You are here

Mid-Sides EQ'ing In Sonar

Sonar Tips & Techniques
By Craig Anderton

The LP EQ is about more than a visually attractive interface, as it offers several advanced features not found in Sonar’s older LP64 linear-phase EQ. Note that the display scaling is adjustable for a  ±6, ±12 or ±18 dB range.The LP EQ is about more than a visually attractive interface, as it offers several advanced features not found in Sonar’s older LP64 linear-phase EQ. Note that the display scaling is adjustable for a ±6, ±12 or ±18 dB range.

Sonar’s new LP EQ lets you perform Mid-Sides EQ’ing with ease.

The Sonar 2016.04 April update featured three new plug-ins: the TH2 Cakewalk Edition amp simulator, the LP MB linear-phase multiband compressor, and the LP EQ linear-phase EQ. Both linear-phase plug-ins offer substantial improvements over the older LP64 EQ and LP64 Multiband versions. In particular, the LP EQ was designed with Mid-Sides processing in mind, as well as conventional stereo. So, that gives us a good reason not just to take a look at the new EQ, but also to discuss Mid-Sides processing in Sonar.

Meet the LP EQ. This is not a refreshed version of the LP64 EQ, but a ground-up redesign. You can add up to 20 nodes, and each can be a low shelf, high shelf, high-pass filter, low-pass filter or bell curve. However, there’s a certain amount of context-sensitivity on display when adding nodes. If you double-click to enter a node close to the highest possible frequency, the LP EQ will insert a low-pass filter; double-click at a somewhat lower frequency, and the new band will have a shelving response. A similar process operates at the other end of the frequency spectrum, and of course you can change the shape of bands once created. The interface is graphical, so you can drag nodes horizontally to change the frequency or up/down to vary amplitude. A right-click drag on a node alters the width, as does using the mouse scroll wheel on a selected node.

You can draw a marquee around multiple nodes to select them (Ctrl+click also works to select multiple nodes) and move them all at once, but there’s an interesting twist. Suppose one selected node is set to boost, and another to cut. If you click on the one that boosts and drag it downward, the amount of boost will decrease, while the one that’s cutting will decrease the amount of cut and start boosting. This kind of complementary motion is a way to increase or decrease the overall emphasis easily; for example, if you think you went too far with the amount of EQ and want to pull it back, this scales back the entire curve equally. (If all the selected nodes are either boost or cut, then their amplitudes vary together.)

These basics give a flavour of the features, but I highly recommend clicking on the UI to give the LP EQ the focus, and then pressing F1 to call up the well-written and comprehensive documentation. Rather than duplicate that information, let’s segue into applications.

Side Show

Mid-Sides (M-S) EQ processing is definitely one of the ‘big deal’ features. If you’re not familiar with M-S processing, it starts by encoding a stereo track into two separate components: the centre becomes the ‘Mid’ component, while the stereo track’s right and left elements become the ‘Sides’ component. You can then process these components individually, before eventually decoding back into stereo. Although it’s possible to ‘construct’ this type of processing (as described toward the end), the LP EQ can automatically encode the signal into the Mid and Sides components for processing, while decoding the results back so you can hear the results in standard stereo.

To get started with Mid-Sides processing, click on the LP 64’s Expert button and under Mode, choose Mid-Side. For best results, set the precision to High. This results in the most latency but the highest accuracy, which is important because with Mid-Sides processing, you don’t want any phase shift or sample misalignment — that will interfere with the decoded stereo imaging.

Processing can be independent for the Mid and Sides components (as well as for left and right channels in conventional stereo applications). You assign a node to the appropriate component by clicking on the node, and then clicking on M or S (toward the LP EQ’s upper right corner). This is exceptionally useful for mastering, because it lets you get ‘inside’ the file to do pseudo-remixing on a stereo track. One typical application is giving a slight boost to the higher-frequency Sides components to provide a bit more ‘air’ and a wider stereo image. And in case you’ve been seduced by vinyl’s comeback, remember that it’s crucial to centre the bass and minimise bass excursions in the sides. Mid-Sides EQ processing is ideal for this application because you can reduce the bass in the Sides, and if needed, increase bass a bit in the Mid channel. Even if you’re not mastering for vinyl, taking this technique further can give a super-anchored, ‘centre-channel’ bass sound.

