Props to you for going lossless.
I settled on mp3s at 44.1khz sample rate and 320kbps bitrate after reviewing the practical differences between that and lossless and then considering space and expense. Best I could do.
That's an
entirely reasonable balance between quality and resource consumption. One of my Linux programs describes that bitrate as "insane" but I don't need my applications to be judgmental! I use those parameters all the time.
I've tried various devices over the years for recording. Those that have a built-in FM radio that can then record directly on the device inevitably have some sort of limitation, aside from FM reception capability generally. For example, Sanza Fuze devices record in WAV format with a 24 kHz sampling rate, which limits frequency response to 12 kHz. For most FM stations, that's probably OK, and likely no worse than the limitations of cassette tapes of an earlier era, but it's still annoying to deal with. But at least that can be compressed further if needed. Some other devices that come out of the Chinese radio-industrial complex have limitations such as brickwall low-pass filters (ranging from 10 to 14 kHz) and, in one case, recording in MP2 format rather than MP3! Best results have come from Tascam recorders. I'm using the DR-05X and, most of the time, recording in WAV format. Battery life is not so great but better than its predecessor, the DR-07, which eats batteries for breakfast, lunch, and dinner all at once, but has more flexible input options. There always seems to be some kind of trade-off.
The Tascams give me good control over levels. I keep a peak limiter on, but at input levels where it rarely triggers. I can normalize levels in post-recording processing, which I do in the Audacity editor.
For most of the recordings I've made with the Tascam recorders, I've also saved the raw WAV files along with a log that indicates what's in each file. This requires an SSD of its own. I'm currently using Samsung T7 SSDs. They are amazingly fast. I also try to have multiple backups; the jury still seems to be out on the longevity of data stored on SSDs, though I've never run into problems.
I would love to use a newer compression format such as AAC or Ogg Vorbis (bonus: Ogg Vorbis uses the same metadata format as FLAC, but support seems limited to the open-source world), which could cut resulting file sizes by half compared to MP3, but MP3 seems to be the lowest common denominator that you can use on almost any platform.
My recent European batch consists mostly of recordings from DAB+, which have varying bitrates from 48 to 112 kbps (AAC), depending on country and allocation. I won't save lossless versions most of those because the broadcast source was already lossy. I feel similarly about HD subchannels and AM HD.
I haven't been good about keeping a catalog of recordings. Every processed recording gets tagged with metadata, so there is some degree of self-documentation. Someday, I'll bestir myself to write a script to read the metadata and compile it into a format that I can import into a spreadsheet. But that requires time, too, and my scripting skills are rusty.
I've been learning as I go along. Thus I'd like to re-do some of my earlier transfers from cassette tape, made more than five years ago by now, because I have a much better idea of how to get reasonable quality from the tapes than I did when I started.
