Coffee Education8 min read

NCT vs F-Series: Kaffelogic Profiles Compared by the Numbers

We measured all 8 F-Series and 24 NCT profile curves and compared them against 58 logged roasts. NCT profiles heat faster and call first crack about 10 °C higher, but they are not always the shorter roast.

BrewedLate Coffee

#kaffelogic profiles#nct profiles#f-series profiles#kaffelogic nano 7#roast profiles

Most comparisons of NCT and F-Series profiles describe the two families in words: NCT is "short and decisive", F-Series is "gentle and forgiving". We wanted numbers. So we measured the programmed curve in all 8 F-Series and 24 NCT profile files in our roast profile library, then checked them against 58 roasts we logged on those profiles on a Kaffelogic Nano 7e.

The short answer: NCT profiles preheat harder, heat faster early, and place the first-crack marker about 10 °C higher. They are not reliably shorter roasts. What they shorten is the stretch after first crack.

If you want the background on what the packs are for, start with Nucleus Link (NCT) profiles explained. This post is only the measurements.

The curves, side by side

F-WSH 1500-2200 vs NCT Filter E vs NCT Filter D

The programmed curve from each .kpro file: bean temperature above, rate of rise below.

Loading roast curves…

Three things show up straight away. Both NCT curves climb faster in the first two minutes: at 2:00 they are 10 to 15 °C ahead. Filter E then stays about 12 °C ahead for the rest of the roast, while Filter D gives its lead back and from five minutes on tracks the F-Series curve within a degree or two. And the rate-of-rise lines cross: the NCT profiles start higher and fall sooner.

What the profile files say

These are medians across every profile in each family, read from the .kpro files. Ranges are in brackets.

MeasureF-Series (8 profiles)NCT Filter A–E (5)All NCT (24)
Preheat power845 W (790–870)1000 W (970–1040)975 W (600–1160)
Curve reaches 150 °C2:30 (2:23–2:40)1:57 (1:53–2:07)2:07 (1:34–3:54)
Curve reaches 200 °C5:29 (4:53–5:47)4:31 (4:24–5:32)4:51 (3:56–8:33)
Expected first crack207 °C (204–213)217.5 °C (215.5–219.5)217 °C (208–223.5)
Curve reaches expected first crack6:18 (6:04–6:27)6:41 (6:15–8:01)7:00 (5:36–11:30)
Rate of rise at that point7.7 °C/min (6.6–8.2)7.3 °C/min (4.8–7.6)5.3 °C/min (3.9–8.8)

Two patterns matter more than any single row.

F-Series is one family. NCT is many. All eight F-Series curves reach their first-crack marker within 23 seconds of each other. The altitude and process variants change the temperatures, not the shape. The 24 NCT profiles span almost six minutes on the same measure, because Espresso and Omni packs are built for a different job than Filter packs. "NCT" tells you much less about a roast than "F-Series" does, so compare pack by pack.

NCT gets hot sooner but calls first crack later. The Filter packs are about 30 seconds ahead at 150 °C and about a minute ahead at 200 °C. Their first-crack marker sits 10 °C higher, though, so the curve takes slightly longer to get there: 6:41 against 6:18.

First crack is a setting, not a sound

The first-crack temperatures in our logs match the profile files almost exactly: 206.1 °C on F-WSH 1500-2200 (the file says 206), 217.6 °C on NCT Filter E (217.5), 215.7 °C on Filter D (215.5), 218.3 °C on Filter B (218).

That is because the roaster does not detect first crack. Each profile carries an expected first-crack temperature, and the log marks first crack where the measured curve crosses it. The same Ethiopia Sidamo "reached first crack" at 206 °C on one profile and 218 °C on another.

So when a guide says NCT profiles have a higher first-crack temperature, it is describing where the profile's designer put the marker. It also means anything computed from that marker, including development time, depends on the profile. For what first crack actually is in the bean, see first crack in coffee roasting.

What happened in real roasts

Curves are plans. These are medians from our own logged roasts on a Nano 7e: one roaster, six beans, mostly 100–120 g batches.

