What changed
Shipped in this update
- Guide hub added so each analysis area has a clear explanation page.
- Analysis-result tabs now link back to the matching guide.
- AI-origin detection copy now frames the score as probability, not proof.
- Genre intent, feedback prompts, SEO recovery checks, and security maintenance were tightened.
What changed this week
This update is based on product work now visible on Mix Analyzer: the browsable guide hub, expanded analysis guides, result-to-guide links, beta AI-generated music detection guidance, and supporting UX, SEO, and security fixes. Mix Analyzer remains a browser-based mix feedback tool, not an AI mastering service.
- The guide hub gives producers one place to browse analysis topics.
- Result pages now point technical flags back to practical explanations.
- AI-origin checks are framed as probability signals, not proof of authorship.
A browsable guide hub for audio analysis topics
The guide hub gives producers a clearer path from a technical result into a usable explanation. If the analyzer flags low-end buildup, weak punch, narrow stereo width, playback risk, or AI-origin uncertainty, the site now has a stronger educational layer around the same checks.
- Frequency spectrum analysis for tonal balance, masking, and sibilance.
- Dynamic range analysis for punch, crest factor, and over-limiting.
- Stereo field analysis for width, phase, and mono safety.
- Playback compatibility analysis for translation to phones, earbuds, cars, and small speakers.
- AI-generated music detection for reading a probability signal without treating it as proof.
Result pages now point back to explanations
A mix feedback tool should not stop at "your mix has an issue." The useful workflow is: notice the flag, understand the production problem, make one revision, then compare. The result-page guide links are meant to keep that loop tight.
- Start from the result tab with the issue you can also hear.
- Read the linked guide before changing plugins.
- Re-export once and compare the second result against the first.
Beta AI-origin checks are framed carefully
Mix Analyzer now includes beta AI-origin style checks and guide copy for AI-generated music detection. The important wording is probability, not verdict. This can be useful as a review signal, but it should not be presented as legal-grade proof of authorship.
- Use high scores to decide where to listen closer.
- Do not use the score alone to accuse a creator or reject a track.
- Keep disclosure, copyright, and platform policy decisions separate from the detector output.
Genre intent and producer feedback got cleaner
Genre context changes how analysis should be read. A club track, acoustic vocal demo, and dense rock mix do not need the same target. Recent work improved the genre flow, feedback prompts, and the way Mix Analyzer asks producers to judge one decision at a time.
- Pick the genre that matches the release goal, not only the closest detector label.
- Use reference-match and playback checks to validate the target.
- Keep intentional choices when they serve the song and still translate.
Security, SEO recovery, and reliability still matter
The product work also included recovery checks after the nonblocking cookie-consent fix, cleaner article discovery paths, and tighter security around generated content and draft storage. Those changes are less visible than a new guide page, but they decide whether the site is crawlable, readable, and safe to operate.
- Cookie consent stays nonblocking and does not blur page content.
- Indexable blog and guide URLs are kept explicit in sitemap and llms.txt.
- Generated article HTML is sanitized before rendering.
What this changes in the actual workflow
The useful path is now shorter. A producer can upload a bounce, see a technical flag, open the matching guide, and make one focused revision instead of guessing what the score means. That matters most for problems that sound vague in the room: low-mid buildup, weak punch, phase risk, dull source quality, and translation on small speakers.
- Use one result tab as the starting point, not the whole report at once.
- Open the linked guide to understand the listening problem behind the metric.
- Change one mix decision, export again, and compare the second result.
What the AI-origin check can and cannot say
The AI-origin module is intentionally framed as a probability signal. It can flag patterns that deserve human review, especially on synthetic vocals or fully generated tracks, but it cannot prove authorship. Mastering, resampling, vocoders, heavy quantization, and new model families can all change the score. Treat the result as a review cue, not a legal conclusion.
- Use high scores to decide where to listen closer.
- Do not use the score alone to accuse a creator or reject a track.
- Keep disclosure, copyright, and platform policy decisions separate from the detector output.
How to use genre intent without flattening the song
Genre intent gives the analysis a target, but it should not erase taste. A club record can carry more low-end weight than an acoustic vocal demo; a dense rock mix can tolerate more midrange energy than sparse electronic music. The improved genre flow helps the analyzer read those tradeoffs without turning every track into the same generic target.
- Pick the genre that matches the release goal, not only the closest detector label.
- Use reference-match and playback checks to validate the genre target.
- Keep intentional choices when they serve the song and still translate.