Full page Federal Tax document is not working

Hi Team, We have been trying to work with Federal Tax Returns 1040 documents using Document Analysis API. We tried with different document samples and made an observation that if the FTR is a full page then the queries are not working properly, however If I try cropping the same FTR into half or until the Dependents section is seen in it, then I could see the extraction from queries is happening without any issues. Could someone please suggest why it is causing that when the FTR document is full page it doesn't work. Could you please suggest us a better way of working with FTRs Thanks Kartheek