Vision API issue detecting text in small images

[Giovanni Marazzi]

I have been using Vision api to detect text in document tif images and i tried using the 2 available processes (async request to detect text in files stored in google storage in GCP and sync request to detect text in images). While the detect text in images works for both large images (around 5 mb) and small images (around 12 kb), detect text in files return invalid file format for the small images. Our team wishes to use async requests but we were unable to get around this issue. Any help is welcome