cosine.club

Search engines work in three separate stages: discovery, crawling/indexing, and ranking. A page can fail at any one of them, and each failure has a different fix. If your page isn't showing up, the fastest path is to check the stages in order — don't jump straight to "ranking factors" before you've confirmed the page is even indexed.

This guide walks through each stage, what blocks it, and a step-by-step diagnostic sequence you can run with free tools.

Stage 1: Discovery — How Search Engines Find Your URLs

Before a search engine can crawl a page, it has to know the URL exists. There are four main discovery paths:

• Links from other sites (external backlinks)
• Internal links from pages already known to the search engine
• XML sitemaps you submit
• Redirects and canonical signals pointing to the URL

What blocks discovery

• Orphan pages: no internal links point to them, and no sitemap includes them. These are effectively invisible.
• Sitemap errors: a sitemap that lists non-canonical URLs, returns errors, or isn't referenced in robots.txt.
• Noindex on linked pages: if the only page linking to your target is itself excluded, the crawler may never follow the path.

Practical fix

1. Add at least one contextual internal link from a page that is already indexed.
2. Confirm the URL appears in your XML sitemap and that the sitemap is submitted.
3. Check robots.txt doesn't disallow the path.

Stage 2: Crawling and Indexing — Getting the Page Stored

Crawling means the bot fetches the page. Indexing means the content is stored and eligible to appear in results. These are not the same thing — a page can be crawled but not indexed.

Common crawl blockers

Common indexing blockers (page is crawled but not stored)

• Thin or duplicate content: near-identical to another URL on your site.
• Canonical tag pointing elsewhere: you're telling the engine "index that page instead."
• Soft 404: page returns 200 but looks empty or error-like.
• Wrong canonical chosen by the engine: often caused by conflicting signals (sitemap says A, canonical says B).

How to check index status

Use a site: query in the search engine (for example, site:example.com/page) as a rough check. It's not exact, but it tells you whether the URL is in the index at all. For a more structured view, use the search engine's own webmaster console if you have one — that's the authoritative source for coverage status.

Stage 3: Ranking — Why an Indexed Page Still Doesn't Appear

Once a page is indexed, ranking depends on relevance and authority signals. The main on-page levers:

Title and headings

• The title tag is still one of the strongest relevance signals. Put the primary topic near the front.
• H1 and subheadings should reflect what the page actually covers, not keyword-stuffed variants.
• Mismatch between title and body content is a common reason a page ranks for nothing.

Content depth and intent match

• Does the page answer the question the searcher is asking? A page about "search engines" that only defines the term will lose to a page that explains crawling, indexing, and ranking.
• Cover the subtopics a searcher would expect. Thin coverage on a broad topic rarely ranks.

Internal links and authority

• Internal links pass context and relative importance. A page with no internal links is treated as low priority.
• External backlinks still matter, but quality and relevance outweigh raw count.

Technical signals

• Mobile rendering: if the mobile version hides content, rankings suffer.
• Core Web Vitals: page experience is a tiebreaker, not a primary driver, but poor performance can hurt.
• HTTPS and clean URL structure: baseline expectations.

A Step-by-Step Diagnostic Sequence

Run these in order. Stop when you find the failure point.

1. Is the URL in the index? Run site:yourdomain.com/page. If nothing appears, go to step 2. If it appears, skip to step 5.
2. Is it blocked by robots? Check robots.txt for a Disallow rule matching the path. Check the page's meta robots and HTTP X-Robots-Tag.
3. Is it discoverable? Confirm the URL is in your sitemap and has at least one internal link from an indexed page.
4. Is it canonicalized elsewhere? Check the rel="canonical" tag. If it points to a different URL, that URL is the one being indexed.
5. Is it indexed but not ranking? Compare your title and H1 against the query. Check whether the page covers the subtopics the top results cover.
6. Check backlinks and keyword position. Free tools like the ones on SmallSEOTools.com can give you a backlink overview and keyword position tracking. Treat these as directional signals, not precise measurements — free backlink and rank tools typically sample data and can differ from what a search engine's own console reports.

Common Misconceptions

"Submit the URL and it indexes instantly." Submission queues a crawl; it doesn't guarantee indexing or timing. Indexing can take hours to weeks depending on the site.

"Meta keywords help ranking." They've been ignored by major search engines for years. Don't spend time on them.

"I can guarantee a #1 ranking." No tool or service can guarantee a specific position. Rankings depend on competition, query, location, and personalization. Anyone promising a fixed position is overstating what's controllable.

"More backlinks always means better rankings." Low-quality or irrelevant links can be ignored or actively harmful. Relevance and trust matter more than volume.

"If it's indexed, it should rank." Indexing is eligibility, not promotion. A page can be indexed and still rank on page 10 because it's less relevant or less authoritative than competitors.

Quick Reference: Which Stage Is Failing?

