This is a topic we’ve looked into before, so for this video, I came up with a brand new chain. It’s a little unorthodox but has a cool sound so let me know what you think of it.
I’m going to start with subtractive EQ, but instead of simply attenuating problem frequencies, I’ll create multiple, high Q value bands, all attenuating out-of-key notes. For the demo in this video, the key is E-Major, so I subtracted all out of key notes by a small amount.
Let’s take a listen, and notice that although subtle, it sounds slightly more transparent, and a little more musical.
This track had sibilance that I wasn’t a fan of, so I introduced the Weiss de-esser to attenuate it - the problem was, it also was cutting into the snare. Using the monitoring function, I found a bandwidth, range, and threshold, that attenuated sibilance, while leaving the snare relatively alone.
Let’s take a listen and notice how the sibilance is tamed, but the snare passes through almost unaffected.
I like upward processing or dynamics plugins that amplify the signal from the quieter parts up instead of affecting the peaks. The Oxford Inflator is great for this - the best way to use it in my opinion is to use 100% of the effect, then reduce the input to compensate.
Let’s listen to how the plugin improves the sound.
With a compressor designed for master bus processing, I’ll subtly attenuate the signal by no more than 1dB. For this track, I enjoyed feed-forward compression, with a low ratio, moderate-knee shape, a 5ms attack, and a roughly 200ms release to keep the sound detailed.
Let’s take a listen to how this slightly thickens the sound, due to the subtle compression and make-up gain.
This next plugin is similar to the inflator we used in chapter 3, but it offers mid-side processing. In short, it amplifies quite details until they hit a ceiling, at which point it compresses - however, if you select the wide setting, the signal is processed as mid/side.
Whenever the mid is attenuated or hits the ceiling, the side will become louder relative to the mid, causing stereo expansion. Let’s take a listen and notice the sound is fuller, and on occasion, wider.
This is the last processor I’ll use for upward compression in the chain - in short, I’m using a multi-band dynamics processor, and using compression with a positive range. This means that whenever the signal is above the threshold, it is amplified until the signal get’s louder.
This way I can control the dynamics a little more while increasing the level of frequencies I want more of. Let’s take a listen.
There are a lot of great saturation plugins, but I like using Saturn 2 since it lets me modulate various aspects, creating more complex and program-dependent saturation. With it, I utilized 3 different saturation emulations, before creating an envelope follower which measures the transient of the incoming signal.
I linked this follower to various parameters including the drive dial and drive’s mid-side panning - this way when the transient hits, there’s more saturation which is oriented toward the mid image in the low and mid frequencies, and the side image in the high frequencies.
I also modulated the dynamics dial of each, so that the transients are expanded, and altered the bandwidths, so that when a transient hits, the low and high frequencies’ settings occupy more of the mids.
Let’s take a listen, and notice how the saturation has a unique quality since it responds to the incoming signal more than a typical saturator.
Next, I’ll send the signal to a bus, and with the free plugin MSED, I’ll isolate the side image. Next, I’ll add the Arturia’s Neve emulation Comp Diode-609, to the side image and compress, both because it sounds great and offers dual-mono processing.
The dual-mono setting is important. Prior to the compression, you’ll notice the isolated side image is phantom mono, meaning the left and right channels are identical.
With dual-mono compression, we get slightly differing responses, causing variation between the left and right channels, and making our side image all the most interesting and complex.
Let’s take a listen.
Right now the original stereo track and the parallel isolated side image are being sent to my stereo output - but I’m going to do something a little weird and separate the stereo output into left and right channels. I’ll then insert the Gullfoss EQ and use identical settings on both.
Even though the settings are identical, each track will be processed differently by the EQ, since each channel is feeding it unique information.
Let’s take a listen and notice how the Gullfoss EQ cleans up the sound and brings forward more details.
Having our signal separated into left and right channels gives us another advantage- that is, we can more closely emulate the processing of an analog chain. In real life, these Pultecs are signal channel processors, meaning, we’d need 2 of them, 1 for the left, and 1 for the right.
That’s the exact setup we have now - I’ll use identical settings on them because I want to avoid some mild phase cancellation between the 2. But if you don’t mind the variation, try subtly differing settings.
With that in mind let’s listen with identical, then slightly varied settings.
Last up, we need to limit the signal - so let’s introduce 2 limiters, one for each channel, and use identical settings to keep the sound cohesive. What’s great about having the channels’ limiting completely separate, is that I can amplify the signals to right before peaking and still achieve loudness.
For example, I barely need to attenuate to get to an integrated -12.5 LUFS on the combined stereo signal, and that’s after lowering the output of each limiter by .5dB.
Let’s listen with limiting enabled, and notice how we achieve loudness while retaining an incredibly clean signal.
