Define the receiver before choosing a codec
The useful question is not whether an old file can be converted. It is which exact file the next player or processor will accept. Keep the source, collect the target's container, codec, sample-rate, bit-depth, channel, and profile requirements, and test with a representative clip before touching a production archive.
Inspect the stream, not just the extension
Start with audio-format-info. A WAV file normally carries PCM, but WAV is a container rather than a guarantee of one sample format. A file can therefore have the expected extension and still fail because its PCM packing, rate, bit depth, channel layout, or embedded stream is outside the receiver's support. Record the mismatch before converting.
Pick a narrow path
Use AMR or GSM only when a telephone, voicemail, handset, or gateway contract asks for that codec. For general processing, MP3 to WAV provides an uncompressed PCM/WAV hand-off, while WAV to AAC, MP3, OGG, or Opus is appropriate only when the delivery environment names that codec. WAV to FLAC is the better branch when lossless compression is wanted.
For sample compatibility, change one parameter at a time: resample for the required rate, change sample format or reduce bit depth for the required PCM representation, dither when reducing depth is appropriate, and remap channels only for a known mono or stereo contract. If DSD or DTS is unsupported, use DSD to PCM or DTS to AC3 according to the destination rather than guessing.
Accept the delivered file
Run audio-format-info on the output and compare every required field with the target contract. Then open the exact output in the real player or processing stage and check for decoder errors, silence, clipped duration, missing channels, and unexpected metadata changes. Link this record with audio-convert for broader format work, audio-analyze for evidence, audio-channel-routing for channel behavior, or audio-utility for adjacent handling.