“Sing Mirror” is a karaoke practice app that lets you visually check the pitch of your singing while you practice. ▼ Import Songs in 3 Ways ・Select an audio file from the Files app ・Choose a song from your iPhone music library ・Record a song using your iPhone microphone Even if you do not have an audio file, you can record music played from speakers or another device using your iPhone microphone and use it for practice. ▼ Analyze Vocals and Pitch with an Inference Model Using an on-device inference model, Uta Mirror separates vocals and accompaniment from imported songs and analyzes the vocal pitch. The analyzed pitch is displayed as blue pitch bars and an orange pitch line. ▼ View Your Own Pitch in Real Time When you sing along with the song, your pitch detected from the microphone is displayed as a red line. Because the original vocal pitch and your own singing pitch are shown on the same graph, you can visually check where your pitch shifts and how your voice moves. ▼ Adjust Volume for Your Practice Style You can individually adjust the volume of the original song, vocals, and accompaniment. For example, you can start by practicing together with the original vocals, then gradually lower the vocal volume as you become more comfortable. ▼ Change the Key You can change the key of each song in semitone steps. Songs that are too high or too low can be adjusted to better match your vocal range. ▼ Record and Review Your Singing You can record your singing during practice and play it back together with the song. Your pitch during recording is also saved, so you can review both your voice and pitch movement later. By comparing the original vocals with your own singing, you can repeatedly practice difficult phrases and pitch changes at your own pace. ※ Model data must be downloaded the first time the inference model is used. ※ Analysis time may vary depending on the length of the song and the performance of your device. ※ The sound quality of recorded songs may vary depending on the playback environment, surrounding noise, and distance from the microphone. ※ Some audio sources, such as DRM-protected songs, may not be available for import.