A studio assembled
around the work.
VTL Studio is developed with Russell Parrish / A.rtist I.nfluencer as part of Parallax Metrology. The Visual Thinking Lens supplies the inquiry and critique framework.
Contributions and tools
The project began from Alice’s Still Wet and claude-paint. Her Easel implementation was a starting instrument, not the limit of the practice. Later work explored libmypaint, native Hands/Canvas drawing, and learned pose and body reconstruction. OpenAI image generation and user-supplied Midjourney/ChatGPT images appear where identified.
The winter brook, cleaners’ window and field/town trials adapted subjects from Still Wet’s public briefs. Their run record states that Alice’s reference paintings were not supplied to the model. The rendered trials are our local run outputs.
How to read the images
Source, native drawing, generated return, diagnostic and rejected branch are different roles. Native drawing may be guided by a source image. A generated raster does not come with the native instrument’s editable mark history. Film sequences document image stages rather than inventing unseen brushwork.
Public edition
This edition preserves visual process and case-study explanations. Internal operating prompts, full protocol and skill texts, raw logs, model files and implementation downloads remain outside the public package. Summaries of methods are included; the full working framework is not published here.
User-supplied starting images are included with Russell’s confirmation of usage rights. Third-party tools and inspirations retain their respective rights. Inclusion assigns no new license to their code or material and makes no blanket claim of exclusive copyright over AI-generated imagery.
Limits stay visible
Source dependence, anatomy errors, over-regular marks and the attraction of familiar symbols remain active problems. The portfolio documents specific experiments. It does not establish superiority over other systems or a general measure of originality.