Lots of audio shenanigans, morse only on the right channel and some definite sections of vocals, but I am having a hard time isolating them and it seems like they are additionally processed to be harder to understand.
The vocals seem to appear twice, once towards beginning ~0:16-0:40 and again ~1:10 - 1:30 ... the segments sound similar but it is hard to tell for sure. I will try later to run analysis on the sections each in its own channel to see if they correlate, could indicate if vocals are same sample used twice or different each time.
edit: short clip which somewhat isolates the vocals but they're still not clear ;
https://clyp.it/5jidyuqt
edit2: Attached better isolates of the forwards & reversed morse code spectro for comparison to by-ear decodings. These can be read as dots and dashes and contain only the right channel audio as the left channel does not have morse on it. If the WPM is constant, lines could be inserted into the image at a constant interval in order to determine word breaks.