Work through the stages in order. Most "my page won't rank" problems turn out to be discovery or indexing problems, and those are usually the fastest to fix.

cosine.club is an electronic music similarity search engine. You give it a track, and it returns other tracks that sound close to it, matched by analyzing the audio itself rather than tags, genres, or listening history. It searches a catalog of over two million electronic tracks, and per the site's own description it is free and requires no account. It is most useful when you are trying to find more music in the same sonic vein as an obscure, untagged, or barely-played record — the exact case where metadata-driven recommenders tend to fail.

How the matching actually works

The mechanism matters, because it explains both what the tool is good at and where it breaks down.

1. Audio is converted to a vector. cosine.club uses the discogs-effnet model from Essentia, a neural network trained to recognize 400 Discogs styles from two-second slices of a spectrogram. The embedding it produces is a vector of 1,280 dimensions describing how a track sounds.
2. Every catalog track already has its vector stored. That's the "over two million" figure — the comparison set is precomputed.
3. Your track is run through the same model. When you paste a track, it goes through the same pipeline, and the result is compared against the whole catalog using cosine similarity (a standard measure of how close two vectors point in the same direction).

One consequence worth knowing: because the model was trained on Discogs styles, those 1,280 dimensions encode era, place, and scene as well as raw sound. So "similar" here means similar in the way a genre-aware listener would group records — not just similar timbre.

What it is not

It is not Shazam. Shazam fingerprints a recording, which is why it survives crowd noise and works from a phone mic in a club. cosine.club matches by sound character, so identification is a side effect at best — the site itself frames identification as "sometimes" possible, not the core function.

It is also not Spotify or Last.fm. Those systems build recommendations from metadata and co-listening behavior. That works well for music lots of people already play and badly for everything else: a record needs tags, plays, and co-listening data before it can be matched at all. cosine.club reads the audio instead, so a 300-pressed white label with no tags and almost no plays sits in the catalog on the same terms as anything else.

Who it is for

• Diggers chasing a specific sound. You have one track and want more like it, including records that share no tags with it.
• Listeners into obscure electronic music. The less metadata a track has, the bigger the advantage over tag-based recommenders.
• Anyone exploring by texture rather than genre label. If you care about how a record feels more than what shelf it sits on, the audio-first approach fits.

It is less useful if you want mainstream discovery, playlist-style radio, or track identification from a noisy recording.

What you can do on the site

Beyond single-track search, the page surfaces a few entry points:

The FAQ is organized as plain text files under /help/faq/ — HOW_IT_WORKS.txt, SEARCH.txt, TRACK_DATABASE.txt, PLAYLISTS.txt, TROUBLESHOOTING.txt, API.txt, and others — which is where the technical detail above comes from.

Practical notes before you start

• No account is needed to search, per the site's description. There is a log-in option on the page, which implies account features exist, but the core search is presented as free and account-free.
• The catalog skews electronic. This is not a general-purpose music search engine.
• Expect era and scene clustering. Because the embedding carries era and place, results for a 90s techno track will tend to stay in that neighborhood rather than jumping across decades — useful if that's what you want, limiting if it isn't.
• Treat "similar" as a starting point, not a verdict. Vector similarity is a ranking signal; your ears still decide.

If your problem is "I have this one record and I want more that sound like it, and the usual recommenders keep giving me nothing," cosine.club is built for exactly that. If your problem is "identify this song playing in the background," use a fingerprinting app instead.

cosine.club matches tracks by comparing audio embeddings rather than tags, genres, or listening history. Every track in its catalog of over two million electronic tracks is stored as a 1,280-dimension vector produced by the discogs-effnet model from essentia. When you paste a track, the same model runs on it and the result is compared against the whole catalog using cosine similarity. The practical consequence: a track with no tags and almost no plays can be matched on the same terms as a popular one.

The matching pipeline

1. Audio becomes a style-aware embedding

The discogs-effnet model is a neural network trained to recognize 400 Discogs styles from two-second slices of a spectrogram. Instead of outputting a single style label, cosine.club uses the embedding the model produces — a vector of 1,280 dimensions describing how the track sounds.

Because those 400 Discogs styles encode era, place, and scene as well as sound, the vector carries that information too. A track's embedding therefore reflects not just timbre but the stylistic neighborhood it belongs to.

2. The catalog is precomputed

Every track in the catalog already has its vector stored. That's the reason search can run against the full set of over two million tracks rather than a small candidate pool.

3. Your input goes through the same model

When you paste a track, cosine.club runs it through the same discogs-effnet model to produce a vector, then compares it against all stored vectors by cosine similarity. Cosine similarity measures the angle between two vectors, so tracks pointing in a similar direction in this 1,280-dimension space are returned as the closest matches.

Why this differs from Spotify or Last.fm

The key difference is the cold-start problem. Metadata and collaborative-filtering systems work well on music plenty of people already play, and badly on everything else: a record needs tags, plays, and co-listening data before it can be matched at all. A 300-pressed white label with no tags and barely any plays has none of that. cosine.club reads the audio instead, so that white label sits in the catalog on the same terms as anything else.

