When saving a waveform to a WAV file, either through the memory list or an autosave step, SoundCheck will normalize the waveform to full scale. The reason for this is that once you create a WAV file, the physical units (V, Pa...etc.) are lost, and it exists as digital information. There are times however, when you would like to store two waveforms and retain their relative levels - here is a way to get around the normalization.
If the units of your waveform are FS (digital full scale) SoundCheck will not normalize the saved WAV file, so the easiest way to store your waveform without normalizing is to simply change the units. This can be done manually through the memory list (Memory -> Units) or automatically in a sequence by using a simple post processing step such as Curve Multiplied by Constant (use a constant of 1 and then set the units box to FS).
There is, however, one caveat. The maximum value of Full Scale is 1. So if your original waveform has any values that exceed 1, they will simply be clipped off when you save to a WAV file. The way to avoid this is to scale the waveform by dividing it by a value that ensures all data values are less than 1. Provided that you use the same constant each time your waveforms will be scaled relative to each other.