In my last post, I wrote about the use of a stereo widener and my panning decisions concerning the string section in My Father’s World. At the end of that post I mentioned some other options for stereo wideners. I thought it would be interesting and possibly educational to use the same familiar source material and perform a shootout between some of the various free stereo widening VSTs available.
[soundcloud url=”http://api.soundcloud.com/tracks/72022905″ params=”” width=” 100%” height=”83″ iframe=”true” /]
Blue Cat Widening Gain
There are a number of ways to create a wider stereo effect with existing stereo material. The most basic, and the one that uses only the existing material, is using a mid/side volume adjustment. Here’s a brief article from Blue Cat Audio on mid-side processing. The most significant advantage of using M/S for widening is that it works well with material that has strong transients that would be smeared by the processing that more invasive VSTs employ.
From the same company is our first stereo widening VST, the Blue Cat Widening Gain. In short, the plug-in separates the incoming audio into the material in the center and the material just in the left and right channels. You can then process (in this case, amplify) either the center or the sides separately. Here are screenshots of the Lissajous display (from the Flux Meter because it’s the best display) of the side amplified 12 dB and 24 dB. Note that when amplifying 24 dB I also had to drop the overall volume 6 dB to keep the output from clipping.
The left display is with the widener at +12 dB and the right display is +24 dB. Visibly, the +12 dB recording is a nice, round, stereo signal. +24 dB is about as flat horizontally as you will see (with one exception, below). When I listen to the tracks, the +12 dB version isn’t bad, although comparatively I like the effect I got with the Flux Stereo Tool better. The +24 dB version loses much of the center signal because I had to reduce the overall gain to keep the signal from clipping, and it also gets lost in the additional side material.
[soundcloud url=”http://api.soundcloud.com/playlists/3092157″ params=”” width=” 100%” height=”145″ iframe=”true” /]
The Voxengo MSED (Mid/Side Encoder/Decoder) pretty much works the same way at the Blue Cat Widening Gain. It can also “decode” a mid-side recording, too.
The Lissajous displays are practically identical to the Blue Cat, and so is the audio.
[soundcloud url=”http://api.soundcloud.com/playlists/3092484″ params=”” width=” 100%” height=”145″ iframe=”true” /]
Jeroen Breebaart Omnisone
Omnisone by Jeroen Breebaart combines the guts of a M/S stereo widener with a basic phase processor exemplified by the VSTs in the last category. It also has a somewhat disappointing Lissajous display.
The Width control is a standard M/S widener, but the Ambiance and Character controls can create an artificial side signal that is mixed in with the main signal to create additional wideness.
There are three displays and three samples for this processor because I created a separate entry for the Ambiance setting. The first two are Width set to 200% and 400% (max) with Ambiance set to zero. The 200% shot doesn’t appear too much different from the unprocessed signal, but 400% display is much rounder.
Then I dialed the width back to 200% and added Ambiance and Character both at 50%. I’m not sure what that funky little spike to the upper right of the display is, but since this is a shapshot of a real-time process, it may have just been a fleeting transient, anyway.
[soundcloud url=”http://api.soundcloud.com/playlists/3092621″ params=”” width=” 100%” height=”165″ iframe=”true” /]
I think the effect on these samples is a little more subtle. The 400% sample I can definitely hear stronger harmony voices. Since they are further out in the panning spectrum they are accordingly increased in volume. At 50% Ambiance I cannot hear a clear difference.
Voxengo Stereo Touch
Another offering by Voxengo, Stereo Touch uses stereo delay to create early reflections, simulating a stereo signal. Wideness of the signal is effected by the length and strength of the delayed signal. The delayed signal can also be filtered in the high and low frequencies.
Both scope displays show a more stereo signal than the original, although the narrow setting danced around surprisingly while active. To my ears the two “with” samples do add some width (and volume – take care) to the original source. It may be more pronounced on transient rich material, and at all settings should be mono-compatible since nothing is being added out of phase. Note that this process can also create a stereo signal from a mono signal, which mid-side and phase inversion cannot do. They must both have some stereo material to work with in the first place.
