First Roast Guide
If you have already installed the app, connected your roaster, and are ready to start your first batch, this guide is for you.
It does not throw you straight into complex automation settings. Instead, it gives you a safer beginner path: copy a reference curve first, understand how one roast moves from start to finish, and only then move on to automation or Replayer.
Who this guide is for
This guide is for users who:
- Have already installed HiBean and connected their device.
- Can see live temperature data updating normally.
- Are about to start their first roast but are not sure what to learn first between manual roasting, automation, profiles, and curves.
- Want to finish a roast they can review later instead of starting with advanced features.
If you have not finished device setup yet, go back to Connect First Device.
One-sentence recommendation
For your first roast, do not start with automation.
A better learning path is:
- In manual mode, find a same-device reference curve first.
- Copy that curve for a few roasts.
- After you understand the basic roast rhythm, use automation or Replayer to simplify your workflow.
The key benefit is that you first learn why the roast moves the way it does, not just where the buttons are.
What to prepare before your first roast
Before you begin, make sure all of the following are true:
- The device connection is stable: live data keeps updating without frequent disconnects.
- Probe data looks normal: bean temperature (BT) is not obviously wrong and the trend makes sense.
- Your setup is safe: ventilation, cooling, bean drop, and power-off handling are all ready.
- Your batch size is confirmed: for the first roast, stay close to a normal batch size for your machine.
- Your preheat plan is decided: you know what state the machine should be in before charge.
- You know how to end the roast: including when to drop, how to stop heating, and how to enter cooling.
If you want to understand the preparation page first, read Roasting Preparation: Pre-roast Checklist.
How to choose your first reference curve
For a first roast, the goal is not to find the "perfect" curve. The goal is to find a curve that is good for copying.
Use this priority:
- A community profile from the same device
- A successful roast of your own saved as a reference curve
- A same-device profile with a similar batch size and roast target
If you cannot find the exact same coffee bean, the priority is still:
same device > same bean
That is because different roasters have very different heat response, airflow behavior, and probe placement. For beginners, device match usually matters more than bean match.
When choosing your first reference curve, also look for these signals:
- Pick a profile that gives you a visible reference curve in manual mode.
- If a profile mainly contains automation actions or replay data but is not useful as a visual reference, skip it for now.
- On the preparation page, if you can clearly see a reference curve, it is usually a better candidate for copying practice.
Why 022 is not the recommended starting point
022 is not the right beginner entry point here.
That does not mean it is bad. It means it serves a different purpose: it is part of an automation approach, not a beginner-friendly reference curve meant for copying. If a user has not yet built a basic sense of roast flow, starting there usually increases cognitive load instead of reducing it.
What to watch during your first manual roast
On your first copied roast, you do not need to act like you are perfectly recreating every point on the graph. What you are really learning is: how one roast moves from beginning to end.
For now, focus on just 3 things.
1. Bean temperature (BT) trend
First, look at whether you are generally keeping up with the reference or falling behind.
- If your live BT stays clearly below the reference curve for a while, the roast is running slow.
- If your live BT stays clearly above the reference curve for a while, the roast is running fast.
On a first roast, you do not need every second to overlap perfectly. Start with the big trend.
2. RoR trend
RoR is not something a beginner needs to "play with" right away.
At this stage, what matters more is noticing:
- Is it pushing too hard overall?
- Is it dropping too fast?
- After you change heat or airflow, does the trend move back toward something more stable?
You do not need to design multiple RoR phases on your first roast, and you do not need frequent control changes just to look advanced.
3. Key milestones
For your first roast, remember these milestones:
- Turning point after charge
- Whether the middle phase is clearly too fast or too slow
- Rough timing and BT at first crack
- Drop timing
If those key points are roughly aligned with the reference curve, the roast is already highly valuable as a learning run.
What not to force on your first roast
For a first roast, do not force the following:
- You do not need every moment to match the reference perfectly.
- You do not need to obsess over manually "holding at temp" right away.
- You do not need complicated heat changes from the start.
- You do not need to react to every small wiggle on the graph.
For a beginner, building a sense of overall roast rhythm is more important than correcting every tiny movement.
How to review the roast afterward
After the batch is done, do not only ask whether the final drop temperature was correct. Review it by phase:
- Was the early phase generally too fast or too slow?
- Did the middle phase push too hard or flatten too much?
- Around first crack, were you earlier or later than the reference?
- Did your key control changes actually move the trend back toward the target?
When you review a roast, record these items first:
- Preheat condition
- Green bean weight
- First crack time
- Drop time
- The 1 or 2 actions that mattered most
If you can explain "why I ran fast here" or "why I failed to keep up here," that roast already did its job as a learning batch.
Three common questions
Can I reproduce a roast with BT?
You can use BT to reproduce the trend and timing of a roast, but you should not read it as "same BT means exactly the same result in every situation."
On the same machine, in a similar environment, with a similar batch size, BT is highly useful. But if the device, probe position, or environment changes, the absolute numbers may no longer transfer directly.
So the better way to think about it is: BT helps you reproduce the general path of the roast, not create a universal answer across machines and environments.
Should I actively chase different RoRs from the beginning?
No.
In the first stage, your main job is to build a basic sense of roast flow. Learn to read the trend, follow the trend, and understand the key milestones first. Only after that should you start shaping more intentional RoR structures.
Is there a direct answer in HiBean similar to V1 022?
For beginners, the first recommendation is not automation. It is copying a reference curve.
If your goal is to finish your first roast quickly while also understanding how that roast actually moved, the better choice is to find a same-device community profile in the app and copy it in manual mode for a few batches. Once the overall process feels familiar, automation or Replayer will make much more sense.
What to learn next
After you have completed 2 or 3 copied roasts, continue in this order:
That sequence matches the learning path: understand the flow first, then control, then profiles, and only then automation.
Device differences still matter
This guide describes the general roasting workflow, not the details of each machine.
Please remember this:
Different machines have different preheat behavior, heat response, airflow effect, and probe placement.
Because of that, the same BT number or the same control rhythm may not produce the same result on another machine.
If you want machine-specific notes, go to Hardware Support and open the page for your device.