Coffee Roast Degree and Agtron Numbers: What They Really Tell You

Roasters talk about "light", "medium" and "dark" because those words are easy, but every roaster's "medium" looks a little different. To repeat a roast from one batch to the next, or compare with someone else's coffee, you need a number. That is what roast-degree measurement is for.

What is being measured?

As green coffee roasts, its colour moves from yellowish to brown and eventually almost black. Roast-degree analysers shine light on ground or whole roasted coffee and measure how much is reflected. Darker coffee reflects less light. The result is expressed as a number on a scale; the best-known scale is the Agtron scale, where a higher number generally means a lighter roast and a lower number a darker one.

See 5% OFF at LeBrew →

Why a number beats your eyes

• Lighting: the same beans look different under daylight, warm bulbs and LEDs.
• Memory: comparing today's batch to last week's by eye is unreliable.
• Communication: telling a customer or a colleague "Agtron around X" is clearer than "medium-light".

Whole bean vs ground readings

Most analysers can read both whole and ground coffee. Ground readings are usually more consistent because the surface is even, and the difference between the whole bean and ground reading can itself tell you something: a large gap can point to uneven roasting, where the outside of the bean is more developed than the inside. Always note which you measured, and keep to one method so your numbers stay comparable.

Colour is not the whole story

Two coffees can show the same colour number and taste different. Development time, airflow, the bean's moisture and density, and how quickly the roast was finished all change flavour independent of colour. That is why roasters record time and temperature curves too, and why other measurements such as moisture, density and water activity are useful companions to a colour reading.

Using roast numbers in practice

1. Set a target for each coffee after you taste a batch you like.
2. Measure every batch the same way and write it down next to the roast log.
3. Look for drift. A creeping change in the number often signals a change in drum temperature, airflow or bean batch.
4. Adjust slowly. Change one variable at a time, then check the number again.

Where tools fit

LeBrew makes a line of coffee analysis instruments that its site lists under the RoastSee name, including a roasting degree analyser, an Agtron and crack analyser, a moisture and density analyser, a water activity analyser and a particle size analyser, plus a temperature analyser. The site describes them as tools for measuring different parts of the roast and grind process. Specifications and prices are not published on the page we reviewed, so check the product pages for what each model measures and which scale it reports before deciding.

Common misreadings

• Comparing numbers from different instruments without calibration notes.
• Assuming a lower number is "better". It is simply darker.
• Reading freshly roasted coffee and coffee that has rested for days and expecting identical numbers. Note the rest time.

Building a simple roast log

A number only becomes useful when it sits next to the rest of the batch record. A basic log needs only a few columns: the date, the coffee and lot, the charge weight, the charge temperature, key times (first crack, end of roast), the final bean temperature, the roast-degree reading and a short tasting note. After ten or twenty batches you will see patterns: which readings you liked, which drifted, and which changes in the roast produced them.

Column	Why record it
Green coffee lot and moisture	Explains differences between lots
Development time	Colour alone can hide under-developed coffee
Roast-degree reading	Objective target to repeat
Rest days before tasting	Flavour changes as coffee degasses
Tasting notes	Connects the numbers to the cup

Sampling and calibration habits

Take a representative sample, not just the beans from the top of the cooling tray, and grind it the same way each time. Let the sample cool and settle before measuring so temperature and trapped gas do not affect the reading. Check the instrument against a reference sample on a schedule, and keep the instrument's surface clean, because oils and chaff can affect the result. If you ever service or replace the device, measure a few known batches before and after so you can relate the new numbers to your historical ones.

Frequently asked questions

Is a lighter roast always more acidic?

Lighter roasts often taste brighter, but acidity in the cup also depends on origin, processing, brewing and water. Roast colour is only one input.

Can I compare my numbers with another roaster's?

Only roughly. Different instruments, calibrations and sample preparation can shift the readings, so compare within your own records unless you have calibrated against the same reference.

Do I need to measure every batch?

Measuring every batch is the habit that builds useful data. If that is too much, measure at least the first batch of each green lot and any batch that tastes different.

Reading trends over time

Plot your readings against date for each coffee. A flat line means your process is stable. A slow slope suggests a drift, perhaps from a dirty roaster, changing room temperature or a new green lot. A sudden jump points to a changed setting or measurement error. Use these patterns to decide when to investigate and when to leave things alone.

Conclusion

Roast degree numbers don't replace your palate, but they give you a repeatable reference. Pair them with a taste notes log and you can reproduce the roasts you love and spot problems earlier.

If you're ready to buy, LeBrew is worth checking — they currently have 5% off with code BR05 live right now.