Many of the items I place in DT3 are images (e.g. maps, photos, timelines, etc). I know that classifying them via See Also & Classify is not possible since there’s no text to use—however, if one puts some text in the Finder Comments or sets up some metadata, that text should also be used for this, right? Alas, for me that doesn’t work. I’m never able to classify images via inspector.
Is there a way to achieve this?
See also & classify support images but it’s limited to the same database. Other metadata or comments don’t matter in this case, only the image is used.
Hi @cgrunenberg, how exactly does see also function works on images containing no text? Can you please share which factors are taken into account by DEVONthink? Also, which method are you suggesting us to get better suggestions: Keeping images as is or OCR to PDF? Thanks!
The thumbnail is considered with media files, e.g., images.
Are you saying visual resemblance is taken into account?
And again, which method do you suggest for getting better results? Keeping images as images or OCR to PDF?
Thanks
Yes the image itself is compared.
I can’t answer the second part without knowing what content is in the file. If it’s a picture of a dog, doing OCR won’t matter.