Okay, purposefully not posting in the Sound Science forum. Disregard whether I can even tell the difference between lossless and `lame -V 0`. Likewise, I don't care if 192/24 is the only way to get the most of my music. Thx.
I enjoy keeping some audio around as lossless, but I can't justify the disk space for anything over 44.1/16. That said, when I come across something with a higher sample rate and/or bit depth, I prefer to convert it. However, recent reading has me worried that I might be doing so incorrectly (or at least sub-optimally). For reducing bit depth, evidently I should dither, but I don't know what dither scale or method I should use. (For the latter, I assume it comes to personal preference?) For reducing sample rate, I have no idea what I should be doing beyond "here's a 48kHz file, give it back as 44.1kHz", if anything.
If you want to know my real world situation, usually WAV or FLAC files come into my hands (from SoundCloud or Bandcamp), and I use either the reference FLAC of FFmpeg CLIs for transcoding (although SoX has recently appeared on my radar). I usually type something like this into the command line:
ffmpeg -i input.file -compression_level 12 [-ar 44100 -sample_fmt 16] output.flac
Thanks in advance.