Speed Analysis

Many tapes in circulation play a little fast or slow, because the deck that recorded them and the one that played them back did not turn at the same speed. Speed? (Ctrl+P) estimates the error for every show and puts the result in the Speed column: "OK?" when nothing points to an error, "1.8 % slow?" or "2.5 % fast?" when something does. The question mark is deliberate: the estimate is a strong hint, not a proof, and the report says how much to trust it.

Click a result to open its report. At the top, a dial shows the speed error the verdict gives, from -3 % to +3 %: green within 0.3 %, yellow up to 1 %. Above, an FM broadcast whose 19 kHz pilot is exactly where it should be: the needle stands at 0. When nothing can be measured, there is no dial.

The Measurements card combines three of them:

• 19 kHz FM pilot: FM stereo stations transmit a tone at exactly 19 000 Hz. When it is on the recording, it measures the speed to the hundredth of a percent.
• Tuning against A = 440 Hz: bands usually tune to A = 440 Hz, so a whole recording sharp or flat by the same amount points to a speed error. It is blind to errors of a whole semitone (5.9 %), and bands that tune down or by ear (common in the early 70s), or old pianos and organs, can fool it.
• Mains hum: hum from the mains (50 Hz in Europe, 60 Hz in America) recorded with the music moves with the tape speed. It may have come in at a later copy, so it only measures the stages after it: a hint, not a proof.

The report of a show also lists every track with its own result and figures, each with a See more button. The Reasoning then gives the correction in the form your editor needs ("speed it up by 1.8 %: play at 101.8 %, +31 cents"), reminds you to check it by ear before applying it, and lists the other possibilities when the error could be larger than a semitone ("it could also be 7.3 % slow or 4.0 % fast: compare by ear with other recordings of the tour"). When nothing can be measured, the result says so rather than guessing.