Why Your Breville Shot Tastes Sour on One Side and Bitter on the Other (And the Cheap Fix)

You pull a shot. One side of the cup tastes sour, thin, almost sharp. The other side is bitter and heavy, like it ran too long. Same beans, same grinder, same dose as yesterday, and yesterday's shot was fine.

That's not two problems. It's one problem showing up as two flavours, and it has a name: channelling.

What Channelling Actually Is

Water always takes the easiest path. Pack a bed of grounds and it's never as uniform as it looks, even from a good grinder. Some patches sit looser than others, from clumping, static, or an uneven dose. Under nine bars of pressure, water finds the loose patches first and rushes through them in the opening seconds of the shot. It barely touches everything else.

The rushed patches under-extract. That's your sour side. The rest of the puck still sees water for the shot's full running time, even though most of the flow has already found an easier way through. Sitting under pressure that long, those denser patches over-extract. That's your bitter side. You taste both because the cup in front of you is one mixed pour from two very different extractions happening at once.

Worth knowing: this isn't something that develops gradually through the shot. It's decided in the first two or three seconds, before you've even started your timer. By the time you notice a thin, fast stream or a shot finishing early, the water has already picked its route. Everything after that is just the shot playing out the decision that got made at the start.

Pre-infusion, where the machine ramps up to full pressure gradually instead of hitting nine bars straight away, is meant to help here. It gives the puck a moment to swell and seal before the full push arrives. But pre-infusion only helps a puck that was even to begin with. On a dose that's already clumped, pre-infusion just channels a little more gently before the same problem takes over.

How to Know It's Channelling, Not Something Else

A few signs point specifically to channelling rather than a grind or bean issue:

• The portafilter spits, spurts, or sprays instead of running two slow, steady streams.
• One side of the spout runs thin and fast while the other barely dribbles.
• The shot finishes faster than usual but tastes worse, not better.
• The same beans and grind setting were fine yesterday and are off today, with nothing else changed.
• Sour and bitter both show up in the same cup, rather than one flavour dominating cleanly.

If the whole shot tastes flat and sour, that's more likely stale beans or a grind set too coarse. If it's evenly bitter and slow to pour, check for too fine a grind or too heavy a dose. Channelling produces a split, not a uniform miss. That split is the tell.

Why People Blame the Wrong Thing First

When a shot goes wrong, the instinct is to blame whatever changed last: new beans, a slightly different grind number, burrs that need a clean. Sometimes that's the right call. But channelling produces exactly the kind of shot that looks like a dial-in problem, because you're tasting sour and bitter in the same cup and assuming you're either under or over-extracting, not both at once in different parts of the same puck.

The giveaway is inconsistency without a clear cause. If you nudge the grind size and the split just moves around instead of resolving cleanly, distribution is the more likely explanation, not extraction time. Grind size controls how fast water moves through the coffee. It doesn't control whether the water is moving through all of the coffee at once in the first place.

Why This Shows Up So Often on Breville Machines

Most Breville home machines run a 54mm basket, the Barista Express, Barista Touch, Bambino Plus and similar. The higher-tier Dual Boiler and Oracle run 58mm. A 54mm basket has noticeably less surface area than a 58mm one, so the same small clump of grounds accounts for a bigger share of the puck. A distribution issue that would barely register on a wider basket shows up fast on a 54mm one.

Breville's integrated conical burr grinder also feeds straight into the portafilter, which is convenient but skips a step most higher-end setups build in: distributing the dose before it's tamped. Grinding straight into the basket tends to leave grounds heavier near the walls than the centre, or the other way round depending on the burr geometry, plus a static cling that makes fine particles clump instead of settling flat. None of that is a fault with the machine. It's just what happens when the dose goes from grinder to basket to tamper with nothing in between.

Some home baristas fight the static at the source with a light spray of water on the beans before grinding, sometimes called RDT. It helps with clumping in the grinder itself. It does nothing about how the dose settles once it's actually sitting in the basket, which is the part that decides whether your shot channels.

The Cheap Fix: Sort Distribution Before You Tamp

Tamping doesn't fix any of this. Tamping locks in whatever distribution is already sitting in the basket when you press down, even or not. So fix the distribution first, then tamp.

Break up the clumps

Run a WDT tool through the dose before you touch the tamper. A small, back-and-forth, criss-cross motion through the grounds breaks up clumps and lets fine particles fall down through the gaps instead of sitting bunched near the surface. It takes ten to fifteen seconds. This is the highest-leverage step in the whole process, because it fixes the actual cause instead of a symptom.

Give the water somewhere even to land

A puck screen sits on top of the tamped puck, between the coffee and your machine's shower screen. It spreads incoming water across the full width of the basket before pre-infusion properly starts, instead of letting it hit wherever the shower screen happens to be uneven or wherever your dose happens to sit a little lower. On a 54mm basket especially, where there's less room for error, that first even contact matters. It's 150 micron stainless mesh, sized by portafilter diameter so it sits flush rather than floats, which is the detail that decides whether it does anything at all.

Where tamping actually fits in

Once the grounds are distributed, tamp level and straight down. A calibrated tamper won't fix uneven distribution. No tamper will. What it does is take pressure variation out of the equation, so when a shot goes wrong you know it wasn't because you leaned on it harder than yesterday. That's one less variable to chase when you're troubleshooting.

Practical Takeaway: Your Next Five Shots

Run this sequence for your next five shots and watch what happens to the spray pattern:

1. Dose as usual.
2. Run a WDT tool through the grounds, criss-cross motion, ten to fifteen seconds.
3. Add a puck screen on top, if you're using one.
4. Tamp level, straight down, consistent pressure.
5. Watch the first three seconds of the pour. Two slow, even streams mean distribution's sorted. A spray or a fast, thin jet from one side means there's still a clump somewhere, usually near the wall of the basket.

If the split disappears, you've confirmed it was distribution, not the grinder, the beans, or the machine itself. If it doesn't, check that your basket size actually matches your machine (54mm and 58mm baskets aren't interchangeable) and confirm your grind setting hasn't drifted since you last dialled in. Worth a look too: a basket that's worn or slightly warped can create its own low-resistance path no amount of distribution will fix, though that's a less common cause than dose and distribution on a machine that's otherwise running fine.

None of this needs a new grinder or a different machine. It needs about forty extra seconds before you tamp, spent making sure the grounds give the water nowhere easy to go. For most home Breville setups, sour on one side and bitter on the other starts, and ends, with what happens to the dose before the tamper ever touches it.