Audimee vs DiffSinger

Audimee is a voice-to-voice platform for transforming vocals with built-in or custom voice models. Users can adjust melody with the Pitch Editor before conversion, isolate lead vocals from instrumentals and background vocals, and remove reverb, echo, and noise. Its harmony feature can turn one recording into up to five tracks with different voices and harmony lines. Users can train custom voice models with 1–30 minutes of vocal data, combine up to four models with Mixed Voices, and split tracks into two or four layers of vocals, drums, bass, and other sounds. The catalogue contains over 100 voices, with new additions released weekly. Audimee is cloud-based and accessible through a browser and/or desktop app. The Free plan is free and includes 15 minutes of conversions, 11 royalty-free voices, and MP3 downloads; free downloads are for personal use only. Starter is $9 a month, billed $108 yearly, and paid plans allow commercial publication of generated vocals. Audimee says uploaded audio and voice models remain private and inputs and outputs are not used for AI training unless separately agreed in writing.

DiffSinger is a free, open-source singing voice synthesis system built around a shallow diffusion mechanism. The maintained version adapts generated audio to 44.1 kHz and incorporates updated acoustic models plus faster diffusion sampling methods. Variance models and parameters provide control over pitch, energy, breathiness, and other vocal characteristics. The linked MakeDiffSinger project supplies dataset-building pipelines, including recording-slicing and labeling tools. Its guide describes preprocessing data, training models, and running inference on DS files with variance or acoustic models. Installation requires Python 3.10 or later, with PyTorch 2.4.0 or later recommended. ONNX is the deployment format. The repository links OpenUtau and DiffScope as production or deployment projects; DiffScope is an editor powered by DiffSinger and remains under development. Users prepare their own data and assets. The project prohibits generating another person’s voice without consent and includes a utility for removing speaker embeddings from checkpoints before distribution. Support channels include tutorials, GitHub issues and discussions, QQ, and Discord.