# Ruth: carry the note, not the barline

The first new solo remains preserved at `experimental/ruth-solo-writing-01`.
This follow-up keeps its complete melody and the original accepted guitar. It
tests one specific writing/articulation decision, not another instrument model,
a harmonic correction, a new chorus, or a claim to have matched Pat Metheny.

## What the actual score does

At soloBeat 59.50 the first draft plays C4 as a common tone into F minor. It
ends its written gate at 59.90, then picks the same C4 again at beat 60.00. The
gap between the written gates is 44.672 ms. Both voices additionally have the
original model's 160 ms release tail, so an empty written gate is not silence.
That is a real repeated attack, not a continuously held note. Repetition can be
musical; the point here is to compare it deliberately rather than assume the
phrase label made the line continuous.

The alternative carries the first C4 through beat 60.40 and removes only the
second excitation. The existing Eb4 at 60.50 still begins the rise. Every other
note, onset, touch, string choice, bass event, shared clock and harmonic change
stays fixed. There are 94 excited voices instead of 95; this is a tie between two
same-pitch manuscript notes, not a strategy of making the solo easier by cutting
material. The whole 32-bar chorus is rendered, not just an isolated oscillator.

## Musical evidence and limits

A bounded independent score audit found no defensible harmony/counterpoint
mistake requiring correction: the chromatic neighbours resolve, and the stated
bass replies avoid written guitar gates. The Db4-to-Bb4 sixth before the crest
is conspicuous but intentional enough that changing it without listening would
be arbitrary. These are score findings, not proof that the solo feels human.

Metheny discusses developing an idea to its conclusion and avoiding excessively
separate attacks in [his phrasing interview](https://www.patmetheny.com/qa/questionView.cfm?queID=233).
He also stresses the relation between note placement and the rhythm section in
[his time-feel answer](https://www.patmetheny.com/qa/questionView.cfm?queID=12812).
Our inference is to compare this particular common-tone seam, not turn those
comments into a formula or a numerical soul score. No recorded solo is copied.

The frozen study-17 renderer is reused directly, with its original guitar model,
excitation, release, highpass, pan and source gain. Score verification checks the
exact one-tie delta independently before the full render. Every published Ruth
source, recording, URL and the first writing attempt remains untouched. No audio
playback, samples, paid reviewers or browser automation are needed for this
local candidate. Export and any separate publication still require actual PCM,
level, codec, preservation and native-player checks. Musical judgment remains
unverified.