Skip to main content
Question

"Content Language" Metadata Auto-Detection on Image Uploads?

Forum|alt.badge.img+1

Has anyone recently encountered a Content Language field being automatically populated when uploading image assets (PNG/JPG) into Seismic?

Yesterday we observed several chart images being assigned unexpected language values (for example, languages outside our organization's standard language taxonomy). The values appeared under a Content Language field rather than our existing custom Language metadata field.

A few observations:

  • The language values did not appear to come from our controlled language metadata list.
  • The field seemed to be populated automatically by the system rather than by the uploader.
  • These were primarily chart/image assets rather than text-heavy documents.
  • We have imported similar assets in the past and do not recall seeing this behavior.

We're trying to understand:

  1. Is Content Language a Seismic-managed system field?
  2. Does Seismic perform OCR or language detection on uploaded images?
  3. Has anyone seen changes to language auto-detection behavior recently?
  4. Can this feature be disabled or overridden?
  5. Are there any known interactions between Content Language and LiveDocs or content retrieval?

Any insights, documentation, or similar experiences would be greatly appreciated.

Thank you!

    2 replies

    Forum|alt.badge.img+1
    • Jimmy
    • Author
    • Quake
    • October 13, 2026

    Here is an image with the dropdown of values for the field for reference. 

    • GraceR
    • Quake
    • October 23, 2026

    This is what I've experienced:

    1. Content Language is a Seismic-managed system field.
    2. Seismic performs language detection on uploaded content. If no language can be detected from the asset (such as text within an image or audio content), it typically defaults to your platform's default language setting (for example, English).
    3. I haven't personally observed any recent changes to Seismic's language auto-detection behavior.
    4. I'm not sure whether the feature can be disabled or overridden, as I haven't had a need to test this.
    5. I haven't seen any known interactions between Content Language and LiveDocs or content retrieval, but I also haven't specifically investigated that area.

    In our environment, we've occasionally seen the Content Language field populated automatically, but it hasn't generally caused any issues. Based on your description, it does sound like Seismic's language detection may have interpreted content within the uploaded images metadata and assigned a value accordingly.

    I'd be interested to hear if anyone else has seen similar behavior recently or has additional insight into whether the detection logic has changed.

      Terms & ConditionsAccessibility statement