digital-preservation / pronom

Simple Maven Artifacts for PRONOM Signatures
http://www.nationalarchives.gov.uk/PRONOM/Default.aspx
Other
8 stars 3 forks source link

Problem identifying fmt/199 files w h246 video and aac sound #21

Open KBspada opened 7 months ago

KBspada commented 7 months ago

Hi. The National Library of Sweden are having problems identifying certain fmt/199 files against the PRONOM registry. Namely ones with h264-video and aac-sound from certain producers, files created after 2016. The issue occurs with both DROID and FIDO. However, Siegried manages to identify them correctly (even when we change the file extension). When we produce derivative files from these files, the derivatives are possible to match against PRONOM with all tools stated above. Is this problem known? See tech details below.

DROID: Type: Binary Version: 109 File name: DROID_SignatureFile_V109.xml Type: Container Version: 20221102 File name: container-signature-20221102.xml

FIDO (with parameters -pronom_only -nocontainer -noextension): FIDO v1.6.1 (formats-v114.xml, container-signature-UPDATE-ME.xml)

Siegfried:

siegfried : 1.10.1 scandate : 2023-10-03T13:12:13+02:00 signature : default.sig created : 2023-03-23T15:09:43Z identifiers :

Dclipsham commented 7 months ago

Hello, the first thing to check is the DROID properties under Tools > Preferences. The Maximum Bytes to Scan value is set at a default 65536 bytes, however the MOOV element that MP4 identification relies on is often found deeper within the file. If you set this to any negative value and save your changes, then hit 'New' to create a new profile with the updated settings, and rescan your files then hopefully this will then identify correctly. Siegfried defaults to a full file scan and the output suggests it is finding the moov atom deep within the file.

If you're already scanning the full file, or updating the settings makes no difference, are you able to share an affected example file?

Note I do not work for The National Archives

David

KBspada commented 7 months ago

Yep, that did the trick! Eternal gratitude!