*2.6. Music Test*

We designed a music trial composed of 4 tests with increasing difficulty. The first 3 tests assessed emotional perception through music. In each of these tests, 6 melodies were played on piano in major (happy) and in minor (sad) modes, representing a total of 12 musical samples of 25 s each. The melodies were unknown to the general public in order to avoid cultural references. The melody line was accompanied by 1–4 note chords. Stimuli were equally tempered MIDI piano notes. All samples were recorded with a 44.1 kHz sampling rate at 16-bit depth. The participant could listen to these samples in a free order and as many times as desired. The subject was asked to categorize these samples as happy or sad with a forced two-choice task. No feedback was given. In the first test (easiest), in addition to the mode difference, happy samples were played faster than sad excerpts with a large difference in tempo (vivace, 140 beats/min for happy versus andante,

80 beats/min for sad). In the second test (intermediate), there was only a small difference in tempo (moderato, 100 versus 90 beats/min). In the third test (most difficult), the tempo was identical (moderato, 90 beats/min) and only the mode difference could allow the distinction. Test 4 evaluated the dissonance perception. Ten melody samples, with (*n* = 5) or without (*n* = 5) dissonance, were presented and the patient had to categorize them as "dissonant" or "harmonious". Melodies had the same characteristics as in the first 3 tests and were played with a moderato tempo (100 beats/min.). All tests were finally marked out of 10. The test interface was a laptop computer screen (Powerpoint 2010, Microsoft Inc. Redmond, VI, USA) where the patient could click on the musical sample to listen and to drag-and-drop the file into the proposed categories represented by happy and sad emojis. Samples were presented on 2 frontal loudspeakers (Sony, SRS-Z510, Tokyo, Japan) at a comfortable level judged by the patient. All tests were conducted in CI-only mode. In patients with residual hearing, the hearing aid was deactivated, and a sound reduction ear plug was placed in the ear. The patient used the interface independently and was only assisted by the investigator for technical issues.

**Table 2. Musical Questionnaire Part 1: Musical Habits.** Numbers indicate the number of choices among proposed responses and the number of positive responses (*n* = 21). Propositions for type of music were not exclusive. For the first question, the numbers indicate mean ± standard error of mean of Likert score [range]. MPIS (*n* = 6) and Crystalis XDP (*n* = 15) refer to the usual strategies used by the patients. HL: hearing loss, CI: cochlear implant.


At the end, an auto questionnaire allowed the participant to rate the clarity of sound, the enjoyment of the melody, and the ease of each test (Likert scale 1–10), and to answer to the question, "which program did you prefer?" in a blinded manner (program 1 or 2, at the end of the second session).
