Introduction
The Royal Cities Infinite Pianos – 2024 (SAO Finetune) is a sophisticated model designed specifically for generating piano samples. This tool can create an endless array of piano compositions optimized for music production workflows. Every piece of output is synchronized to the BPM and locked to any note within the 12-tone chromatic scale, whether in major or minor keys. Let’s dive into how you can utilize this model effectively!
Model Features
- Multiple Types of Stem Generation: The model can generate Chord Progressions alone, Melodies solely, or a combination of both.
- Dynamic FX Chain: Use simple text prompts to control effects such as reverb and tremolo.
- Tonal Versatility: Generate piano stems in any key across the 12-tone chromatic scale.
- Simplified Scale Notation: Designed to use sharps only for scales.
Training Methodology
This model was specifically trained to understand different structures of piano samples, including:
- Chord Progression Only Samples
- Chord Progression with Melody Samples
- Melody Only Samples
The training process enabled the model to distinguish various musical motifs, allowing it to generate specific types of music samples based on the input commands. Think of it like a chef who, after examining different recipes, can whip up a fresh meal based on a general idea of what the dish should be.
Usage Guide
To harness the full capabilities of the model, simply follow the steps below:
- Ensure you have the correct .ckpt file and config .json in their respective folders.
- Use a supported interface such as the improved Stable Audio Gradio to work with the model efficiently. This version includes features like:
- Dynamic Model Loading: Switch between models seamlessly.
- Random Prompt Button: Generate random prompts with one click.
- BPM & Bar Selector: Auto-fill prompts with the required BPM/Bar info.
- Automatic Sample to MIDI Converter: Convert audio to .MID format automatically.
- Automatic Sample Trimming: Trim generated samples for easier DAW import.
You can find a direct link to the GitHub fork here. If you wish to use the original Stable Audio Github, you can follow this link.
Prompt Structure
To achieve optimal results, please ensure that your prompts follow this format:
[Piano Type], [Modifier][Chord Progression], [Melody Type], [Key], [FX], [BPM], [Bar Count]
Example prompts could include:
- Chord Progression Only: “Grand Piano, slow chord progression only, E minor, Medium Reverb, 110BPM, 8 bars”
- Chord Progression with Melody: “Medium E. Piano, low chord progression with top catchy melody, B minor, High Spacey Reverb, 128BPM, 8 bars”
- Melody Only: “Soft E. Piano, alternating top arp melody only, F minor, High Reverb, 128BPM, 8 bars”
Technical Specifications
- Platform: Runpod
- Monitoring Tool: Weights and Biases
- Steps: 2850
- Optimization and Learning Rate: 5e-5 with AdamW optimizer
- Batch Size: 32
- Hardware: 2x NVIDIA A6000 GPUs
Troubleshooting
As with any innovative technology, you may encounter some hiccups. Below are some common issues and their solutions to assist you:
- Sample Not Generating Correctly: Ensure that your prompts adhere strictly to the format provided above. Minor changes can drastically affect output.
- Inconsistent Key Performance: If you notice certain keys aren’t producing expected results, try switching to relative major/minor keys to see if that helps with tonal output.
- Random Prompt Not Functioning: If it seems the random prompt button is stuck, simply refresh the interface or restart the application.
For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.
Conclusion
At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.
License
This model operates under the Stability AI Community License, available for non-commercial use or limited commercial use with specific revenue restrictions. For additional details, refer to the License Info.
