What the result tells you
Three separate readings, deliberately not blended into one verdict. The harmonic one says whether the two keys are a move the transition table endorses, and which kind. The tempo one says how far apart the BPMs are as a percentage, and whether that is inside the margin a crossfade normally survives. The pitch one says what happens to the incoming track's key once you have pulled it onto the other one's tempo.
They are separate because they disagree often, and a single traffic light would have to pick one to lie about. Two tracks in perfect keys can be nine percent apart. Two tracks at identical BPMs can be in keys that fight. Both of those are useful things to know and neither is 'compatible: no'.
Pitch and key lock
On a deck with key lock off, tempo and pitch are one knob. Speed a record up and everything in it rises, including its key. That is not a bug to be corrected — it is what vinyl did and what a lot of mixes are built on — but it means the key you matched on paper is not the key that plays.
A semitone is a frequency ratio of about 5.95%, which is why a track pulled six percent lands almost exactly one semitone away and one pulled three percent lands between two keys, sounding slightly out of tune against everything. The tool reports both: how many semitones the move is worth, and which key that leaves you in — flagged when the answer falls between keys rather than on one.
With key lock on, none of that happens: the deck holds the pitch and only the tempo moves. Turn it on in the tool and the key reading says so instead of pretending the move still shifts it.
Half-time and double-time
87 into 174 is not a 100% tempo jump. It is one beat against two, and it is one of the most reliable transitions in the book — the whole relationship between drum and bass and its halftime intros is built on it. A checker that reported it as impossible would be wrong in exactly the sets where it matters most.
So the tempo reading matches half and double before it measures anything, and then measures the gap against whichever relationship is closest. 86 into 174 comes back as double-time with a 1.2% gap, because that is what your ears and your pitch fader are both dealing with.
Your fader's range is not the same question
Two different limits get confused constantly. A ±8% pitch fader is a fact about your equipment: past that the deck simply cannot go. The seven percent the tool mentions is a musical judgement about what a crossfade survives — past it the two tracks drift audibly during the blend even though the fader reached comfortably.
A 7.5% move clears a ±8 fader and fails the margin. That is not a contradiction, it is two answers to two questions, and the tool gives both rather than averaging them into a number that answers neither.
When to override what it says
A verdict about two tracks is a verdict about two tracks, and you are mixing a night. A move the table does not endorse can be exactly right at the moment you make it — after a break, over a drop, into a track the room already knows. The checker cannot hear any of that, and the tag it is reading may be wrong in the first place: key detection disagrees with itself often enough that a code nobody verified is a strong guess rather than a fact.
Where it earns its place is the case you cannot hear in advance: planning at home, with two tracks you have not tried together, deciding whether the transition is worth building a section around. That is also why the third reading exists — a mix that works in isolation can still be the wrong move for where the set has got to, and only the whole tracklist can answer that.