You have probably stumbled upon the difficult task of a simple fade out (or in) in mastering (and/or mixing). It sounds as if there are high- and a low-pass filters that are slowly closing the frequency spectrum. This got me thinking of other ways to improve the flaws of the fade.
Maybe you have stumbled upon this when mastering (or mixing) to loud or to silent and when adjusting the volume some frequencies has been to exaggerated or subdued. This is because the human hearing isn’t flat. Our perception of sound pressure is different over the frequency spectrum and it differs depending on sound pressure.
This is an audio (and visual) visualization of three different bit reduction conversions from 24bit to 16bit audio. With / without dither and noise shaping.
I often hear people describing the attack time of compressors in the completely wrong way. I hear this everywhere, from home studio enthusiasts to professional studio owners and music educators. Maybe it’s just bad communication but I thought that it was worth looking into.
We use a 3M M79 mastering magnetic tape recorder in the studio. This example is the reel to reel 2-track version made for mastering with input and output transformers.
I did some in depth research and testing of the EQ algorithm with the LANDR website and some sinus tones. The tests are made with mp3 quality files but I still believe you can do a relevant reading out of the result, the most relevant reading being the handling of peaks.
Every mastering facility needs a great stereo/bus compressor with that nice smoothing effect to use as glue for the sound. This compressor is a further development of the G-series compressor made famous by the SSL (solid state logic) company.