CJ Carr

Pedro Sarmento, Adarsh Kumar, Yu-Hua Chen et al.

Recently, symbolic music generation with deep learning techniques has witnessed steady improvements. Most works on this topic focus on MIDI representations, but less attention has been paid to symbolic music generation using guitar tablatures (tabs) which can be used to encode multiple instruments. Tabs include information on expressive techniques and fingerings for fretted string instruments in addition to rhythm and pitch. In this work, we use the DadaGP dataset for guitar tab music generation, a corpus of over 26k songs in GuitarPro and token formats. We introduce methods to condition a Transformer-XL deep learning model to generate guitar tabs (GTR-CTRL) based on desired instrumentation (inst-CTRL) and genre (genre-CTRL). Special control tokens are appended at the beginning of each song in the training corpus. We assess the performance of the model with and without conditioning. We propose instrument presence metrics to assess the inst-CTRL model's response to a given instrumentation prompt. We trained a BERT model for downstream genre classification and used it to assess the results obtained with the genre-CTRL model. Statistical analyses evidence significant differences between the conditioned and unconditioned models. Overall, results indicate that the GTR-CTRL methods provide more flexibility and control for guitar-focused symbolic music generation than an unconditioned model.

Zach Evans, Julian D. Parker, CJ Carr et al.

Open generative models are vitally important for the community, allowing for fine-tunes and serving as baselines when presenting new models. However, most current text-to-audio models are private and not accessible for artists and researchers to build upon. Here we describe the architecture and training process of a new open-weights text-to-audio model trained with Creative Commons data. Our evaluation shows that the model's performance is competitive with the state-of-the-art across various metrics. Notably, the reported FDopenl3 results (measuring the realism of the generations) showcase its potential for high-quality stereo sound synthesis at 44.1kHz.

Jackson Loth, Pedro Sarmento, CJ Carr et al.

Recent work in the field of symbolic music generation has shown value in using a tokenization based on the GuitarPro format, a symbolic representation supporting guitar expressive attributes, as an input and output representation. We extend this work by fine-tuning a pre-trained Transformer model on ProgGP, a custom dataset of 173 progressive metal songs, for the purposes of creating compositions from that genre through a human-AI partnership. Our model is able to generate multiple guitar, bass guitar, drums, piano and orchestral parts. We examine the validity of the generated music using a mixed methods approach by combining quantitative analyses following a computational musicology paradigm and qualitative analyses following a practice-based research paradigm. Finally, we demonstrate the value of the model by using it as a tool to create a progressive metal song, fully produced and mixed by a human metal producer based on AI-generated music.

Julian D. Parker, Zach Evans, CJ Carr et al.

Latent representations are at the heart of the majority of modern generative models. In the audio domain they are typically produced by a neural-audio-codec autoencoder. In this work we introduce SAME (Semantically-Aligned Music autoEncoder), an autoencoder for stereo music and general audio that reaches a 4096$\times$ temporal compression ratio while maintaining reconstruction quality and downstream generative performance. We achieve this by combining a tranformer-based backbone with set of semantic regularisation approaches, phase-aware reconstruction losses and improved discriminator designs. The architecture delivers substantial computational cost benefits, through both its high compression ratio and its reliance on well-optimised transformer primitives. Two variants (a large SAME-L and a CPU-deployable SAME-S) are released in open-weights form.

Zach Evans, Julian D. Parker, Matthew Rice et al.

Stable Audio 3 is a family of fast latent diffusion models (small, medium, large) for variable-length audio generation and editing. Since our models can generate several minutes of audio, variable-length generations are key to avoid the cost of producing full-length generations for short sounds. We also support inpainting, enabling targeted audio editing and the continuation of short recordings. Our latent diffusion models operate on top of a novel semantic-acoustic autoencoder that projects audio into a compact latent space, enabling efficient diffusion-based generation while preserving audio fidelity and encouraging semantic structure in the latent. Finally, we run adversarial post-training to both accelerate inference and improve generation quality, reducing the number of inference steps while improving fidelity and prompt adherence. Stable Audio 3 models are trained on licensed and Creative Commons data to generate music and sounds in less than a 2s on an H200 GPU and less than a few seconds on a MacBook Pro M4. We release the weights of small and medium, that can run on consumer-grade hardware, together with their training and inference pipeline.

Zachary Novack, Zack Zukowski, CJ Carr et al.

Generative audio requires fine-grained controllable outputs, yet most existing methods require model retraining on specific controls or inference-time controls (\textit{e.g.}, guidance) that can also be computationally demanding. By examining the bottlenecks of existing guidance-based controls, in particular their high cost-per-step due to decoder backpropagation, we introduce a guidance-based approach through selective TFG and Latent-Control Heads (LatCHs), which enables controlling latent audio diffusion models with low computational overhead. LatCHs operate directly in latent space, avoiding the expensive decoder step, and requiring minimal training resources (7M parameters and $\approx$ 4 hours of training). Experiments with Stable Audio Open demonstrate effective control over intensity, pitch, and beats (and a combination of those) while maintaining generation quality. Our method balances precision and audio fidelity with far lower computational costs than standard end-to-end guidance. Demo examples can be found at https://zacharynovack.github.io/latch/latch.html.

