Added support for 32 and 64bit float files

[gee-ell]

I've added support for 32 and 64bit floating point WAV files inputs - modified 1.33 ssrc.c: https://desperatebeauty.com/_files/ssrc_float_support/ssrc.c

• only tested on Windows (byte order from simple casting may not be correct on other platforms (search for 'case 8' and 'case 4' to correct if needed).

• only tested with Reaper 6 and Wavelab 8 generated 64bit float files (my use case).

• handles 'JUNK' chunk that Wavelab 8 inserts before the 'fmt ' chunk.