Neatbo.

Save original embedded audio pictures

Keep embedded picture bytes, audio originals and every provenance record without screenshots or reencoding.

Choose originals and extract the entire queue

Select audio originals with embedded artwork, recover every eligible picture in encounter order and download the complete original/picture/report ZIP.

1–4 native FLAC or MP3 files: 50 MiB per file and 100 MiB total; UTF-8 filenames up to 512 bytes. Preserve originals and use the files you own.

Any refusal stops the entire queue with no partial ZIP. Correct the source or choose supported originals and retry. After cancellation, rerun the same selected files.

  • Keep the audio originals, then select the whole queue before starting extraction.
  • Download pictures.zip and open pictures.json to match each picture with its source and offset.
  • Open recovered pictures in an image reader; exact extracted bytes can still belong to an image that will not open.

Open the complete deliverables

Numbered pictures follow selected-file order and encounter order inside each source. Identical payloads still have separate rows. The paginated screen contains every picture; all structural records remain in JSON.

Delivered contents
ArtifactUsePreservation
originals/file-NNNN.flac or .mp3Audit laterEvery original audio byte
pictures/picture-NNNN.png or .jpgSave a separate assetExact original payload
pictures.jsonAudit full provenanceSource index/name, offsets, raw descriptions, every structural row and hashes
pictures.csvRead every picture in a spreadsheetText has a leading apostrophe; exact values remain in JSON

Separate declarations from decoded results

Image and audio codecs are not decoded. A signature-matching damaged image is recovered byte for byte and may fail to open. FLAC dimensions/depth/colours are source declarations, not verified image dimensions; no audio CRC/PCM/playability claim.

The example FLAC declares its back picture as 999999×0, while the original image is 9×6. The tool preserves that declaration and the 677-byte JPEG. It neither resizes from the declaration nor presents independent test decoding as its own measurement.

Recognize finite support and duplicate rules

Native FLAC PICTURE and prepended ID3v2.3.0/2.4.0 APIC only; PNG/JPEG MIME and byte signature pairs. APIC png/jpeg shorthand is supported. Front/back and batch pictures share one extraction task.

At most 128 pictures, 4 MiB per payload, 16 MiB total picture bytes, 4096 metadata records, 4096 raw bytes per description and 262144 raw description bytes overall. ID3v2.3 descriptors also allow at most 64 characters.

FLAC descriptions use strict UTF-8. APIC supports Latin-1 and BOM UCS-2 in v2.3; Latin-1, BOM UTF-16, UTF-16BE and UTF-8 in v2.4. Literal text and description raw hex stay in JSON.

Unknown/reserved metadata, linked pictures, unknown frames, nested/extra tags, duplicate APIC descriptions or duplicate icons refuse the whole queue. Repeated ordinary FLAC pictures, including identical payloads, remain separate.

Unsynchronisation, compression, encryption, grouping, data length indicators, extended headers, experimental headers and footers are unsupported even when valid. No de-stuffing, repair or silently skipped tags.

Check budgets and report safety

Download originals, every numbered picture and full pictures.json/pictures.csv in one pictures.zip. JSON and CSV allow 4 MiB each, ZIP 128 MiB; stricter input limits dominate these protective output guards. CSV textual values receive an apostrophe; JSON/raw hex preserve exact values.

No upload, remote URI fetch, resizing or reencoding. Any malformed, unsupported or over-budget input produces no partial archive; cancellation discards unfinished output.

Source descriptions render as ordinary text without executing HTML. Every CSV textual value receives an apostrophe even before a literal BOM or hidden control prefix. Use JSON and descriptionRawHex when exact text matters.

References

Tools in this category

Expand a tool to see its steps, options and supported formats, then open its workspace.

Extract embedded audio picturesRecover every embedded PNG/JPEG payload from local FLAC/MP3 without reencoding, with originals and complete JSON/CSV ZIP.

Select audio originals with embedded artwork, recover every eligible picture in encounter order and download the complete original/picture/report ZIP.

Steps

  1. Select 1–4 audio originals.
  2. Run and inspect every picture type, description, offset and hash row.
  3. Download pictures.zip and open numbered pictures.
  4. Audit full JSON/CSV provenance and keep original audio.

Capabilities and limits

  • 1–4 native FLAC or MP3 files: 50 MiB per file and 100 MiB total; UTF-8 filenames up to 512 bytes. Preserve originals and use the files you own.
  • Native FLAC PICTURE and prepended ID3v2.3.0/2.4.0 APIC only; PNG/JPEG MIME and byte signature pairs. APIC png/jpeg shorthand is supported. Front/back and batch pictures share one extraction task.
  • At most 128 pictures, 4 MiB per payload, 16 MiB total picture bytes, 4096 metadata records, 4096 raw bytes per description and 262144 raw description bytes overall. ID3v2.3 descriptors also allow at most 64 characters.
  • FLAC descriptions use strict UTF-8. APIC supports Latin-1 and BOM UCS-2 in v2.3; Latin-1, BOM UTF-16, UTF-16BE and UTF-8 in v2.4. Literal text and description raw hex stay in JSON.
  • Unknown/reserved metadata, linked pictures, unknown frames, nested/extra tags, duplicate APIC descriptions or duplicate icons refuse the whole queue. Repeated ordinary FLAC pictures, including identical payloads, remain separate.
  • Unsynchronisation, compression, encryption, grouping, data length indicators, extended headers, experimental headers and footers are unsupported even when valid. No de-stuffing, repair or silently skipped tags.
  • Image and audio codecs are not decoded. A signature-matching damaged image is recovered byte for byte and may fail to open. FLAC dimensions/depth/colours are source declarations, not verified image dimensions; no audio CRC/PCM/playability claim.
  • Download originals, every numbered picture and full pictures.json/pictures.csv in one pictures.zip. JSON and CSV allow 4 MiB each, ZIP 128 MiB; stricter input limits dominate these protective output guards. CSV textual values receive an apostrophe; JSON/raw hex preserve exact values.
  • No upload, remote URI fetch, resizing or reencoding. Any malformed, unsupported or over-budget input produces no partial archive; cancellation discards unfinished output.
Open Extract embedded audio pictures →