruments to high, let’s consider what’s in this range.
The kick’s fundamental is likely below this range - let’s say it’s 80Hz.
The fundamental will likely be highest in amplitude; however, the 2nd, 3rd, and 4th order harmonics will also be high in amplitude.
In this example, that would be 160Hz, 240Hz, and 320Hz.
So the kick starts in the lows, but a lot of it’s information is in the low mids.
Say the root note of the bass is 110Hz. Again, the bass starts in the lows, but its 2nd and 3rd order harmonics are 220Hz and 330Hz. Respectively.
Then, say a vocalist sings notes that have fundamentals between 220 and 330Hz.
An acoustic guitar plays chords with the majority of the fundamentals between 110 and 440Hz, and so on.
You get the idea, so many overtones or harmonics, and fundamentals occupy this range.
Not every instrument occupies the lows. Not every instrument occupies the highs or even the high mids.
But almost every instrument in a recording will occupy the low mids.
So, if you’ve run into the incredibly common problem of a muddy mix, that’s why. Simply put, the majority of instruments and vocalists occupy this range.
So, what about the high mids?
If you’ve ever heard someone say that the high mid range adds clarity, they’re not being subjective. Our ear canals resonant at 3.5kHz in order to prioritize speech. Our vocal tracts create clusters of vowel and consonant information called formants. The most important of which is in this area - that being from 2-5kHz.
If something or someone has ever sounded muffled, that means this range has been attenuated, and the vital information our ears are designed to prioritize is lower than we’re accustomed.
So this range, the high mids, 2kHz-5kHz is incredibly important to listeners.
Here’s where it gets interesting - since 150 to 400Hz is often so high in amplitude due to all of the reasons we just discussed, it masks or covers the ranges above it. This includes 2kHz-5kHz.
So not only can the low mid range be too high in amplitude, it can make other vitally important ranges sound quieter.
I believe this is a big reason why the the 20Hz. boost and cut technique on a Pultec sounds so good, especially when processing a stereo mix or a master. The lows are boosted, meanwhile there’s a dip to the low mids.
By boosting it, an engineer’s first reaction might be, oh the lows sound so much better. The kick and bass are more impressive. But that’s a small piece of what’s actually happening.
The kick is higher in amplitude, while the low mids are attenuated. Subsequently, masking to the high mids is reduced, adding clarity.
Pair this with a small boost to 3kHz and you have 2 filters with the potential to greatly improve the balance of a mix.
So, this just ties into not focusing too much on narrow-bell or notch filters. They can be important on occasion, but always keep the low mids and high mids in mind.
Let’s check out those 2 simple and gradual Pultec filters I mentioned and notice how much they can improve a mix.