I am currently working on the new Opéra de Paris website that uses Cloudinary as an images CDN.
In order to display a copyright on each image, I am using the Cloudinary API text overlay parameter.
As an exemple, twitter:image meta contains the following url : https://res.cloudinary.com/opera-national-de-paris/image/upload/c_crop%2Ch_1409%2Cw_2500%2Cx_0%2Cy_328/c_fill%2Ch_315%2Cw_600/a_270%2Cc_fit%2Cco_rgb:fff%2Cg_south_west%2Cl_text:Helvetica_10_left:©%20Cyrus%20Cornut%20%2F%20Dolce%20Vita%20%2F%20Picturetank%2Cw_315%2Cx_5%2Cy_5/v1/user_photos/yuhfiukjubiiyz0dotrf.
That kind of url is successfully displayed by Facebook meta scraper. It was also working great with Twitter Cards before adding the copyright transformation.
Any idea how I could encode/format those urls so that Twitter cards can display images correctly ?