However, don’t ignore what Mid-Sides processing can do with individual instruments. Drums with lots of room ambience can benefit from a bit of upper mids in the Sides channel, and a little bit of lower mids in the centre to accent the kick. With synth basses whose wide image ‘steps on’ other instruments, you can bring down the bass in the Sides.

Mid-Sides processing is also helpful when taming reverb. Set a node to Mid, select the high-pass curve, and slide it all the way to the right to take out essentially everything. Then you can shape the remaining reverb with the Sides EQ, but the main point here is that the reverb will be out of the centre, where it can muddy the bass and kick.

LP EQ’s Expert Mode is the key to doing Mid-Sides EQ processing. Also note the Mix control for doing parallel processing.LP EQ’s Expert Mode is the key to doing Mid-Sides EQ processing. Also note the Mix control for doing parallel processing.

But how do you know whether you’re really making an improvement to the sound or not? The LP EQ includes a Mix control (accessed in the Expert section), so you can vary the mix from full EQ to no EQ. Yes, parallel processing for EQ! Very handy, and even better, the Mix control can be automated (like virtually all other parameters, including display characteristics and bypass). If you’d rather do a quick switch than a mix, there’s also an A/B comparison function.

Stuck In The Middle

After playing with the LP EQ you might find yourself hooked on Mid-Sides processing. Fortunately, if you want to take this beyond EQ, you can use Channel Tools to encode files into Mid-Sides for processing with any plug-in, and then decode back into stereo again. Conveniently, the Channel Tools settings are the same whether you want to encode or decode.

To encode, insert Channel Tools in your stereo track. Set the Mid and Sides gains to 0, tick the Mid-Sides box, and pan the left slider all the way left and the right slider all the way right. Now the stereo track’s centre (Mid) is in the left channel, while the Sides information is in the right channel. Select the clip, then choose Process / Apply Effect / Audio Effects. Your file has now been encoded into a Mid-Sides file. (Karaoke fans, take note: because vocals are usually in the centre, if you listen to only the Sides channel, you can sing along and not hear the original vocals — although you’ll also lose anything else that was panned centre, like the kick and bass).

To split the two into individual tracks for processing, select the clip and, under the Tracks tab, choose Bounce to Tracks. Under Channel Format, select Split Mono. For Source Category, choose Tracks. You’ll end up with two individual mono tracks, which you can process individually.

These Channel Tools settings work both for encoding stereo files into Mid-Sides components, and for decoding Mid-Sides audio back into conventional stereo.These Channel Tools settings work both for encoding stereo files into Mid-Sides components, and for decoding Mid-Sides audio back into conventional stereo.

After processing, you need to recombine and then decode the tracks. Reduce the level of each track by 3.0dB, or the Master bus level by 3.0dB (if the processing did additional boosting, you’ll want to lower the level further; you can always raise the level of the decoded file as needed). Pan the Mid track fully left, and the Sides track fully right. Select both tracks. Under the Tracks tab, again choose Bounce to Tracks but this time, under Channel Format select Stereo. For Source Category, choose Entire Mix. After bouncing, you now have a processed mix that’s ready to decode. To do the actual decoding, apply the same Channel Tools settings used for encoding.

However, there are some caveats that apply when breaking tracks into separate Mid and Sides components. These need to be aligned with sample accuracy to maintain proper stereo imaging after decoding, and this is why the LP EQ’s phase-linear design is important, because it doesn’t cause phase shifts that can alter the imaging. And although Sonar is excellent in terms of handling delay compensation, that doesn’t help if the plug-in itself emulates analogue operation and doesn’t have phase integrity.

Regardless, rules are made to be broken and sometimes a less-than-perfect adventure in Mid-Sides processing might be what you actually want. Why be normal? Never be afraid to experiment!

Published June 2016