There are two sets of 2hr 20min linked WAV files (nature sounds) to manipulate:
A: A mono center channel WAV of the nature scene (shotgun mic)
B: A synchronized stereo WAV of the same nature scene made using the same WAV recorder.
C: A completely unrelated stereo WAV of a stream which is only 2 hours long.
These are weak nature sounds with some recorder and ambient noise.
There are several distinct tasks in this project:
1) Clean up the first two sets of WAV files (A,B) to eliminate electronic and mechanical noises (such as cars and aircraft) by splicing (and/or band equalizing) to recover as much usable sound track as possible in both channels. Leave all desired sounds which will be described later. The final clean version WAV files should be 120 minutes in length. Some sections can be repeated to fill out the time, so long as these replicated sections are not detectable. The two sets of files must be precisely time synchronized so as to be playable when mixed together.
2) Combine the two sets of WAV files (A, B) into one new L-R stereo WAV, distributing the mono WAV (A) among the L-R stereo channels (B) with gain adjustment to provide balance and reasonable gain matching for a pleasant, balanced headphone experience. The addition of (A) to (B) should not overwhelm (B).
3) Bandwidth limit the low end result of (2) as necessary to eliminate residual mechanical noises and wind noise.
4) Enhance stereo separation of the result of (3).
5) Apply very mild compression on the result of (4) without introducing audible phase and gain artifacts when heard in high quality headphones.
Save out the result of (5) as a reference stereo WAV.
6) Take the result of (5) and add in WAV file (C) at a low level to appear as a very distant background sound, spatially located slightly off-axis.
Save out the result of (6) as the final work product. Convert this file also to MP3.
These files are too large to upload and will be provided by google docs.