ProfileRoastsLevelTotal timeTime after first-crack markerDevelopment ratio
F-WSH 1500-2200222.27:2875 s16.3%
F-WSH 1000-170052.57:2480 s17.4%
F-NAT 0-1200113.47:3797 s20.8%
NCT Filter E72.47:0149 s11.7%
NCT Filter B42.27:4655 s11.8%
NCT Filter D82.48:3544 s8.7%

Only Filter E was the shorter roast. Filter B ran 18 seconds longer than F-WSH 1500-2200 and Filter D more than a minute longer. The consistent difference is the last column: NCT filter packs spend 44 to 55 seconds past their marker where F-Series spends 75 to 97.

Same bean, both families

Six beans were roasted on at least one profile from each family. Scores are our own cup ratings out of 5, shown as the median with the number of roasts in brackets.

BeanBest F-Series resultBest NCT result
Ethiopia Sidamo (washed)F-WSH 1500-2200: 2.25 (3)Filter E: 4.5 (3)
Ethiopia Yirgacheffe (washed)F-WSH 1500-2200: 2 (5)Filter E: 4 (3)
Colombia ASOSPAC (washed)F-WSH 1500-2200: 3.25 (2)Filter B: 4.5 (2)
Honduras La Flor (washed)F-WSH 1000-1700: 2.5 (4)Filter D: 3.5 (2)
Kenya Gatugi peaberry (washed)F-WSH 1500-2200: 3.5 (12)Filter D: 4 (1)
Swiss Water decaf BrazilF-NAT 0-1200: 4 (11)Filter D: 3 (1)

NCT came out ahead on four of the six, clearly so on the two washed Ethiopians. Kenya is too close to call on a single NCT roast, and the decaf did better on F-Series. These are small counts from one roaster's palate, and the Kenya and decaf rows reflect where we spent our batches as much as what the profiles can do. The full attempt history for each bean is in the roast logs, for example Ethiopia Sidamo.

How to use this

  • Do not carry a level across families. A 2.2 on F-WSH 1500-2200 dropped at about 214 °C in our roasts; a 2.2 on Filter B dropped at about 223 °C.
  • Do not compare development ratios across families. 8.7% on Filter D and 16.3% on F-WSH 1500-2200 scored the same median 3.5. The marker moved, not the roast.
  • Pick the NCT pack, not the family. Filter E and Filter D differ from each other more than Filter E differs from F-WSH 1500-2200.
  • For washed Ethiopians, try Filter E first. It is the clearest result in our data.

Browse 106 Kaffelogic roast profiles

Search, filter and compare real community profiles, ranked by downloads — no signup needed.

Family
Process
Origin
Brew
Roast
Altitude

24 profiles · tap a profile to open its full page with the roast curve and community outcomes.

Like a profile? Plan the roast, log the curve and get AI next-step suggestions.

Plan this roast in the app
Free for Nano owners

Compare any two profiles yourself

Overlay up to three Kaffelogic profile curves in the library, then plan the batch and log the result in BrewedLate.

Related Reading

Frequently Asked Questions

What is the difference between NCT and F-Series Kaffelogic profiles?
NCT profiles preheat harder (a median 975 W against 845 W for F-Series), reach 150 °C about 30 seconds sooner, and set the first-crack marker around 217 °C instead of 207 °C. F-Series profiles are one tightly grouped family; NCT packs vary widely from one pack to the next.
Are NCT profiles shorter roasts than F-Series?
Not always. In our logged roasts NCT Filter E finished in a median 7:01 against 7:28 for F-WSH 1500-2200, but NCT Filter B took 7:46 and Filter D took 8:35. What NCT filter packs consistently shorten is the time after the first-crack marker: 44 to 55 seconds against 75.
Why is first crack at a different temperature on NCT profiles?
Each profile file carries its own expected first-crack temperature, and roast logs mark first crack where the curve crosses it. F-Series files set it between 204 and 213 °C; NCT Filter packs set it between 215.5 and 219.5 °C. It is a setting in the profile, not something the roaster hears.
Which is better for washed Ethiopian coffee, NCT or F-Series?
On our two washed Ethiopians NCT Filter E scored higher: a median 4.5 out of 5 against 2.25 on Sidamo, and 4 against 2 on Yirgacheffe, compared with F-WSH 1500-2200. That is one roaster and a handful of batches per profile, so treat it as a starting point.
Can I use the same roast level on NCT and F-Series profiles?
No. Levels are relative to each profile's own temperature ladder, and the two families run on different temperature scales. Start from the level the profile recommends and move in 0.1 to 0.2 steps.

Related Articles