dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 378108023.375
num_examples: 1581
download_size: 373552088
dataset_size: 378108023.375
MDCT-1k
Over 1000 audio clips from the Google music captions dataset represented as 512x512 time-frequency images.
The time-frequency images are created from the MDCT coefficients of the 0-12kHz frequency band for 20 second audio clips.
Please see this notebook showing how to load the dataset and convert from the MDCT images back to audio
Most other audio diffusion models operate in the space of the magnitude spectrogram or mel magnitude spectrogram. Since the phase is discarded, this requires the use of a vocoder for audio generation. When operating in the space of the mel-spectrogram, high frequencies are represented with insufficient time resolution, leading to a noticable loss of quality.
Operating in the MDCT space does not require a vocoder, nor does it oversample or undersample any range of frequencies.