What this means in practice

• You can search with audio, not text. The input is a track, not a genre name or a tag.
• Obscure material is not penalized. Matching does not depend on popularity or community data.
• Style, era, and scene are baked in. Because the underlying model was trained on Discogs styles, the embedding encodes those dimensions alongside pure sound.
• It is not Shazam. Shazam fingerprints a recording, which is why it survives crowd noise. cosine.club's approach is similarity, not identification — it can sometimes help identify a track, but not the way Shazam does.

If your goal is finding tracks that sound close to something you already have — especially underground or untagged electronic records — audio-based matching is the mechanism to rely on. If your goal is recommendations driven by what similar listeners played, a metadata or collaborative system is the better fit.

cosine.club is an electronic music similarity search engine that matches tracks by analyzing audio rather than tags. Its main sections are search, Today's Gems, Community Playlists, Journey, plus account-level entries like favorites, history, settings, and an API. You can use it without an account, according to the site's own description.

Core search: find tracks that sound alike

The central feature is similarity search. You paste in a track, and cosine.club runs it through the same model used for its catalog and compares it against every stored track by cosine similarity.

How it works, in the site's own terms:

• It uses the discogs-effnet model from essentia, a neural network trained to recognize 400 Discogs styles from two-second spectrogram slices.
• Each track is represented by a 1,280-dimension embedding describing how it sounds. Those dimensions carry era, place, and scene as well as sound.
• The catalog holds over two million tracks, each with its vector already stored.

The practical difference from Spotify or Last.fm: those systems build recommendations from metadata and co-listening data, so obscure records with few tags or plays are hard to match. cosine.club reads the audio instead, so a 300-pressed white label with no tags sits in the catalog on the same terms as anything else.

Today's Gems

A rotating selection of tracks, presented as a numbered list. The current example on the site is a 90s techno and tech house selection, with entries like Phreax – Jitterballz, Mike Parker – Druma, and Expander – Silk Screening. A "View today's gems" link leads to the full set.

Use this if you want a curated entry point without searching first.

Community Playlists

Playlists created by cosine.club users. Examples shown on the site include:

A "View all playlists" link opens the full list. This is the section to browse if you want human-curated groupings rather than algorithmic matches.

Journey

Journey lets you hear the path between any two tracks. Instead of a single similarity result, it walks from one track to another through the catalog. Useful when you want to understand how two distant sounds connect, or to build a transition between them.

Account and developer entries

The site lists these entries alongside the main features:

• Playlists – your own saved playlists
• Favorites – tracks you've marked
• History – past searches or activity
• Settings – account preferences
• API – programmatic access to the search
• Log in – account access

The site describes itself as free with no account required for search, so these entries are for users who want persistence or programmatic use.

Can it identify a track like Shazam?

Sometimes, but not the same way. Shazam fingerprints a recording, which is why it survives crowd noise. cosine.club matches by sound similarity, so it works differently and is better suited to finding tracks that sound like a given one than to naming a noisy clip.

Help and documentation

The F.A.Q section is an index of plain-text files under /help/faq/, including:

• HOW_IT_WORKS.txt
• SUBMITTING.txt
• SEARCH.txt
• TRACK_DATABASE.txt
• PLAYLISTS.txt
• TROUBLESHOOTING.txt
• API.txt
• SUPPORT.txt

If you want the exact mechanics or want to submit tracks, start with these files rather than guessing at the interface.

Not in the same way. cosine.club can sometimes surface the track you're after, but it works by audio similarity rather than fingerprinting. Shazam matches an exact recording against a fingerprint database and tolerates crowd noise, phone mics, and partial clips. cosine.club analyzes how a track sounds — its embedding — and returns the closest matches in its catalog. That means it can identify a track when the audio is clean enough and the track is in the catalog, but it is not built to survive a noisy club recording the way Shazam is.

How the two approaches differ

Shazam fingerprints a recording. A fingerprint is a compact signature of the specific waveform, designed so that even a distorted, partial, or noisy capture still lines up with the original. That is why it works from a phone in a loud room.

cosine.club instead runs your audio through the discogs-effnet model from essentia — a neural network trained to recognize 400 Discogs styles from two-second spectrogram slices. The model produces a 1,280-dimension embedding describing how the track sounds, and every track in the catalog already has its vector stored. When you paste a track, it goes through the same model and is compared against the catalog by cosine similarity.

So the match is "this sounds closest to," not "this is the same recording." The dimensions carry era, place, and scene as well as sound, because that is what Discogs styles encode.

When cosine.club will find your track

• The track is in the catalog (over two million electronic tracks).
• Your audio is clean enough that the embedding lands near the original's vector.
• The track has distinctive enough sonic character that it separates from its neighbors.

When it won't

• The audio is a noisy club recording or a heavily processed clip — the embedding drifts.
• The track isn't in the catalog at all.
• The track sits in a dense cluster of similar-sounding records, so the top matches are close but not the exact one.

What this means in practice

If you have a clean file or a decent rip and the track is in the catalog, cosine.club can often point you to it — or to something close enough to identify it by ear. If you're holding up a phone in a loud room, Shazam is the right tool. The two are complementary: Shazam for exact identification from messy audio, cosine.club for finding what a track sounds like when tags and metadata aren't enough.

The catalog is free to search and requires no account.