Ping-pong delay widens the stereo image when the left/right repeats are just different enough in timing and level to decorrelate the sides while your brain still perceives one coherent event. It stirs (shifts, swirls, or destabilizes) the image when the repeats become strong “second arrivals” that compete with the dry sound for localization, or when feedback builds a moving pattern that keeps re-centering your attention.
What “widening” actually means with ping-pong delay
A stereo image feels wider when the two channels stop behaving like identical copies. Ping-pong delay can do that by making the left and right channels carry similar content at slightly different times, so the ear treats them as spaciousness rather than as two separate sources. In practice, widening is strongest when:
- The dry sound stays centered and dominant (or at least stable in its panning).
- The delays are clearly stereo, but clearly secondary (lower level than the dry, limited feedback).
- The timing difference creates decorrelation without turning into a distinct echo.
This is closely related to the precedence/Haas family of effects: within a short time window, the first arrival dominates localization, while later arrivals mostly contribute apparent width and spaciousness rather than creating a new “object.” (Q-SYS)
The timing zones: widen vs stir (practical ranges)
Think in three timing zones. The boundaries aren’t hard laws (material matters), but these ranges are reliable starting points.
1) Very short offsets (roughly 1–10 ms): “widening,” but fragile
If your ping-pong setup creates extremely short interchannel offsets, it can read as width because left and right are no longer identical. The risk is mono compatibility: short offsets can collapse or comb-filter when summed to mono, and the tonal change can be obvious on vocals, bass, or anything steady.
Use this zone when you can check mono and you’re working with sources that tolerate phasey coloration (many synths, guitars, textured percussion).
2) Short delays (roughly 10–35 ms): “widening” with clearer depth
This is the sweet spot where the delays are late enough to feel like space, but early enough that they usually don’t become a separate rhythmic event. The dry sound “owns” the position; the alternating repeats “paint” the sides. This is the zone most people mean when they say delay creates stereo width without sounding like delay. (Q-SYS)
3) Audible repeats (roughly 35–120+ ms, and/or synced note values): “stirring” becomes likely
Once repeats are clearly perceived as repeats, ping-pong becomes motion by definition: the energy keeps jumping sides in a way the listener can follow. That can be great, but the image is now being actively animated. Depending on arrangement, it may feel like:
- the source is moving side to side,
- the phantom center is weakened,
- or the whole mix gets “busy” in the stereo field.
Why feedback changes the result more than people expect
The same delay time can widen or stir depending on feedback.
- Low feedback (0–20%): You mostly get one bounce to each side. That tends to widen because the stereo information is sparse and doesn’t compete with the dry for long.
- Medium feedback (20–45%): Repeats build a pattern. Now you’re not just adding stereo difference; you’re creating a moving stereo object. This is where stirring becomes noticeable, especially on sustained sources.
- High feedback (45%+): The ping-pong line becomes part of the groove. Image stability depends on how rhythmic and how filtered the feedback is. Without filtering, this is where clutter and masking rise quickly.
A useful mental model: widening is “a little stereo evidence,” stirring is “a stereo storyline.”
Filtering determines whether the motion feels wide or messy
Filtering inside the delay loop (or after the delay) is one of the cleanest ways to keep ping-pong widening instead of stirring.
- High-pass the delay returns to keep low frequencies from alternating left/right. Low end that bounces sides tends to feel like the whole mix is wobbling, and it can reduce punch.
- Low-pass the delay returns so repeats get darker each bounce. Darker repeats read as depth, not distraction.
- Narrow the delay bandwidth so it sits behind the dry signal; you preserve width without making the stereo picture feel “busy.”
If you do nothing else: roll off lows on the delay return and keep feedback modest.
“Widening” setups: stable center, wider sides
These are patterns that typically widen without stirring.
Centered dry + stereo ping-pong return (subtle)
- Keep the dry track centered (or wherever it belongs).
- Put ping-pong delay on a send/aux.
- Set delay time in the short-delay zone (often 10–35 ms for “space,” or longer if you keep feedback very low).
- Keep feedback low so you get only a couple of audible bounces.
- Filter the return.
This works because the dry stays a stable “anchor,” and the delay is only a spatial cue.
Unequal left/right delay times (micro-asymmetry)
If your tool lets you set different L and R times, a small mismatch can increase decorrelation without increasing level. Many delay designs explicitly support separate L/R delay times or spread controls. (Valhalla DSP)
The key is to keep the mismatch small enough that it reads as width, not as a rhythmic flam.
“Stirring” setups: when ping-pong starts pulling the image around
You’ll usually hear stirring when one or more of these is true:
The delay return is too loud relative to the dry
If the first repeat is close in level to the dry, your brain starts treating it as a competing localization cue. The stereo field can feel like it “tilts” toward whichever side has the most salient repeat at that moment.
The delay time is long enough to sound like a second event
Once the repeat is clearly separate (often beyond ~35–50 ms, depending on source), the listener can track it. Alternating sides becomes attention-grabbing motion.
Feedback creates a repeating left-right “meter”
Even at moderate levels, multiple repeats can turn into a stereo pattern that draws focus away from the main element. The image isn’t merely wider; it’s moving.
Wideband repeats on dense material
Full-range repeats on vocals, busy synths, or dense guitars can stack into a stereo wash that feels like the stereo image is being churned rather than widened.
How to tell which one you’re getting (fast checks)
1) Mono check (non-negotiable for widening claims)
If the sound gets hollow, quieter, or changes tone drastically in mono, your “widening” is partly phase interaction. That may be acceptable, but it’s not a free win.
2) Correlation meter / vectorscope
- Widening typically pushes correlation downward a bit but stays mostly positive.
- Stirring (and especially phasey stirring) often shows big swings and can flirt with negative correlation during strong repeats.
3) Headphone vs speaker reality check
Ping-pong motion can feel larger on headphones. On speakers, strong side-to-side repeats can pull attention in a way that feels less “wide” and more “restless.” Check both.
A simple decision rule
- If you want wider but stable: keep the dry anchored, keep repeats quieter, keep feedback low, and filter the return.
- If you want animated stereo movement: turn up feedback and/or return level, use audible times (or tempo-sync), and let the repeats stay bright enough to be noticed.
why does this matter
Ping-pong delay is one of the fastest ways to change how “big” a mix feels, but it can just as quickly destabilize placement and clarity. Knowing when it widens versus stirs lets you choose space or motion intentionally instead of by accident.
Sources:
- https://blogs.qsc.com/live-sound/what-is-haas-effect-and-how-to-take-advantage-of-it/ (Q-SYS)
- https://www.soundonsound.com/sound-advice/q-how-should-set-stereo-widening-delay-effects (soundonsound.com)
- https://valhalladsp.com/2019/04/16/valhalladelay-the-controls/ (Valhalla DSP)
