Milk Type Experiments

The central question behind these experiments is straightforward: how does the type of milk (or milk alternative) affect the microfoam quality, texture stability, and visual definition achievable in latte art? Rather than relying on preference alone, each milk was steamed under controlled conditions and evaluated against a consistent set of observable criteria.

Experiment Setup & Controls

All trials used the same espresso base — a double shot pulled at roughly 18 g in, 36 g out, over 26–28 seconds — to minimise variation from the coffee side. (For more on how espresso extraction and crema density influence results, see Espresso Base & Crema Analysis.) Steam pressure was held at approximately 1.0–1.2 bar, and each milk type was steamed from a starting temperature of 4 °C (refrigerated) to a target of 60–65 °C, measured with an instant-read thermometer. A 350 mL stainless steel pitcher was used for all pours, filled to 200 mL each time.

Five milk types were tested across multiple sessions:

• Whole cow’s milk (approximately 3.5% fat)
• Skim cow’s milk (approximately 0.1% fat)
• Oat milk (a widely available “barista” formulation, containing added oil and acidity regulators)
• Soy milk (unsweetened, shelf-stable)
• Almond milk (a “barista” formulation with added sunflower lecithin)

Each milk was steamed five times, and the best three pours (judged by consistency of technique) were used for evaluation.

Observations: Foam Structure & Stability

Whole milk produced the most consistently fine, velvety microfoam. The combination of fat content (which contributes to body and mouthfeel) and whey proteins (which act as surfactant proteins, stabilising air bubbles at the liquid–air interface) gave this milk a clear advantage in foam longevity. Poured patterns — primarily rosettes — held their definition for well over 60 seconds before visible degradation.

Skim milk foamed readily and actually produced a slightly greater volume of foam for the same steaming duration, likely because the lower fat content reduces the antifoaming effect that fat globules can exert on protein films. However, the resulting foam was noticeably less dense, with larger and less uniform bubbles. Latte art definition was acceptable but degraded more quickly (visible softening of lines within roughly 30–40 seconds).

Oat milk (barista formulation) performed surprisingly close to whole milk in terms of microfoam density, though the texture had a subtly different viscosity — slightly thinner on the palate, with a tendency to flow faster during the pour. The added oil (typically rapeseed or sunflower) appears to serve a role analogous to dairy fat, helping to stabilise the foam. Art definition was good, though fine details like thin lines in a tulip were marginally less crisp.

Soy milk was the most temperature-sensitive of the group. When steamed above approximately 65 °C, it curdled visibly on contact with the espresso’s acidity — a well-documented reaction between soy protein and chlorogenic acid. At 58–62 °C, the result was a workable microfoam, though somewhat prone to large-bubble formation if air was introduced too aggressively at the start of steaming.

Almond milk produced the thinnest foam with the least stability. Even the barista formulation (with added lecithin as an emulsifier) yielded microfoam that collapsed quickly, making detailed pours difficult. The result suggests that almond milk’s low protein content (roughly 1 g per 100 mL, compared to approximately 3.4 g for whole cow’s milk) is the primary limiting factor.

Ranking by Art Definition

Based on the clarity of a simple rosetta pattern held for 45 seconds after pouring:

1. Whole milk — most defined, most stable
2. Oat milk (barista formulation) — close second, slightly faster flow
3. Skim milk — good initial definition, faster degradation
4. Soy milk — workable within a narrow temperature window
5. Almond milk — limited definition, rapid foam collapse

Takeaways

The data consistently point to protein content and fat content as the two most influential variables in foam quality. Milks (or alternatives) with higher protein tend to form more stable bubble structures, while moderate fat content contributes to the dense, glossy character associated with good microfoam. Temperature sensitivity — particularly relevant for soy — introduces an additional variable worth controlling carefully. For more detail on how steaming temperature interacts with these findings, the Temperature & Steam Pressure Tests page extends these observations. Readers newer to structured latte art practice may also find the Beginner’s Guide a useful starting point for foundational technique before experimenting with different milk types.