The ‘doom loop’ where it will transcribe one character or phrase several times will appear often when an image is inserted upside down or sideways, but strangely enough it can still transcribe a large portion of the letter accurately. I think a potential way to fix this would be to make it so that you can rotate an image within the program (so you don’t have to open up the pdf, rotate it, and then reupload, and then retranscribe which uses credits).
1 Like
We’re planning on introducing image manipulation features (including cropping, rotating, inverting) within the web-app. Hopefully this will be released within a few weeks. 
2 Likes
