We ran all twelve tracks of Second Flight through an AI music analyzer. Then we checked what the score actually means.

On 13 September 2026 we uploaded the twelve released masters of Wings & Prayers: Second Flight — the 48 kHz / 24-bit files the release shipped from — to DropTrack’s Music Analyzer. Every track came back between 94 and 97 out of 100.

That is a strong result inside the tool’s own scale, and the reflex here is not to frame it. It is to ask what the number is made of. This entry publishes the complete result, then separates the parts of the report that anyone can check from the parts nobody outside DropTrack can. The governing rule of this Studio applies to a good number as much as a bad one: explain what you can measure, and be explicit about what you cannot.

The result

#	Track	Score
01	First Wings (Second Flight Version)	95
02	Tailwind (Up, Up, Up) [Second Flight Version]	96
03	The Little Blue Pilot (Second Flight Version)	94
04	Read the Storm (Second Flight Version)	95
05	Little Brave Heart (You Can Do It) [Second Flight Version]	96
06	Higher Than the Clouds (Second Flight Version)	97
07	Wheels Down (Second Flight Version)	94
08	Wingtips and Dew (Second Flight Version)	94
09	Tailwind (Up, Up, Up) [Remix, Second Flight Version]	96
10	The Little Blue Pilot (Sunshine Squad Remix, Second Flight Version)	96
11	The Little Blue Pilot (Night Flight, Second Flight Version)	94
12	Little Brave Heart (We Can Do It) [Second Flight Version]	96

Average 95.25; range 94–97; one track at 97, five at 96, two at 95, four at 94. Source: DropTrack Music Analyzer, “Overall Quality Score”, run 13 September 2026 on the released masters. Per-track PDF reports were saved for eight of the twelve (tracks 01, 04, 05, 07, 08, 09, 11 and 12); the other four scores were recorded from the analyzer’s result pages on the day, and their full reports were not kept.

This is a proprietary AI readiness score reported by DropTrack. It is not an independent review, and it is not a measure of artistic quality.

What the score represents

DropTrack built the analyzer and describes it, in its own words, as “an overall grade based on industry-standard spectral balance and dynamic range” with “AI-powered A&R feedback” alongside (DropTrack, May 2026). The same company sells the next step: “when your track is ready, you can move straight into promotion inside DropTrack by submitting it to record labels, DJs, playlist curators, bloggers, and radio opportunities” (DropTrack, April 2026). Every report we received ends with a “Promote this track” button. None of that makes the tool useless — a readiness check before a pitch is a reasonable thing to sell — but it does make the score a product feature rather than a neutral measurement, and the reader should know who made the ruler.

The scoring formula is not published. Nothing on DropTrack’s site says how a 94 differs from a 97, so nobody outside the company can reproduce the number, and neither can we. That single fact is the spine of this entry.

The Billboard benchmark, for scale

DropTrack ran its own analyzer over the Billboard Hot 100 Top 10 for the chart of 11 July 2026 and published the results: ten recordings scoring between 86 and 93, average 90.4, with three tied at 93 and the lowest at 86 (DropTrack, July 2026). The same post says, in a sentence worth keeping: “A score should be treated as a diagnostic, not a talent grade.”

So the precise statement is this: using DropTrack’s own scoring system, Second Flight’s twelve tracks scored 94–97; in DropTrack’s published test of the Billboard Hot 100 Top 10, those ten recordings scored 86–93. That gives the number scale, and nothing more. Chart position, artistic quality and a proprietary readiness score measure three different things, and the second of those is not measured by anything on this page. We are not saying Second Flight is “better” than those recordings; the tool’s own author says its score cannot say that either.

A score, and a measurement

The left column is a number a company produced from the files with a method it has not disclosed. The right column is a set of measurements anyone with the same files, a BS.1770 meter and a sample counter will get again. Both are real. Only one of them is evidence in the sense this Studio uses the word, and there is no WPAudio Engine equivalent of the left column — the engine does not produce a score, only measurements.

What we can measure

The released masters were made to targets chosen in advance and then checked. The full account is the Second Flight mastering case study; the facts it establishes, each one measured from the same files this entry is about, are these:

• One target for the whole record: −14.0 LUFS integrated under a −1.0 dBTP ceiling. Every released master measures −14.0 LUFS integrated, and no master peaks above −1.1 dBTP.
• Zero samples at full scale in any master. Four of the twelve mixes had arrived with clipped samples (185 in all); none survived into the masters.
• Seven tracks needed the true-peak limiter (02, 04, 07, 08, 10, 11, 12); the other five are gain and a DC blocker, nothing else. No equalisation, no multiband processing, no stereo widening.
• Loudness range preserved to within −0.1 LU on the worst-affected track.
• Delivered at 48 kHz / 24-bit, no resampling.
• Every figure measured twice, from the same files: once by the engine’s own run, once with ffmpeg’s BS.1770 implementation and a direct count of full-scale samples, and published only where the two agreed.