Zach Evans, CJ Carr, Josiah Taylor et al.

Generating long-form 44.1kHz stereo audio from text prompts can be computationally demanding. Further, most previous works do not tackle that music and sound effects naturally vary in their duration. Our research focuses on the efficient generation of long-form, variable-length stereo music and sounds at 44.1kHz using text prompts with a generative model. Stable Audio is based on latent diffusion, with its latent defined by a fully-convolutional variational autoencoder. It is conditioned on text prompts as well as timing embeddings, allowing for fine control over both the content and length of the generated music and sounds. Stable Audio is capable of rendering stereo signals of up to 95 sec at 44.1kHz in 8 sec on an A100 GPU. Despite its compute efficiency and fast inference, it is one of the best in two public text-to-music and -audio benchmarks and, differently from state-of-the-art models, can generate music with structure and stereo sounds.

Zach Evans, Julian D. Parker, CJ Carr et al.

Audio-based generative models for music have seen great strides recently, but so far have not managed to produce full-length music tracks with coherent musical structure from text prompts. We show that by training a generative model on long temporal contexts it is possible to produce long-form music of up to 4m45s. Our model consists of a diffusion-transformer operating on a highly downsampled continuous latent representation (latent rate of 21.5Hz). It obtains state-of-the-art generations according to metrics on audio quality and prompt alignment, and subjective tests reveal that it produces full-length music with coherent structure.

Julian D Parker, Anton Smirnov, Jordi Pons et al.

The tokenization of speech with neural audio codec models is a vital part of modern AI pipelines for the generation or understanding of speech, alone or in a multimodal context. Traditionally such tokenization models have concentrated on low parameter-count architectures using only components with strong inductive biases. In this work we show that by scaling a transformer architecture with large parameter count to this problem, and applying a flexible Finite Scalar Quantization (FSQ) based bottleneck, it is possible to reach state-of-the-art speech quality at extremely low bit-rates of $400$ or $700$ bits-per-second. The trained models strongly out-perform existing baselines in both objective and subjective tests.

Zachary Novack, Zach Evans, Zack Zukowski et al.

Text-to-audio systems, while increasingly performant, are slow at inference time, thus making their latency unpractical for many creative applications. We present Adversarial Relativistic-Contrastive (ARC) post-training, the first adversarial acceleration algorithm for diffusion/flow models not based on distillation. While past adversarial post-training methods have struggled to compare against their expensive distillation counterparts, ARC post-training is a simple procedure that (1) extends a recent relativistic adversarial formulation to diffusion/flow post-training and (2) combines it with a novel contrastive discriminator objective to encourage better prompt adherence. We pair ARC post-training with a number optimizations to Stable Audio Open and build a model capable of generating $\approx$12s of 44.1kHz stereo audio in $\approx$75ms on an H100, and $\approx$7s on a mobile edge-device, the fastest text-to-audio model to our knowledge.

Pedro Sarmento, Adarsh Kumar, CJ Carr et al.

Originating in the Renaissance and burgeoning in the digital era, tablatures are a commonly used music notation system which provides explicit representations of instrument fingerings rather than pitches. GuitarPro has established itself as a widely used tablature format and software enabling musicians to edit and share songs for musical practice, learning, and composition. In this work, we present DadaGP, a new symbolic music dataset comprising 26,181 song scores in the GuitarPro format covering 739 musical genres, along with an accompanying tokenized format well-suited for generative sequence models such as the Transformer. The tokenized format is inspired by event-based MIDI encodings, often used in symbolic music generation models. The dataset is released with an encoder/decoder which converts GuitarPro files to tokens and back. We present results of a use case in which DadaGP is used to train a Transformer-based model to generate new songs in GuitarPro format. We discuss other relevant use cases for the dataset (guitar-bass transcription, music style transfer and artist/genre classification) as well as ethical implications. DadaGP opens up the possibility to train GuitarPro score generators, fine-tune models on custom data, create new styles of music, AI-powered songwriting apps, and human-AI improvisation.

Zack Zukowski, CJ Carr

We use a modified SampleRNN architecture to generate music in modern genres such as black metal and math rock. Unlike MIDI and symbolic models, SampleRNN generates raw audio in the time domain. This requirement becomes increasingly important in modern music styles where timbre and space are used compositionally. Long developmental compositions with rapid transitions between sections are possible by increasing the depth of the network beyond the number used for speech datasets. We are delighted by the unique characteristic artifacts of neural synthesis.

CJ Carr, Zack Zukowski

This early example of neural synthesis is a proof-of-concept for how machine learning can drive new types of music software. Creating music can be as simple as specifying a set of music influences on which a model trains. We demonstrate a method for generating albums that imitate bands in experimental music genres previously unrealized by traditional synthesis techniques (e.g. additive, subtractive, FM, granular, concatenative). Raw audio is generated autoregressively in the time-domain using an unconditional SampleRNN. We create six albums this way. Artwork and song titles are also generated using materials from the original artists' back catalog as training data. We try a fully-automated method and a human-curated method. We discuss its potential for machine-assisted production.