Choose ten hard retrieval tasks
Include three badly named files, two loops with BPM/key requirements, two “same but different” replacements, two arrangement complements, and one blend between references.
Reviews and buyer evidence
Do not buy Sample because we praised Sample.
There are not yet enough verified public customer ratings to publish an aggregate score. We would rather show zero than manufacture five stars. Use this page to test the product claim against your own library.
0fabricated testimonials or synthetic star ratings published here
30 minrecommended buyer evaluation using your own difficult searches
14 daysmoney-back window after purchase; email support to request it
10retrieval tasks in the scorecard below—enough to expose weak fit quickly
Review status
No aggregate rating is displayed because genuine visible ratings have not yet reached a defensible sample size. When reviews are published, incentivized access will be labeled, criticism will not be removed for being negative, and structured rating markup will match visible data.
The skeptical buyer’s 30-minute test.
Use sounds you routinely fail to find. Do not build the test around Sample’s demo queries.
Establish the current baseline
Use Finder or your existing browser. Start a timer for every task. Count previews until the first acceptable result—not just the first click.
Run the same ten tasks in Sample
Use text, similarity, refine, complementary, and blend where appropriate. Keep failed tasks in the score rather than rewriting them away.
Make a purchase decision from evidence
Record median retrieval time, accepted results, previews per result, and subjective interruption. If Sample does not win on your library, use the refund policy.
A scorecard harder than five stars.
Copy this table into a note. Score each task 0–2: failed, usable, or clearly better.
| Test | Pass condition | Record |
|---|---|---|
| Bad filename | Usable result without knowing the vendor’s words | Seconds + previews |
| Exact filename | Known token remains easy to retrieve | Seconds |
| Texture search | Top results reflect an audible character | Top-10 relevance |
| BPM query | Explicit tempo meaningfully narrows eligible loops | Correct/incorrect |
| Key query | Explicit key filters reliably eligible tonal audio | Correct/incorrect |
| Similar replacement | At least one top candidate can replace the reference | Rank of winner |
| Refined replacement | “More/less” produces audible directional movement | Blind preference |
| Complement | Candidate supports rather than duplicates the source | In-track preference |
| Blend | Candidate shares useful qualities with both references | In-track preference |
| DAW handoff | Chosen audio reaches the timeline without path friction | Clicks + seconds |
Evidence available before reviews.
None replaces independent use, but each claim is inspectable.
Actual interface screenshots
Search, refine, blend, and drag-and-drop imagery on this site comes from the shipping product—not conceptual UI.
Inspect search →Measured engineering probes
The benchmark page gives hardware, runs, distributions, limitations, and the next missing measurements.
Inspect benchmarks →Repository-backed change history
The changelog condenses dated product work across search reliability, watched folders, offline auth, updates, and performance.
Inspect changes →Independent reviewer brief.
Reviewers receive access, not editorial approval. Suggested disclosures are explicit.
What to disclose
- Whether the license was provided
- Library size and drive type
- Mac model, memory, and macOS version
- Analysis duration and excluded formats
- Any affiliate relationship
What to challenge
- Large-library startup and background analysis
- Weak filenames and non-English pack names
- BPM/key false positives
- Similarity diversity, not only obvious wins
- Memory, battery, offline use, and path changes
One-time purchase · $129 launch price
Give Sample ten difficult searches. Keep it only if it wins.
Sample searches and analyzes locally on Apple Silicon. Your audio stays on your Mac. Use it offline after activation, drag results into any DAW, and request a refund within 14 days if it does not improve your workflow.