Nothing in that list is a grade. It is what the files are.

DropTrack’s technical panel, beside our measurements

Each report carries an “Audio Technical Analysis” panel. For the eight tracks with saved reports, its fields can be set beside our two-meter measurements of the same files.

#	Track	DropTrack “Loudness”	Measured	DropTrack “True Peak”	Measured
01	First Wings	−17.3 LUFS	−14.0 LUFS	−2.0 dBTP	−1.9 dBTP
04	Read the Storm	−17.7 LUFS	−14.0 LUFS	−2.0 dBTP	−1.1 dBTP
05	Little Brave Heart (You Can Do It)	−17.6 LUFS	−14.0 LUFS	−2.2 dBTP	−2.2 dBTP
07	Wheels Down	−18.1 LUFS	−14.0 LUFS	−1.9 dBTP	−1.1 dBTP
08	Wingtips and Dew	−17.8 LUFS	−14.0 LUFS	−1.4 dBTP	−1.1 dBTP
09	Tailwind (Remix)	−17.2 LUFS	−14.0 LUFS	−0.9 dBTP	−1.4 dBTP
11	The Little Blue Pilot (Night Flight)	−17.6 LUFS	−14.0 LUFS	−1.6 dBTP	−1.1 dBTP
12	Little Brave Heart (We Can Do It)	−17.7 LUFS	−14.0 LUFS	−1.1 dBTP	−1.1 dBTP

“Measured” is the released master’s integrated loudness (ITU-R BS.1770-4) and true peak (4× oversampled), as published in the case study.

What agreed

• Sample rate and bit depth: 48 kHz / 24-bit on every report, as delivered.
• Duration: every report’s running time matches the file to the second.
• Clipping: “None detected” on every report, which agrees with the zero full-scale samples we counted in every master.

The parts a file header and a threshold can answer, the panel answers.

What disagreed

• The field labelled “Loudness” reads between −17.2 and −18.1 LUFS across the eight reports. The same files measure −14.0 LUFS integrated on two independent BS.1770 meters. On every report the “Loudness” figure sits within a decibel of the report’s own “RMS Level”, so it does not appear to represent integrated BS.1770 loudness and behaves more like the adjacent RMS measurement. We cannot see the tool’s method, so we cannot say more than that.
• The field labelled “True Peak” matched the report’s “Peak Level” (a sample-peak figure in dBFS) to the decimal on all eight saved reports. An oversampled true-peak reading is typically at or above the sample peak and is not expected to coincide with it on every file. Where the field disagrees with our measurement it does so in both directions: track 09 is reported at −0.9 dBTP and measures −1.4 on both of our meters; track 04 is reported at −2.0 and measures −1.1.
• “Dynamic Range: 12 dB” is the value on all eight reports, for tracks whose loudness range we measured between 4.9 LU and 10.5 LU. A figure that does not vary across eight different masters is not one we can interpret.

We therefore would not use these three fields as substitutes for a dedicated BS.1770 loudness meter or a true-peak meter. That is not a claim about the overall score — we do not know how much, if anything, the score depends on these fields — only about what the panel’s labels can be taken to mean.

The AI A&R feedback across the reports

Each report closes with five “Recommended Next Steps” from an “AI A&R rep”. Across the eight saved reports, seven suggest adding a bridge section, all eight suggest additional harmonies, and three suggest raising the vocal presence — in the same words, for a singer-songwriter track and an EDM remix alike. The executive summaries do read each song (the night-flight remix is described as “a little blue bird soaring through the night sky”, which it is), but the prescriptions read as a template.

What 95.25 can legitimately tell us

That a commercial readiness tool, on whatever it measures, places all twelve released masters near the top of its scale — and in a higher band than the ten chart recordings it chose to test itself. That is a real result about DropTrack’s output, and it is the reason this entry exists.

What it cannot tell us

Why the number is 95 rather than 93 or 97. How much of it is the mastering, the mix, the arrangement or the song. Whether a 97 sounds different from a 94. Whether any of it would survive a different tool. And it cannot tell a listener anything about artistic quality, because its author says it is not built to.

Its technical panel, where we could check it, is a rougher instrument than its labels imply. So the score goes into the record as what it is: a diagnostic from a tool we do not control, published in full, beside measurements we do.

Why this is here and not on the artist’s site

littlebluepilot.com is the pilot’s world: songs, the storybook, the films. Its own editorial rule is that the site makes no claims about the music’s reception, and a score is one. The Studio is where the bird has a spreadsheet.

Related

• Mastering Second Flight — one target, every number measured twice: the measurements this entry checks against, and how they were made.
• Mastering Wings & Prayers: the first record, with the audio to A/B for yourself.
• Publish only what you can measure: the decision this entry is an application of.