So I have a static field of sprites that a camera is passing through, and I'm trying to figure out how to apply depth of field. Most DOF info on the web has been non-hardware render related, but I don't really have a choice, since I'm using sprites. Hardware rendering only

I've tried rendering in different output formats like .tiff, .rla, .iff. With formats like .tiff and .rla, I would batch render and get the main pass as well as a depth image for each frame, but the only program that can open those depth images is fCheck. It correctly displays the grayscale depth coloration, though the Z view shows my sprites as cards...not as the images they contain. I tried saving that Z image out and using it in AE as a luma matte for the main render, but of course it didn't work in the comp since they're grayscale cards, not images.
I tried the technique listed here by yenvalmar: https://forums.cgsociety.org/archive/.../t-314991.html
...but I never saw a second alpha with Photoshop, even when using software rendering of a polygon object.
I have a very simple understanding of expressions and MEL...not much of a code monkey.
Need help!
(sorry to seem like the guy that joins/starts a thread because I have a question...but I do.)
edit: here's a shot of the sprites I'm working with, and the zDepth result I get from fCheck, both at 25% scale (if it helps)