Grind, dose, yield, time
How to dial in espresso at home
Dialing in means finding a grind, dose and yield that taste good for this coffee on this machine today.
This guide assumes a basic kitchen scale that reads to 0.1 g and a timer. You do not need a smart scale or flow profiling. It covers the controls, a symptom table and a logging habit. It does not repeat the ratio conventions in the espresso ratio guide or the strength and extraction maths in the extraction guide.
The four controls
- Dose. Dry coffee in the basket, in grams. Fix it. Use the same number every shot, and weigh it to 0.1 g.
- Yield. Grams of espresso in the cup. Put the cup on the scale and stop at your target. Yield is the control that sets strength.
- Time. Seconds from pump start to stop. Time is a clue, not a control. You do not set it. It results from grind, dose and yield.
- Grind. The only control you will move most of the time. Finer slows the flow and raises extraction. Coarser speeds it up and lowers it.
A starting shot such as 18 g in, 36 g out in about 30 seconds is CupTune's starting recipe. It is a place to begin, not a target.
One change at a time
Change one control, pull the next shot, taste, and write it down. If you move grind and yield together, you cannot tell which one fixed it. This is why the sequence is:
- Fix dose.
- Fix yield.
- Move grind until time and taste agree.
- Only then change yield for strength or dose for a different basket fit.
Make grind moves small. One or two steps on a stepped grinder, or a quarter turn on a stepless one. Purge a few grams of coffee after each change, since the grinder holds some of the old setting.
Symptom table
These are most-likely causes, not diagnoses. Taste words overlap, and two symptoms that point opposite ways mean the picture is unclear, so change nothing and pull another shot.
| Symptom | Most likely cause | One next move |
|---|---|---|
| Sour, sharp, salty | Under-extraction, often a loose bed | Grind finer one step |
| Bitter, dry finish | Over-extraction or too long | Grind coarser one step |
| Harsh, astringent | Over-extraction or channeling | Grind coarser one step, then check puck prep |
| Weak, watery | Yield too high for the dose | Stop 3 g earlier |
| Thin, no body | Too fast a flow or high yield | Grind finer one step |
| Gushing, well under 22 s | Grind too coarse or dose too low | Grind finer two steps |
| Choking, well over 35 s, dripping | Grind too fine | Grind coarser two steps |
| Spurts, pale streaks, uneven flow | Channeling | Redistribute grounds evenly and tamp level |
The time bands are authored rules of thumb. The SCA's espresso article reports about 25 to 30 seconds as current practice, which is a description of habits, not a rule.
Signs of channeling
Water finds a path of least resistance through the puck. Look for a bottomless-portafilter stream that sprays sideways, darker and paler zones in the flow, or a puck that is wet and cratered after the shot. If you see these, check distribution and a level tamp before you touch the grind. A grind change cannot fix an uneven puck.
How bean age changes a dial-in
Roasted coffee holds carbon dioxide that is released over time. Smrke and colleagues (2018) measured this degassing by weight loss. Their abstract describes gas that stays trapped in the bean and is released later, during storage, grinding or extraction. That the gas affects extraction and crema is CupTune's inference, not a result we found in that paper, and they did not test shot times against days off roast.
Here is the part CupTune is authoring, not citing. Very fresh beans tend to give faster, more variable shots, and a grind that worked on day four may need moving by day twelve. Bubbling gas in the bed is the likely reason, but this has not been verified in a controlled espresso study that we found. So treat each new bag, and each week of the same bag, as a fresh start. Pull one shot at your last setting, and expect to move grind by a step. Beans two to four weeks off roast are usually calmer to dial. That window is a preference and varies with roast level.
Logging shots so the next change is informed
A log needs six fields: date, beans and days since roast, dose, yield, time, grind setting. Add one line of taste in plain words, and the single change you made last. Compare shots only when dose and yield match.
- Write down the setting before you pull the shot, not after.
- Note the changes you did not make. "No change, same time" is useful data.
- Mark the shot you liked. That row becomes your starting point for the next bag.
- Re-read the last three rows before you change anything.
Next steps
Try the CupTune dial-in tool, which takes your symptoms and proposes one change, and say when they conflict. Then record each shot in your brew journal.
Evidence and limits
Only the Smrke degassing paper and the SCA's description of current practice come from published sources, and neither tests a home dial-in. Every row of the symptom table, the step sizes and the time bands are authored rules of thumb. Machines, baskets and grinders differ widely. If a change makes the shot worse, undo it.