Dropbox.filesDownload returns corrupt data for Word Documents

When I use the filesDownload function to retrieve a Word Document, the file writes successfully, but with missing bytes (rendered in a text editor as ?). I’ve tried to decode the data into the file from binary to utf-8 to hex to base64 With no luck. Binary and UTF-8 encoding seem the closest to the original file bitwise, but I am nonetheless getting corrupt data back.

dbx.filesDownload({path: path})
      .then(function (response) {
         saveFile(null, response.fileBinary));
      })
      .catch(saveFile);

Note: this is on a Node Server.

[t-hg]

Same issue here with PDF

v 2.5.+

%PDF-1.3
%âãÏÓ
[...]

v 3.0.4

%PDF-1.3
%����
[...]

[kyle-west]

Here is an example line from the file when I open it up in a text editor and compare it to the original:

Original

l"%3÷Þ3VƃÑÚšl	µw%ë=–“^i7+Ù×ä-d&á”0ÞAÉ6€l4¼½L6
0#µÃ’ÍS

Response from filesDownload

l"%3��3Vƃ�ښl	�w%�=���^i7+���-d&��0�A�6�l4��L6
0#�Ò�S

As you may see, there are missing values in this line that are not in the original.