I ran the same eight-bar disco loop through three stereo width tools last week, and only one of them made it to the second pass. The loop came off a Hilda Garman edit, one of the tighter ones, with a dry kick, a clipped guitar chop, and a string pad that already sits wide. Disco edits leave little room for fake width. Push the sides too far and the whole thing turns into a wet smear that sounds fine in headphones and falls apart on a club system when the room sums to mono.
So I treated it as a mono-compatibility test with a side of “does this feel good.” Three tools, same loop, same chain, same four-bar phrase on repeat until my ears got tired. One of them did the job. One was usable with caveats. One was too much, and I’ll tell you exactly where it broke.
First, the loop and the monitoring
Nothing exotic. The edit runs at 122 BPM, the kick is mono, the pad is a stereo patch from the original 1979 session, and the guitar chop has a hard pan dance going on. I printed it at 24-bit and kept the level peaking around -6 dBFS so none of the wideners had an excuse to pump.
I monitored on a pair of nearfields in a treated room, then checked on a single mono Auratone-style box, then on headphones. That mono check is the one that decides things for me. If a widener can’t survive a mono fold-down, it’s not a width tool, it’s a phase hazard.
Tool one: the standard mid/side widener
The first was a plain M/S widener, the kind that shows up all over the freeware scene. You know the principle. These tools “simply increase the volume of the Side signal relative to the Mid signal,” as one survey of free plugins puts it, and the whole trick boils down to that one fader move.
On the disco loop it felt good for about ten seconds. The pad bloomed out, the guitar chops spread past the speakers, and the kick stayed put because it’s mono anyway. Then I A/B’d against the dry loop and heard the problem. The center image went soft and hollow, like someone had scooped a tablespoon out of the middle of the mix. That’s the phasey center those M/S tools are known for.
Mono check: the guitar chop lost about a third of its body, and the pad developed a comb-filter flutter. Not a collapse, but not clean either.
Verdict: usable at maybe 15 to 20 percent wet, mostly on the pad, and never on the full mix. It stretches left and right and that’s all it does. It cannot put anything behind you or outside the speakers, because M/S processing only reshapes the stereo field you already have.
Tool two: Transpanner 2
This was the interesting one. Transpanner 2 doesn’t do mid/side at all. It works on crosstalk cancellation plus head-related transfer functions, the same physics your ears use to localize sound. In a real room, a sound from the left hits your left ear first and loudest, then wraps around your skull, slightly delayed and filtered, to reach the right ear. Your brain reads those interaural time and level differences and says “left.” When you listen to speakers, the left speaker leaks into your right ear and vice versa, and that crosstalk muddies the cues.
Transpanner 2 generates an inverted signal to cancel that crosstalk, so the audio beams more directly to the intended ear. Practically, it can place a sound outside the speaker cone or even behind you. On the disco loop I used it on the string pad, pushed the position slightly left and back, and the pad opened up in a way the M/S tool never managed. It felt like depth, not just width.
Caveat, and it’s a big one: this trick depends on a sweet spot. Move your head a foot off-axis and the illusion weakens. On a club system with people scattered everywhere, most of them will hear a slightly phasey pad rather than a 3D image. And it’s a monitoring-dependent effect, so if you’re mixing on headphones you’ll get a different result than the nearfields.
Verdict: great for a single element, especially a pad or a reverb return, where you want placement rather than loudness. Bad idea on a full drum bus. It didn’t ruin mono, but it did nothing useful there either.
Tool three: the one that was too much
The third was a “stereo imager” style plugin, the sort that combines M/S width with a Haas-style delay and a little high-shelf excitement. I won’t name it because the failure wasn’t unique to that brand. It’s a whole category.
At 30 percent width the loop sounded huge. At 50 percent it sounded like the record was playing in a tiled bathroom next door. The guitar chop started arriving a few milliseconds late on one side, the kick grew a ghost, and the pad smeared into the cymbals. That’s the Haas delay doing what Haas delays do: it creates width by offsetting one channel in time, and on a busy disco loop it just adds a second, blurrier copy of everything.
Mono check: the loop lost the kick punch entirely and the guitar turned into a phasey whisper. That’s the collapse the freeware M/S crowd warns about, except this one collapsed harder because the delay made the phase relationships worse before the fold-down even happened.
Verdict: too much. Not because the tool is broken, but because it solves a problem this loop didn’t have. A dry disco edit with a wide pad already has its width. Adding a Haas delay on top is like putting a second coat of paint on a wet wall.
What I actually kept
On this loop, the M/S widener stayed at 15 percent on the pad only, Transpanner 2 handled the pad’s placement, and the imager went back in the folder. The final print kept the kick dead mono, the guitar chop untouched, and the pad just wide enough that the mono check still held together.
One thing the mastering crowd gets right that I keep stealing: mid/side compression on the master can make the sides feel more pronounced without reaching for a widener at all. A gentle 1 to 2 dB of mid-channel compression does more for perceived width than most dedicated width plugins, and it doesn’t touch phase. If you’re already compressing the mid, you may not need a widener on the mix bus.
That’s the lesson. Width is not a knob, it’s a decision about where things sit. Two of these tools understood that. The third just made the room bigger and hoped nobody would check in mono. If you’re working with old disco stems, check in mono first. It’ll save you a revision.