[soundcloud url=”http://api.soundcloud.com/playlists/3109795″ params=”” width=” 100%” height=”145″ iframe=”true” /]
German plug-in developer Brainworx specializes in plugs utilizing M/S processing, such as equalizers and compressors. It’s a little ironic, maybe, that this freebie from them uses Phase Inversion to create a wider stereo signal. Phase Inversion is a process whereby a little of the signal from the right side is phase-inverted and fed to the left speaker, and vice versa. Here is a great short article from Sound on Sound explaining how phase inversion creates a sound “outside” of the speakers. The downside is the characteristic “smearing” of transient-rich material and lack of mono-compatibility (the phase inverted information cancels out its original when collapsed to mono).
The bx_solo is marketed as a tool to check your mix – soloing the left or right side, the mid or the center. It also has a stereo width control that uses the phase inversion technique to create a wider signal.
At 200% the display shows a pleasantly round stereo shape, while the 400% sample is a bit squashed. Even at 400%, though, the audio clip is not unpleasant to listen to, although the middle is beginning to disappear a little (not so bad as the M/S at +24 dB, however).
[soundcloud url=”http://api.soundcloud.com/playlists/3096695″ params=”” width=” 100%” height=”145″ iframe=”true” /]
Hbasm’s Stereoizer (not to be confused with the commercial plug of the same name by Nugen Audio) is the free version of their Stereoizer Pro. This plug-in uses the most complete implementation of phase inversion that I’ve found. It includes an LFO (low frequency oscillator) to add some motion to the effect.
The pay version includes a low-pass and high-pass filter to limit the range of the effect.
When set at 90-degrees, the result is a fairly round stereo image. At 180-degrees, though, the display shows a straight horizontal line, indicating completely out of phase material. Played back in mono the signal would completely disappear. Listening to the audio, the 90-degree sample has a nasty artifact, a wobble in the mix that is distracting and ugly. In the 180-degree sample the wobble takes on the character of an old-timey recording, and, interestingly, the waveform is a sausage, indicating no dynamic range within the clip. Overall, I would avoid this VST and have already deleted it from my lineup. You can draw your own conclusions.
[soundcloud url=”http://api.soundcloud.com/playlists/3097907″ params=”” width=” 100%” height=”145″ iframe=”true” /]
Upstereo by quikquak is a simple-to-operate phase inversion stereo widener. None of the controls have numbers on them, so I estimated 150% and 200% widening. There are also simple EQ controls for high and low boost/cut, and a loudness saturation setting. There is also a gain control that I had to use because this VST increased the volume significantly. Also, when this VST was active I could not do a fast bounce of the audio – only real-time.
From the Lissajous displays, the 150% setting is slightly squashed, but the 200% setting looks like an elliptical galaxy. Listening to the samples reveals a much higher quality rendering than Stereoizer with more widening effect than bx_solo.
[soundcloud url=”http://api.soundcloud.com/playlists/3110113″ params=”” width=” 100%” height=”145″ iframe=”true” /]
Flux Stereo Tool
My last entry in this shoot is my favorite – the Flux Stereo Tool. This was the first stereo widener I ever tried and only used it first because it came recommended. There are separate gain and pan controls for the left and right input. The pan controls can be used for further refinement of the stereo width of the output. The additional pan slider at the bottom re-centers the entire stereo signal. You can have two different setups for comparison purposes and also use automation to switch between them during playback.
For this exercise I used the 50% (+6dB) and 100% (+12dB) settings to get an even comparison with the other limiters.
The displays show a nice round signal at +6 dB and a very squashed signal at +12, the flattest of any of the phase inversion wideners except for the Stereoizer. The Mid/Side wideners achieve signals this flat.
Listening to the samples, the +6 dB reveals a significantly wider image than without processing, and the extreme +12 dB setting loses much of the high-frequency center, including the solo violin.
[soundcloud url=”http://api.soundcloud.com/playlists/3110378″ params=”” width=” 100%” height=”145″ iframe=”true” /]
It’s good that we have these various options for stereo widening, because they all serve different purposes and program material. I will probably give Upstereo a closer look in the future, as I really liked some of the results I was getting in this shootout. If I have material with a lot of transients I will try out one of the Mid/Side wideners, or Voxengo’s Stereo Touch first. For material like my strings I’ll probably stick with the phase inversion method. For mastering where I’m looking for a subtle effect, I’ll try out several of these to see what works best.
To wrap up, here are all the samples of the wideners working at their moderate settings:
[soundcloud url=”http://api.soundcloud.com/playlists/3124244″ params=”” width=” 100%” height=”300″ iframe=”true” /]